Posts

Why More Traffic Won’t Fix Your Growth Problem (But This Will)

/0 Comments/in /by

Key Takeaways

  • Conversion rate drives more revenue impact than traffic volume, but most teams still spend their effort chasing clicks instead of fixing the funnel.
  • More impressions don’t guarantee more clicks, more sessions don’t guarantee more conversions, and more leads don’t guarantee more revenue.
  • Conversion loss usually traces back to four sources: messaging, friction, trust, and experience problems.
  • A four-step CRO framework (research, prioritize, test, scale) keeps testing systematic instead of reactive.
  • People convert when uncertainty gets removed, not when a page simply looks nicer.

Have you been pouring more budget into traffic, only to watch revenue stay flat? You’re not alone, and it’s not a traffic problem. According to NP Digital research, for businesses generating at least $10 million in annual revenue that invest in both conversion rate optimization and traffic generation, conversion rate accounted for 68 percent with website traffic just 32 percent. Yet 93 percent of teams still put most of their effort into driving more traffic, and only 7 percent focus on conversion rate.

Most companies don’t have a traffic problem. They have a conversion problem they haven’t identified yet, and that gap is exactly where competitors are gaining ground.

The Traffic Trap

It’s easy to fall into over-fixation on traffic. Often, more traffic feels like progress. Impressions climb, sessions go up, and the dashboard looks busy. None of that guarantees revenue follows, though.

NP Digital’s research found three assumptions that consistently trip up growth teams: more impressions doesn’t mean more clicks, more sessions doesn’t mean more conversions, and more leads doesn’t mean more revenue. 

Bar chart comparing revenue impact and team focus for website traffic versus conversion rate

Source

Each step in that chain has its own leak point, and pushing more traffic into the top of the funnel just sends more visitors through the same broken experience.

Meanwhile, the cost of acquiring that traffic keeps climbing. Cost per click is rising across every major ad platform, competition for the same high-intent keywords keeps intensifying, and the number of touchpoints a buyer needs before converting keeps growing. Organic click-through rates are dropping as AI answers replace traditional search results, and attention spans on landing pages keep shrinking. Growth is getting harder the more you spend, which is exactly why the acquisition side of your funnel has a ceiling that conversion optimization doesn’t.

Why Conversion Is the Highest-Leverage Layer

At some point, acquisition spend hits a ceiling. Costs climb, and the extra traffic you’re buying tends to skew toward lower lead quality, so you end up with diminishing returns on the surplus. Conversion doesn’t have that same ceiling.

NP Digital data backs this up. Companies running fewer than five tests per quarter saw 2.7 percent revenue growth. Companies running 11 to 15 tests per quarter saw 5.6 percent, more than double. That dip points to a sweet spot of roughly four tests a month. Beyond that, teams don’t have enough time to review results and build on what they learn before the next test launches, so the extra volume stops paying off.

Bar chart showing revenue growth by number of experiments run per quarter.

Source

The reason is structural. You can only bid so high on a keyword or an audience before the math stops working. A landing page, a checkout flow, or a lead form has no such ceiling. Every percentage point of improvement compounds across every channel already sending you traffic, which is why testing your way to a higher conversion rate tends to outperform simply increasing spend.

The Four Main Sources of Conversion Loss

Most conversion problems aren’t hidden. Teams just aren’t measuring the right things. When NP Digital audits funnels across industries, the drop-off almost always traces back to one of four categories.

  1. Messaging problems. Visitors land on a page and can’t quickly answer three questions: What do you do? Why are you different? Why should I trust you? If a visitor has to dig for the answer, they leave before finding it.
  2. Friction problems. Long forms, too many steps, slow page speed, and complicated checkout flows all add resistance at exactly the moment a visitor is deciding whether to commit.
  3. Trust problems. Reviews, testimonials, case studies, and social proof do the work of convincing a stranger to hand over their money or their information. Skip this step and you’re asking for a decision the visitor doesn’t have enough evidence to make.
  4. Experience problems. Poor mobile UX, weak navigation, and confusing journeys pull visitors away from the path you built for them, even when your messaging and trust signals are solid.

NP Digital data found B2B teams lose the most ground between leads and marketing-qualified leads, at a 59.4 percent drop-off, while B2C teams lose the most between add-to-cart and checkout, at 32.8 percent. 

Funnel chart showing drop-off percentages at each stage for B2B and B2C conversion paths.

Source 

Where Conversion Problems Actually Happen

Average conversion rates vary wildly by page type. Checkout pages convert at 26.8 percent on average, lead generation pages at 4.1 percent, product pages at 2.0 percent, contact pages at 1.6 percent, and homepages at just 0.02 percent.

That split is more telling than it looks. Product and contact pages should, in theory, convert well, since the visitors landing there already want a specific product or want to talk to someone. Instead they’re among the lowest-converting page types on the list. And while a 26.8 percent checkout conversion rate sounds strong on its own, it also means nearly 75 percent of visitors who reach checkout still abandon their order, so something in that final step is working against you.

In our findings, the biggest gains from testing come from the middle of the funnel, at 49 percent for B2B and 43 percent for B2C, not the bottom. 

Chart showing average conversion rate by page type and where marketers see the biggest A/B testing gains.

Source

That’s a useful correction. Most teams focus their testing energy on the bottom of the funnel, since that’s where the purchase happens. But the conversion problem showing up at the bottom is frequently created higher up, in the messaging and targeting that brought the visitor there in the first place.

Conversion Is a Human Problem Before It’s a Marketing Problem

Most CRO conversations focus on buttons, colors, layouts, and A/B tests. But people don’t convert because of a design change, but because a barrier to their decision disappeared.

Four drivers move people to act:

  1. Clarity: what you do, who it’s for, and why it matters.
  2. Relevance: different landing pages by audience and different messaging by funnel stage.
  3. Trust: reviews, case studies, and brand authority.
  4. Risk reduction: guarantees, free trials, flexible contracts, and transparent pricing.

Most visitors don’t buy because they’re convinced. They buy because uncertainty has been removed.

That gap is bigger than most brands realize. In one NP Digital study, only 18 percent of visitors agree they feel a sense of trust landing on a site, while 46 percent of companies believe visitors feel that trust. A visitor who doesn’t trust what they’re reading won’t convert, no matter how strong the offer is underneath it.

A bar chart showing whether or not website visitors get a sense of trust when landing on a site.

Source

The Four-Step CRO Framework

The best CRO programs rely on testing, and they run that testing through four repeatable phases.

  • Research. Pull analytics insights, heatmaps, session recordings, and customer feedback to see how visitors actually move through your site, not how you assume they move through it.
  • Prioritize. Rank opportunities by highest-traffic pages, largest funnel drop-offs, and effort versus impact. Fix the leak that costs you the most, not the one that’s easiest to notice.
  • Test. Run experiments on headlines, offers, CTAs, form length, layouts, and pricing presentation. Test one variable at a time so you know exactly what moved the number.
  • Scale. Deploy winning experiences across the rest of the funnel, not just the page where you tested them. A winning headline on one landing page is often a winning headline on five more.
Circular diagram of the four-step CRO framework: research, prioritize, test, scale; and donut chart showing organizational barriers to conversion rate optimization.

The organizational barriers that stall this framework are consistent across companies: resources (not enough time, budget, or headcount), followed by lack of stakeholder buy-in and lack of in-house testing expertise. Most CRO programs not necessarily due to a bad framework, but because the organization hasn’t agreed on who owns it or why it matters.

A test isn’t automatically valid just because it hit statistical significance. Before you trust a result enough to scale it, check five things: whether bot or AI-agent traffic skewed the numbers, whether engagement with the test matched the intended experience, whether users actually followed the path to conversion you expected, whether the variant increased abandonment anywhere else in the funnel, and whether the projected revenue lift justifies the time it took to build and launch. Skipping this check is how teams end up scaling a “winner” that isn’t actually one.

Measuring CRO Success Beyond Conversion Rate

A higher conversion rate doesn’t always mean a better business. Two of the biggest mistakes here are focusing only on conversion rate and optimizing for volume instead of value.

A test that lifts conversion rate by adding a discount code field, for instance, might also lower average order value or attract lower-quality leads that never close. The scorecard that actually protects revenue includes four metrics: revenue per visitor, customer acquisition cost, lead quality, and customer lifetime value.

The more of these metrics you track, the better you can tell whether a win in testing is a win for the business. Focus only on front-end conversions and you’ll optimize yourself into more leads that cost more to close and stick around for less time.

Four-part scorecard for measuring CRO success: revenue per visitor, customer acquisition cost, lead quality, and customer lifetime value.

The 60-Day CRO Growth Plan

Most companies need a structured plan to remove friction and improve conversion funnel optimization rather than a fancy website redesign. Our recommended 60-day plan runs in four phases.

Days 1 to 15 (Diagnose): Review landing pages, product pages, pricing pages, lead forms, and checkout experiences. Evaluate heatmaps, session recordings, scroll depth, and funnel abandonment data. Identify your highest- and lowest-converting channels.

Four-phase 60-day CRO growth plan: diagnose (days 1-15)

Days 16 to 30 (Prioritize and test): Focus on your highest-traffic pages, highest-revenue pages, and largest drop-off points. Test headlines, value propositions, CTAs, form length, and social proof. Align marketing, product, design, and analytics around a shared testing calendar.

: "Four-phase 60-day CRO growth plan: prioritize and test (days 16-30)

Days 31 to 45 (Scale): Apply winning elements to additional landing pages, product pages, lead-generation flows, and checkout experiences. Build variants by traffic source, funnel stage, customer intent, and visitor type.

: "Four-phase 60-day CRO growth plan: scale (days 31-45)

Days 46 to 60 (Operationalize): Establish a monthly experiment calendar, a testing backlog, and a prioritization framework so testing continues past the initial 60 days. Track revenue per visitor, conversion rate, lead quality, pipeline contribution, and CAC improvements.

Four-phase 60-day CRO growth plan: build a CRO operation system (days 45-60)

Teams that build this system early move faster in every quarter that follows, because they’re refining a working process instead of starting from scratch each time.

FAQs

What is CRO in marketing?

Conversion rate optimization (CRO) is the practice of increasing the percentage of website visitors who complete a desired action, such as making a purchase, submitting a lead form, or starting a trial. It works by identifying friction points in a funnel and testing changes that remove them.

Why isn’t more traffic fixing my conversion rate?

Traffic and conversion rate solve different problems. If your funnel has messaging, friction, trust, or experience issues, sending more visitors through it just means more people encountering the same drop-off points. Fixing the funnel itself typically produces more revenue than adding volume on top of it.

What causes the most conversion loss?

Conversion loss typically comes from four sources: unclear messaging, friction in forms or checkout flows, insufficient trust signals, and poor site experience. B2B teams tend to lose the most between the lead and marketing-qualified-lead stage, while B2C teams tend to lose the most between add-to-cart and checkout.

How often should I run CRO tests?

Companies running 11 to 15 tests per quarter saw the strongest revenue growth in NP Digital’s 2026 survey. The right cadence depends on your traffic volume, since you need enough visitors per variant to reach statistically significant results.

Conclusion

More traffic doesn’t fix a broken funnel,  it just sends more visitors through the same leak.

The organizations that win at growth treat CRO as an ongoing discipline.

They combine process, psychology, and measurement, and they build testing programs that compound instead of resetting every quarter. You don’t need twice the traffic, it’s better to focus on more value from the traffic you already have. If you want help finding where your funnel is leaking, talk to a CRO team that audits funnels for a living.

Read more at Read More

Which Fintech Brands Win in AI Search? We Analyzed 68,000 Answers to Find Out

/0 Comments/in /by

Key Findings

Note: Note: Our AI Visibility Score is a 100-point measure combining how often a brand appears in AI answers, how early it’s mentioned, and how many categories it shows up in. 

  • Apple ranks #1 for AI visibility with a score of 65.6/100, boosted by Apple Pay’s saturation in AI answers and established product presence.
  • AI platforms don’t share the same information sources. Reddit, for example, is mentioned in ChatGPT answers more often than of Perplexity
  • and Claude answers.
  • Among 7k brands studied, none cleared a score of 66. In contrast, by category, brands like Klarna own answers for BNPL (shows up in 92% of answers), suggesting more strategic focus on building category saturation versus broader presence.
  • SoFi leads pure fintech brands for AI visibility, scoring 64.6/100.
  • Raw volume of mentions alone doesn’t win, position at the top of the answer matters too: PayPal has the highest mention count in the dataset at 14,294 but ranks third overall because they’re being mentioned after other competitors.
  • Category concentration varies sharply. Klarna appears in 91.8 percent of buy now, pay later (BNPL) answers, while no lending or credit brand cleared 43.7 percent.
  • Third-party coverage is now an AI visibility lever. Comparison site and financial media coverage can influence which brands appear in AI-generated fintech answers.
  • By category, the top 3 brands for volume of answers and position in answers owned visibility, while brands scoring below the top 3 were close to invisible.

“What’s the best budget tracking app?” 

“How do I send money to friends and family?” 

These are the types of financial questions millions of people are bringing to the internet every day. Traditionally, you could find answers to these questions on blogs or fintech product homepages, but AI has become one of the primary channels for discovering fintech tools in recent years. Instead of Google, people are bringing their financial queries to channels like ChatGPT and Claude.  

Unfortunately, brands in the space have struggled to how often they show up on these platforms as consumers have started to rely on them more heavily for answers. 

Our study examines 68,334 answers to 250 prompts across five different AI platforms. Our queries also covered 10 categories of consumer fintech products
to deliver a clear picture of the space. 

Why AI Brand Presence in Fintech Matters Now

Fintech data platform Plaid reports that more than 200 million people ask ChatGPT personal finance questions every month. That’s only one platform. When you factor in other channels, such as Google AI Overviews and Claude, the volume of financial decisions shaped by AI answers grows very quickly.  

A LendingTree survey found that nearly half (49 percent) of AI chatbot users say AI has influenced at least one of their financial decisions, most often related to budgeting, taxes, or opening or closing an account or loan. 

Success in traditional search doesn’t automatically translate to AI visibility, however. You can dominate Google for your priority terms and still fail to appear when ChatGPT provides an answer to a similar query. 

The size of that gap is what stood out most. Across the 68,334 answers we analyzed, 7,317 distinct brands appeared. The highest AI Visibility Score any single brand achieved was 65.6 out of 100. Among pure fintech brands, no one cleared 65, granted, the study covers multiple fintech channels, so someone scoring 90 to 100 isn’t likely. 

While an extremely high score isn’t likely, the scores and dispersion of our fintech brands show us there’s an opportunity for newcomers to show up in a fragmented space.  

How We Measured Fintech AI Visibility

Here’s the breakdown of how we got our results: 

  • Platforms. We ran the same prompts across five AI systems: ChatGPT (GPT-5), Google AI Overviews, Gemini 2.5, Perplexity, and Claude Sonnet 4. We chose these because they were the highest-performing platforms in our previous AI platform study. Google AI Overviews is included because it appears in standard Google search results regardless of whether the user opts in. 
  • Prompts. We wrote 250 prompts (25 per fintech category) to mirror how real people ask AI about financial products. Instead of generic “best of” queries, we built each one around a purchase decision with context and product-fit specifics. We ran the prompts across all five platforms, and the analysis covered a total of 68,334 answers. 
  • Scoring. The AI Visibility Score is a 100-point scale built from three components: Frequency (50 points, the share of answers in which a brand appears), Position (30 points, how early the brand shows up in the answer text), and Coverage (20 points, how many of the 10 fintech categories the brand appears in). 

A few things worth calling out about the methodology: Our 10 categories were selected off industry relevance. We also inferred position from the first mention in the answer text, and we scored sentiment from the surrounding keyword context. Both are text-based proxies, not human-coded labels, which we’re flagging so readers can weigh the numbers with that context in mind. 

The Fintech AI Visibility Index: Overall Rankings

The top-scoring brand across our entire fintech-focused dataset is Apple, with an AI Visibility Score of 65.6 out of 100. Although Apple is a nuanced case with some financial products (Apple Pay, Apple Card, etc.), it’s worth noting that any fintech brand competing for AI visibility is also competing with tech-adjacent products it might not have considered rivals. 

Among pure fintech brands, SoFi leads at 64.6. It wins for a specific reason: SoFi is the only brand in our study to appear in all 10 fintech categories, including banking, lending, investing, and money transfers. Frequency (how consistently a brand is recommended within a specific topic) accounts for half the total score, but breadth (how often they’re mentioned across those 10 topics) accounts for another 20 percent, and no other fintech brand comes close to SoFi. 

Behind the top two, the leaderboard clusters tightly. PayPal (59) is third overall. The middle of the pack is a cluster of four brands within a 2-point range: 

  • Chase (57.8) 
  • Intuit (57.2) 
  • Fidelity (56.8)
  • Wise (55.4) 

Charles Schwab (53.6), Chime (53.0), and Revolut (52.6) complete the top 10. 

The top 10 brands in NP Digital’s Fintech AI Visibility study ranked by overall AI Visibility Index score.  

Fidelity and Wise both score higher than their raw mention volume would predict. Both brands average a first-mention rank of around three, meaning they appear early in the answer. Early mentions are what convert raw visibility into ranking, but the data shows that how often brands show up across categories matters as well. 

Let’s look at Wealthfront as an example. With an overall score of 46, the brand has a high mention count, but it covers only seven of the 10 fintech categories in our dataset. That coverage gap costs it enough points to drop it well below brands with lower raw mention rates but broader category presence. 

The top score on the leaderboard is a significant data point. No brand had an AI Visibility score above 66, and no pure fintech brand exceeded 65. The market leader is barely two-thirds of the way up a 100-point scale. That lack of fintech AI visibility spells opportunity. Fintech brands reading this piece are closer to the ceiling than they might realize. 

Category-by-Category: Who AI Recommends by Topic

The fintech brands in our study exist across the following categories: 

  1. Buy Now, Pay Later (BNPL) 
  1. Money Transfer Services 
  1. Investing Platforms 
  1. Robo-Advisors 
  1. Banking/Neobanks 
  1. Crypto Platforms 
  1. Budgeting Tools 
  1. Lending/Credit 
  1. Business Finance Tools 
  1. Payment Apps 

Each of the 10 fintech categories has its own AI visibility landscape, and the brand that leads the category is often not the brand that leads the overall index. Brand concentration also varies sharply within each category. 

Buy now, pay later (BNPL) is the most consolidated. The top three brands in that category account for nearly every mention. Lending and credit is the most fragmented, with no brand clearing a mention rate of 43.7 percent. 

The rest of the categories sit between those extremes, effectively owned by two or three brands with a long tail behind them. However, these top performers in every category aren’t defined by brand size or search volume. With multiple brands showing up in each answer, there’s potential for challenger brands to rival the top performers in certain categories. 

Here’s how they all break down. 

Buy Now, Pay Later (BNPL)

Klarna appears in 91.8 percent of BNPL answers, the highest single-category saturation rate in the entire dataset. Affirm (83.1 percent) and Afterpay (80 percent) make up the top three in this category. 

Bar chart ranking BNPL brands by AI mention rate, with Klarna leading at 91.8 percent, Affirm at 83.1 percent, and Afterpay at 80.0 percent. 

PayPal shows up in 64.7 percent of answers and Zip Co. in 54.1 percent, but as supporting mentions rather than contenders for the lead position. Klarna’s average first-mention rank of 2.89 means it appears first in the vast majority of answers in which it’s present. 

Money Transfer Services

Wise leads with 84.7 percent of mentions in the category and maintains that rate consistently across all five AI platforms. That kind of cross-platform consistency is rare throughout our data. Most brands over-index on one or two systems and drop off elsewhere. Wise doesn’t. 

Remitly (66.8 percent) and Western Union (55.2 percent) significantly trail Wise. Wise ranked number 1 in 57.7 percent of answers where it appeared, meaning it’s the first brand mentioned in more than half of the money transfer responses across the dataset. This is the cleanest example of a challenger brand in the study claiming category authority over a legacy brand like Western Union in AI answers. 

Bar chart ranking money transfer services by AI mention rate, with Wise leading at 84.7 percent, Remitly at 66.8 percent, and Western Union at 55.2 percent. 

Investing Platforms

Fidelity (74.5 percent) and Charles Schwab (73.8 percent) are close enough that AI treats them as co-leaders, contrary to how these two brands typically compete for market share elsewhere. AI doesn’t pick a winner between them. 

Robinhood holds a clear third at 56.3 percent, reflecting its brand recognition among newer investors. Webull (34.5 percent) and Interactive Brokers (33.8 percent) are nearly tied for fourth and fifth, representing a significant drop-off after the top three. 

Established platforms like Vanguard and E*Trade do appear in the data, but they’re below Interactive Brokers. This is despite the fact that Vanguard and E*Trade have massively larger brand monthly search volume, further proving that AI recommendations do not reflect brand size or consumer recognition. 

Smaller brands with strong feature-comparison content are competing directly with category giants in AI answers. Interactive Brokers features comparison content around commission-free trading, fractional shares, and platform tools tend to be surfaced by AI more than material purely based on brand size.  

Bar chart ranking investing platforms by AI mention rate, with Fidelity leading at 74.5 percent, Charles Schwab at 73.8 percent, and Robinhood at 56.3 percent. 

Robo-Advisors

Betterment (84.2 percent) and Wealthfront (75.7 percent) form the clearest brand pair in the study. When AI answers a robo-advisor question, both names usually appear in the same response. 

Betterment’s average first-mention rank of 2.56 is one of the lowest of any brand in the study, meaning it leads the answer more often than not. 

Fidelity (62.8 percent) and Charles Schwab (58.9 percent) hold supporting positions largely because they offer their own robo-advisor products. Functionally, this is a two-brand category with major institutional players filling the supporting slots. 

Bar chart ranking robo-advisor brands by AI mention rate, with Betterment leading at 84.2 percent, Wealthfront at 75.7 percent, and Fidelity at 62.8 percent. 

Banking / Neobanks

SoFi (59 percent) and Chime (54.2 percent) lead here, and this is where SoFi’s No. 2 overall Fintech AI Visibility Score finds its strongest support. Neither, however, holds the kind of dominant position seen in BNPL or robo-advisors. 

Capital One (40.3 percent) and Ally Bank (40.2 percent) are essentially tied for third, a notable result given how differently the two brands position themselves. Capital One positions itself as the neighborhood bank, opening up brick-and-mortar banking cafes, while Ally focuses on its strictly online presence as its unique selling proposition (USP).  

 These two notable names are followed by Varo (33.5 percent) and Current (28.7 percent), rounding out the top five and showing that AI recognizes challenger banks beyond the largest names. 

Bar chart ranking banking and neobank brands by AI mention rate, with SoFi leading at 59.0 percent, Chime at 54.2 percent, and Capital One at 40.3 percent. 

Crypto Platforms

Coinbase is one of the few brands in the dataset with a dominant first-place position at a 72.9 percent mention rate, with no legitimate competitor nearby.  

Kraken (64.9 percent) and Binance (52.5 percent) hold a clear second and third. Gemini, the crypto exchange (41.2 percent), and Crypto.com (38.8 percent) compete for the lower tier. 

Binance’s presence varies more across platforms than that of most brands in the study, (44.0% on Google AI Overviews to 61.9% on Claude, for example)  likely reflecting how each AI system handles regulatory and geopolitical sensitivities within this category. 

Bar chart ranking crypto platforms by AI mention rate, with Coinbase leading at 72.9 percent, Kraken at 64.9 percent, and Binance at 52.5 percent. 

Budgeting Tools

YNAB leads the budgeting category with a 72.8 percent mention rate and an average first-mention rank of 2.93. It’s almost always the first budgeting tool AI names. 

Monarch Money (55.5 percent) has emerged as a strong second, filling the gap left by Mint when Intuit shut it down in March 2024. PocketGuard and Goodbudget occupy the middle tier with 41.3 and 38.9 percent mention rates, respectively. 

Bar chart ranking budgeting tools by AI mention rate, with YNAB leading at 72.8 percent, Monarch Money at 55.5 percent, and PocketGuard at 41.3 percent.

 This category has recently been reshuffled with the loss of Mint, and AI has already settled on YNAB and Monarch as its new top two. 

Lending / Credit

SoFi leads this category with a 43.7 percent mention rate, but this is the most fragmented top five of any category in our list. Upstart (36.5 percent), LightStream (33.1 percent), LendingClub (26.9 percent, prior to its name change to Happen Bank), and EarnIn (26.7 percent) all sit within close range of each other. 

Bar chart ranking lending and credit brands by AI mention rate, with SoFi leading at 43.7 percent, Upstart at 36.5 percent, and LightStream at 33.1 percent. 

No brand has a commanding position, and the gap between first and fifth place is smaller here than anywhere else in the study. For a well-positioned challenger brand, this is the category with the most opportunity to gain AI visibility. 

Business Finance Tools

Intuit leads this category at 58.3 percent, carried by the combined strength of QuickBooks and TurboTax across multiple prompt types. Xero (41.2 percent) is the clearest runner-up and ranks higher in prompts with an international context. Wave, Expensify, and Zoho are all grouped around 21 percent, competing for the small-business and freelancer tier. 

Bar chart ranking business finance tools by AI mention rate, with Intuit leading at 58.3 percent, Xero at 41.2 percent, and Wave, Expensify, and Zoho clustered near 21 percent. 

This category also carries the highest positive sentiment rates. FreshBooks and QuickBooks both clear 75 percent positive framing, driven by AI’s tendency to recommend them in helpful, task-oriented contexts. 

Payment Apps

PayPal leads with a mention rate of 81.5 percent and is the most-mentioned brand in the entire dataset (14,294 total mentions). However, its average first-mention rank of 4.46 tells a more nuanced story. It appears frequently, more than any brand in the study, but not typically first, suggesting AI treats PayPal as a known option to include rather than a default recommendation. 

Venmo follows PayPal at 64.2 percent and frequently appears in the same answers. Cash App (48.1 percent) and Zelle (46.4 percent) sit close together in third and fourth, and Apple Pay rounds out the top five at 39.2 percent.  

Notably, Apple Pay ranks fifth in its category, despite Apple being the overall AI visibility leader across the study, a reminder that broad presence doesn’t translate to category dominance in every subcategory. 

Bar chart ranking payment app brands by AI mention rate, with PayPal leading at 81.5 percent, Venmo at 64.2 percent, and Cash App at 48.1 percent. 

How AI Describes Fintech Brands

In our study, fewer than 2 percent of fintech brand mentions are negatively framed. AI answers about financial products are overwhelmingly positive or neutral, but the lack of criticism isn’t what’s interesting here. The interesting finding is the split between brands AI actively recommends and brands AI merely lists. 

A bar graph showing the split of positive, negative, and neutral AI answer context among the top [10/15] brands. 

The positive sentiment leaders are primarily business finance tools. FreshBooks, QuickBooks, and Wave all exceed 75 percent positive framing. AI reaches for these brands in instructional, task-oriented queries, where the natural framing is recommendation language (“QuickBooks makes it easy to…”) rather than neutral cataloging. Task-oriented prompt intent produces a recommendation-shaped sentence structure, and a recommendation-shaped structure reads as positive framing. 

Two consumer-facing fintech brands stand out for a different reason. Wise (69 percent positive) and Capital One (68.6 percent positive) both consistently receive genuinely favorable descriptor language (“reliable,” “transparent,” and “well-established”). That framing isn’t from AI’s own judgment. It shows up because AI is pulling from third-party coverage, such as being included in listicles by outside financial publications, that describe these tools that way. 

On the other end, Cash App and Wells Fargo carry the highest relative negative framing among frequently mentioned brands, at 1.7 percent and 1.5 percent, respectively. Both are tied to fee and friction language that appears in complaint-adjacent contexts. 

The strategic read for a fintech marketer is that AI constructs sentiment from the language of the sources it pulls from and the prompt context it operates within. If your brand appears mostly in complaint-heavy contexts, that language shapes how AI describes you. If your brand appears in task-oriented, instructional coverage, the language of those contexts shapes the description too. 

In other words, sentiment reflects your source mix, not your reputation. 

How Brand Presence Shifts by Platform

Running the same prompts across five AI systems revealed three dynamics that measurably separate the platforms: how many brands they name per answer, which specific brands they favor, and where they pull their information from. 

A chart showing the average number of brands mentioned per answer across all five AI platforms in this study. 

ChatGPT returned the highest mention rates for PayPal (25.3 percent) and SoFi (20.5 percent), both exceeding their respective cross-platform averages. Claude Sonnet 4 returned the lowest mention rates for most major brands, consistent with a more selective response style. 

The most extreme single-brand variance in the entire dataset belongs to NerdWallet, one of the most popular financial comparison/review sites. The brand appeared in 11.5 percent of Google AI Overviews answers versus 1.3 percent of Gemini answers, a multiplier of nine for the same brand appearing on two different platforms. 

Answer Density by Platform

Gemini produced the most brand-dense answers in the study, averaging 9.8 brands per response. Perplexity and Claude sat at the other end, averaging just 6.  

A brand appearing in a Perplexity answer is competing against five others for reader attention; the same brand appearing in a Gemini answer is competing against nine. If a brand has to prioritize which platform to build fintech AI visibility on, the amount of brands per answer should factor into that decision. A platform that pulls information from fewer brands means less competition.  

Brand Mention Patterns Across Platforms

Third-party brand mentions vary meaningfully across AI platforms, and some patterns track with how those platforms operate. Reddit, for example, is named in 12.0 percent of ChatGPT answers and 7.8 percent of Google AI Overviews answers, but in fewer than 0.2 percent of Perplexity and Claude answers.  

ChatGPT and Google both have formal data-sharing agreements with Reddit, while Perplexity and Claude do not. That correlation is notable, though the dataset cannot confirm that the agreements caused the difference. 

Comparison sites show similar platform-specific patterns. NerdWallet is mentioned in 11.5 percent of Google AI Overviews answers but only 1.3 percent of Gemini answers, the largest single-brand cross-platform gap in the dataset. Bankrate also is mentioned most frequently in Google AI Overviews. 

For fintech brands, the takeaway is about where visibility is built. Brands featured consistently across comparison sites and relevant financial media or social discussions tend to appear more often in certain AI outputs. Strong brand-owned content paired with an active social strategy may also help newer brands punch above their weight, as SoFi’s visibility relative to more established financial institutions suggests. 

What This Means for Fintech Marketers

So, what does all this AI data mean for your fintech brand’s marketing strategy? Here are five practical takeaways from the study that marketers can take into account when approaching fintech AI visibility going forward: 

  • Frequency alone does not win the index. PayPal has the highest raw mention count in the entire dataset at 14,294, but ranks third overall because its average first-mention position (4.46) and category coverage (how many of the 10 categories a brand appeared in) cap its score. High-volume mentions without early positioning leave Visibility Index points on the table.  
  • Category authority beats broad presence for position scores. In the context of our study, category authority is how consistently and prominently a brand is recommended within a specific topic, whereas broad presence would be how many different fintech topics a brand appears across. Betterment and Klarna consistently lead their category answers with average first-mention ranks under three, despite lower total mention volume than PayPal. Owning your category in AI answers is worth more than being adjacent to many.  
  • Recommendation language matters more than positive sentiment. Fewer than 2 percent of mentions in the dataset were negative, which means positive framing is already the baseline. The big question is whether AI names your brand as a recommendation in task-oriented queries or lists it as a possible option among many.  
  • Each platform requires its own strategy. ChatGPT, Gemini, and Google AI Overviews weigh source types differently. ChatGPT favors Reddit mentions, while Google AI Overviews mentions NerdWallet and Bankrate at rates that other platforms don’t match. Gemini pulls from a broader mention mix than either. A single visibility strategy across all five platforms will underperform one calibrated to each platform.  
  • Third-party visibility can strengthen a brand’s presence in AI-generated answers. Coverage in financial media, placement on comparison sites, and relevant Reddit discussions may increase the likelihood that a brand will appear on certain AI platforms. As a result, earned media and generative engine optimization (GEO) should be planned together, even if each is measured against different outcomes. 

Conclusion

Fintech AI visibility is measurable, but it’s a metric that’s unevenly dispersed and still being defined.  

The ceiling of our study is an AI Visibility Score of 65.6, and no pure fintech brand cleared 65. There’s room for growth for the brands that take the right approach. 

The brands that will define fintech AI visibility over the next few years are the ones already measuring their standing today. They know which categories and platforms they’re winning or losing, and they’re acting on that intelligence before the market settles. 

NP Digital tracks these dynamics across dozens of enterprise-level brands, helping teams benchmark their AI visibility by platform and category and build strategies around what the data shows. If you’re trying to understand where your brand stands in AI answers, that’s the work we do every day.

Read more at Read More

WithinWP podcast: Humans, agents, and authenticity: rethinking search and sites in WordPress with Alex Moss

/0 Comments/in , /by

WithinWP podcast: Humans, agents, and authenticity: rethinking search and sites in WordPress with Alex Moss



Took place on

Hosts & Guests

Alex Moss

Principal SEO

Remkus de Vries

WithinWP Podcast Host

In this episode of Within WordPress, Remkus sat down with Alex Moss – Manchester-based SEO veteran, agency owner, and Principal SEO at Yoast – to explore how WordPress, SEO, and AI are colliding in real time. They cover idea-making and product risk, why optimism beats doom, and what tools like “OpenClaw” and Grok mean for everyday users versus developers.

You’ll hear practical multi-LLM workflows, how non-technical folks (including parents) are changing research habits, and where Yoast leans on ICPs to guide content. They look ahead to an agent-first web: data integrity, artifacts/markdown, and deeply personalized delivery – plus where WordPress fits.

Along the way: diagnosing an old Mercedes with an LLM, turning kids’ drawings into coloring pages, and speed-building ad‑free PWAs with Claude.

We wrap on authenticity, IRL experiences, jobs, and even AI, ASI, and faith – staying curious and optimistic.

First upcoming events

Introduction to Yoast SEO webinar
1 October 2026

Hosts

Mabel Adekola

Support engineer

Michael Quaranta

Support Team Lead

Introduction to Yoast SEO webinar

A practical, demo-driven webinar on using Yoast SEO for WordPress with confidence.

Register now (opens in a new tab)

Learn more

SEO for paws 2026
September 23, 2026

SEO for paws 2026

Team Yoast is Sponsoring at SEO for paws 2026! Click through to…



Learn more

The SEO Update by Yoast – September 2026
5 October 2026

Hosts

Alex Moss

Principal SEO

Carolyn Shelby

Principal SEO

The SEO Update by Yoast – September 2026

Expert analysis of the latest SEO & AI news developments with Carolyn…

Register now (opens in a new tab)

Learn more

Increase your site’s visibility

  • Get instant guidance on key phrases
  • Generate better titles and descriptions
  • Get clear SEO and readability checks
  • Trusted by 13M+ site owners

Get Yoast SEO (free) today

Reviews from real users

Rated 4.8 stars by 27,819 users

(opens in a new tab)

The post WithinWP podcast: Humans, agents, and authenticity: rethinking search and sites in WordPress with Alex Moss appeared first on Yoast.

Read more at Read More

GEO by Engine: How ChatGPT, Claude, Gemini, and Perplexity Really Decide What to Say

/0 Comments/in /by

Key Takeaways

  • All four major AI engines share a similar training recipe (pretraining, instruction tuning, and preference optimization), but each one is tuned toward a different outcome.
  • Public preference-training datasets show a documented shift from a single accept-or-reject judgment to a five-axis grading rubric, and that newer rubric appears to reward more structured, list-shaped answers.
  • Perplexity leans on live retrieval and citations. Claude’s edge is depth and long-form reasoning, while ChatGPT covers the broadest set of use cases and Gemini’s advantage comes from native access to Google’s own ecosystem.
  • Third-party benchmark testing changes every few months, so treat any single comparison as a snapshot, not a permanent ranking.
  • Measuring AI visibility means watching citation frequency, brand mentions, and source diversity, not just keyword rank.

You already know what GEO is. You’ve read the definitions, sat through the LLMO comparisons, and seen the “AI search is changing everything” takes. What’s still missing from most of that content is GEO by platform: the actual, mechanical differences in how ChatGPT, Claude, Gemini, and Perplexity decide what to surface. Clients are asking this question directly, and most of the guidance out there treats “optimize for AI search” as one strategy instead of four. 

Here’s the ground rule before we go further: nobody outside OpenAI, Anthropic, Google, and Perplexity knows the current, exact formula any of these engines uses to rank or select content. Those systems are proprietary, and they shift on a rolling basis. What we can do is walk through the documented mechanisms these systems share, where they genuinely diverge in practice, and what that means for your tactics and measurement, platform by platform. 

Why AI Engines Aren’t the Same

Every major AI engine on the market runs a version of the same recipe: pretraining on a massive amount of text, instruction tuning to make the model follow directions, preference optimization to shape its behavior, and product-level decisions about what to retrieve and rank when someone asks it something. That shared foundation is exactly why a lot of GEO optimization by platform advice sounds interchangeable. It’s also why that advice tends to underperform once you actually test it against a specific engine. 

Each engine bends that shared recipe toward a different job. Perplexity built its product around source-based retrieval, so its answers read closer to a research assistant than a chatbot: citation-heavy, and grounded in whatever’s currently live on the web. Claude was tuned to go deep, reasoning through multi-step problems, holding context across long documents, and working well inside agent-style workflows where accuracy compounds over many steps. ChatGPT covers the widest surface area of the group, with the largest install base and voice, vision, and multimodal features layered on top of a general-purpose core. Gemini’s biggest edge shows up the moment a task touches anything inside Google’s own ecosystem, whether that’s Docs, Sheets, YouTube, or Workspace data the other three simply can’t see. 

None of that comes down to taste. A single, generic “GEO best practices” checklist misses this entirely, because the differences between these systems come from how each one was built and what it’s optimized to reward, not from surface-level style choices. 

How GenAI Actually Decides What to Say

Decision-making inside these systems happens in three layers, and understanding them explains most of what looks like unpredictable behavior from the outside. 

Training-time shaping comes first. A model reads an enormous amount of text during pretraining, then goes through instruction tuning to learn to follow directions, then preference optimization, commonly RLHF or RLAIF, to learn which of two possible answers a human grader preferred. This is the stage where a model’s default habits get set: how much it hedges, how much detail it volunteers, how polite it sounds by default. 

Inference-time selection happens next, every time you send a prompt. The model scores and weights candidate responses, usually through a reward model or alignment layer trained on those same human preference judgments from the step above. 

Product-time retrieval is the layer that varies most visibly by platform. Some engines pull in outside sources at the moment you ask a question, a process called RAG, or retrieval-augmented generation, while others rely more on patterns already baked into the model’s weights through fine-tuning. This layer explains a lot of why some products feel more like a search engine and others feel more like a conversation. 

Two public datasets make the preference-tuning layer concrete instead of theoretical. Anthropic published hh-rlhf in 2022: 169,000 rows, each one a binary call on which of two responses a human grader preferred. Nvidia’s HelpSteer2, published in 2024, grades responses across five separate axes instead of one: helpfulness, correctness, coherence, complexity, and verbosity. 

A comparison between hh-rlgf and HelpSteer2,

 An independent analysis of both files found that under HelpSteer2’s five-axis rubric, more list-shaped, enumerated answers scored higher in a majority of the pairs examined. That’s a plausible partial explanation for why AI-generated answers so often default to bullets and numbered steps, but it’s worth treating as an informed inference rather than a settled fact. The researcher who ran the analysis was careful to frame it the same way. 

One caveat matters more than the data itself: neither file reflects how Anthropic or Nvidia trains models today. They’re historical snapshots from a specific year at a specific lab, not a live map of any current engine’s ranking logic. Treat them as a useful, citable window into how preference tuning has worked in at least these documented cases, not a spec sheet for what’s happening right now. 

What Changes by Engine

Perplexity: Built For Sourced Answers

Perplexity’s whole product is oriented around citations and fresh retrieval, and that shows up in the output. Independent testing that runs ChatGPT vs. Claude vs. Gemini vs. Perplexity through identical prompt batches consistently ranks Perplexity ahead on citation accuracy and real-time grounding, which tracks given it’s pulling from the live web at query time instead of relying mostly on training data. Where it falls short is anything creative or long-form. Ask it to draft a full article, and the output tends to read functional rather than polished. 

A sample Perplexity response for the question "what are some low impact ways I can get more exercise during a busy day?"

Claude: Built For Depth

Claude tends to show up strongest on tasks that require holding a lot of context and reasoning through it carefully. Testing rounds that track calibration, meaning how often a model’s confidence matches whether it’s actually right, have repeatedly put Claude ahead of the field, particularly on claims where being wrong actually matters. That combination of depth and caution is a big reason teams lean on Claude for long-form content and multi-step agent workflows rather than quick answers. 

A sample Claude answer for the question, "what are some low impact ways I can get more exercise during a busy day?"

ChatGPT: Built For Breadth

ChatGPT still carries the largest install base of the group and the widest feature set, voice, vision, image generation, browsing, and a long list of plugins layered onto a general-purpose core. That breadth is a real advantage. Several independent testers also note the output quality swings more than the other three without detailed prompting. It’s the most flexible tool here, and flexibility cuts both ways. 

A sample ChatGPT answer for the question: "what are some low impact ways I can get more exercise during a busy day?"

Gemini: Built For The Google Ecosystem

Gemini’s advantage rarely shows up in raw model quality alone. It shows up the moment a task touches Gmail, Docs, Sheets, or YouTube, where Gemini can actually read and act on your own data instead of talking about it in the abstract. For teams already living inside Google Workspace, that access matters more day to day than a benchmark score. 

Keep the bigger takeaway simple: none of these four wins across the board, and most credible testing in this space lands on some version of using more than one tool, matched to the task, rather than crowning a single winner. Revisit that assumption every few months. Standings shift, and last quarter’s leaderboard isn’t this quarter’s. 

A sample response in Gemini for the question: "what are some low impact ways I can get more exercise during a busy day?"

What Tactics Change by Engine

Here’s where this gets practical. Once you understand the mechanism and positioning differences above, the tactical shifts stop feeling arbitrary. They follow directly from what each engine actually rewards. 

Perplexity: Lead With Sources

Because Perplexity leans this hard on retrieval, prioritize factual, source-rich content that matches a query directly. Original research, cited statistics, and clearly attributable claims perform better here than persuasive copy. Structure your content so a system pulling live answers can lift a clean, self-contained statement out of it without needing the surrounding context.  

The Marketing Stats page on NeilPatel.com

Claude: Lead With Structure And POV

Claude rewards depth over surface-level breadth, so long-form content with strong headings, a clear argument, and an experiential point of view earns more traction here than a shallow listicle covering the same ground. If you have real experience running the strategy you’re writing about, say so directly. That’s the kind of signal this engine’s reasoning tends to weight. [Internal link suggestion: long-form content / E-E-A-T guide] 

An author page for Nikki Lam on SearchEngineLand

ChatGPT: Lead With Flexibility

ChatGPT serves answers across the widest range of surfaces, so content that works as both a full explainer and a set of shorter, atomized pieces tends to travel further here. Conversational framing helps too, since a large share of ChatGPT’s traffic comes through follow-up questions rather than one query. 

An example of key takeaways on an Entrust blog.

Gemini: Lead With Entities And Structure

Gemini’s advantage comes from the Google ecosystem, so optimize for visibility inside it specifically: clean entity definitions, structured data, and schema markup that helps Google’s own systems understand what your content is actually about. This is less about persuasive writing and more about making your content legible to a system that already has your brand’s data sitting in other Google products.  

An example of schema markup in action.

One more distinction is worth real space here, because it’s one of the most practical takeaways in this piece: these engines don’t weight sources the same way. Some lean more heavily on a brand’s own site content. Others lean more on third-party mentions and citations from outlets a brand doesn’t control. Several reward structured, schema-marked data over plain prose, regardless of who published it. Knowing which lever matters most for a given engine changes where you spend your time: publishing more on your own domain, earning more third-party citations, or investing in structured data that makes your existing content easier to parse. 

KPIs for Tracking Performance Across Engines

Traditional rank tracking doesn’t map cleanly onto AI answers, so you need a different measurement set once you’re optimizing for GEO by platform instead of just SEO. 

Start with citation frequency: how often your content actually gets cited as a source across these engines, not just mentioned. Track brand mentions separately, since showing up by name in an AI answer still carries value even without a citation attached. Watch whether your content lands inside the direct answer summary or gets pushed to a secondary link a user has to click through to find, because that placement difference matters more here than it ever did in traditional search. 

AI Citations in the Writesonic platform.

Source diversity is worth tracking at the query level: how many distinct domains a given engine pulls from for a topic you care about, and whether your brand is consistently one of them. Query match quality matters more than keyword match here. Look at how closely your content actually maps to the real prompts people are typing, not just the keywords you targeted. 

Content freshness rounds this out, especially for retrieval-heavy engines like Perplexity, where how recently you updated a page can affect whether it gets pulled into an answer at all. None of these metrics replace the KPIs you already track. They sit alongside them, and together they give you a fuller picture of whether your content is showing up where your buyers are actually asking questions.

FAQs

 

Do ChatGPT, Claude, Gemini, and Perplexity Use the Same Ranking Algorithm? 

No. All four share a similar training foundation, pretraining, instruction tuning, and preference optimization, but each product layers different retrieval and ranking choices on top of it. That’s why the same query can produce different answers across platforms. 

What Is RLHF, And Why Does It Matter For GEO? 

RLHF stands for reinforcement learning from human feedback: human graders rank possible responses, and the model gets tuned to favor the ones graders preferred. It matters for GEO because it shapes what a model considers a “good” answer, including, in at least some documented cases, a preference for more structured, list-style content. 

Is GEO Different for Every AI Platform? 

Yes, meaningfully so. Each engine optimizes for a different outcome and pulls from different sources, so a single, generic GEO checklist will underperform a platform-specific approach. 

Conclusion

The mechanics genuinely differ by engine, and that’s not changing as these products keep evolving. But the goal was never to master four black boxes. It’s to understand the shared foundations well enough to make an informed, platform-specific bet, then revisit that bet as the landscape shifts, because it will. 

That’s the kind of work we do at NP Digital: tracking how these engines behave in practice, testing content against real prompts, and adjusting strategy as each platform updates. Start with one platform where your buyers already spend time, apply what’s above, and measure it. Then expand from there. 

Read more at Read More

The follow-up query: Rethinking SEO for conversational search

/0 Comments/in , /by
The follow-up query- Rethinking SEO for conversational search

Today, the most revealing search query may be the follow-up. The question people ask after they learn, compare, or reconsider often reveals what they actually need.

For brands, that next question represents a valuable opportunity to earn visibility, secure trust, and support a meaningful action. The objective isn’t to predict every possible prompt. It’s to understand the audience well enough that your content, evidence, and experience remain useful as the conversation evolves.

The first query doesn’t tell the whole story

Traditional SEO strategy has focused primarily on the query that brings someone to a result. But what happens after the first answer is delivered?

A user can now begin with a broad question, narrow the request, add a photo, ask for a comparison, and move toward an action without restarting the search process. The follow-up query is the connective tissue in that journey.

This dynamic more closely mirrors how people actually think. We rarely begin with a perfectly formed question. We learn, reconsider what matters, add context, and proceed down the rabbit hole.

The biggest change isn’t that people suddenly began asking related questions. It’s that the interface can now remember the relationship between them.

Traditional SEO best practices still matter. The strategic workflow has simply expanded. Instead of treating a single query and landing page as the whole assignment, brands must consider the first question, likely follow-ups, credible proof, and the next useful action.

The goal is simple: Help people answer the question they have now and make it easy to answer the question they’re likely to have next.

Be the brand AI recommends.

See where your brand appears in AI search, where competitors are winning, and what it takes to become the answer AI recommends.

See your AI visibility

What conversational search actually means

Conversational search allows people to ask natural questions, carry context from one request to the next, and refine their needs along the way. It can happen in a search engine, AI assistant, voice interface, on-site chatbot, shopping assistant, or visual search tool.

Three qualities distinguish conversational search from conventional search:

  • Context carries forward: “What about one under $200?” makes sense because the system remembers what “one” refers to.
  • Intent can evolve: A user may move from learning to comparing to buying within one interaction.
  • The format can change: A journey may combine typed text, speech, images, video, maps, charts, or product feeds.

Not every voice query or AI summary is conversational. The defining question is: Can the person continue the task without rebuilding the context?

Conversational search also overlaps with personalization, but the two aren’t identical. Personalization changes an answer based on what a system knows about the individual. 

Conversational search changes an answer based on what the individual reveals during the exchange. Increasingly, search systems combine both forms of context to deliver more individualized responses.

Here’s a simple example:

  • Initial query: “What is the best carry-on for a five-day work trip?”
  • Follow-up: “I need a laptop sleeve and I fly budget airlines.”
  • Visual turn: The user uploads a photo of a bag and asks, “Is this likely to fit?”
  • Decision turn: “Compare two options that are under $250.”
  • Action turn: “Which choice can arrive by Friday?”

Traditional keyword research might stop at “best carry-on luggage.” Conversational strategy follows the full decision arc. Specifications, comparisons, images, policies, inventory, delivery data, and expert guidance all have a role to play.

A sample search can move from question to image, and follow-up to action.
A sample search can move from question to image, and follow-up to action.

How search became conversational

Search has gradually evolved from keyword-driven interactions to contextual, multimodal conversations. 
Search has gradually evolved from keyword-driven interactions to contextual, multimodal conversations. 

Conversational search didn’t begin with ChatGPT or other AI platforms. Its building blocks developed over decades alongside changing user behavior.

Keywords and reformulation

Early web search encouraged short noun phrases that were often stripped of natural grammar. If the results missed the central need, users manually reformulated the search: “running shoes,” then “running shoes flat feet,” then “best stability running shoes women.”

The user carried the context between searches. SEO centered on keyword matching and landing pages built around primary phrases.

Semantic and contextual understanding

Search engines gradually improved at understanding entities, relationships, intent, and natural phrasing. Google’s 2019 BERT announcement emphasized the context and relationships among words in a query. Small terms such as “to,” “for,” and “no” could materially change intent.

For SEO, that shift reduced the value of repetitive exact-match language and increased the importance of satisfying underlying needs.

Voice and answer-first interfaces

Voice assistants normalized complete questions and concise spoken answers. They also introduced local, immediate, and hands-free situations. Related searches could have been “Where is the nearest pharmacy open now?” or “How long do I bake salmon at 400 degrees?”

Many voice interactions remained single-turn, but the lasting lesson was clear: Provide concise answers, accurate facts, and content that works when heard rather than read on a page.

Complex and multimodal understanding

Google’s 2021 MUM announcement framed complex tasks as journeys that could require multiple searches. In 2022, Lens multisearch enabled people to combine an image with text such as a color, attribute, or question.

The direction was already clear: Let people authentically express their needs while the system performs more of the work behind the scenes.

What conversational search looks like now

Generative AI can interpret natural language, retrieve up-to-date information, combine sources, and retain context within a single interface. That dynamic changes both how people express their needs and how platforms search on their behalf.

One question can trigger many searches

Google says AI Overviews and AI Mode may use query fan-out, running multiple related searches across subtopics and data sources. For example, a lawn care question may prompt research into treatment, prevention, safety, cost, climate, and timing.

The visible prompt doesn’t tell the whole story. A page can support part of an answer even when it doesn’t mirror the wording of the initial question. Keyword datasets still reveal demand, but they represent only part of the picture. Support questions, on-site searches, reviews, sales conversations, and prompt testing can expose the needs that come next.

Dig deeper: Query fan-out optimization guide: How to rank in AI searches

Follow-ups turn results into journeys

Google has made the shift visible by connecting follow-up questions in AI Overviews to a continuing conversation in AI Mode. ChatGPT search similarly blends conversational responses with timely web information and source links.

On a practical level, people can reveal more information with every turn:

  • “Explain heat pumps.”
  • “Would one work in a 1920s house?”
  • “What if the electrical panel is only 100 amps?”
  • “Estimate the trade-offs in Southern California.”
  • “What should I ask contractors?”

Each question changes the best answer. A page that handles only the definition may support the opening request and disappear from the rest of the journey. The initial query identifies the topic. The follow-ups reveal what truly matters: budget, risk, location, use case, or deadline.

This shift is already affecting the search interface. In January, Barry Schwartz documented how follow-up questions can move users directly from AI Overviews into AI Mode, a feature that Google later pushed worldwide across desktop and mobile.

Multimodal inputs make the conversation more natural

People don’t need to translate everything they see into keywords. They can show a system an object, screen, plant, product, room, or broken part and ask a direct question.

In May 2025, Google reported that Lens handles more than 25 billion queries per month. Search Live has further expanded this behavior by allowing people to discuss a live camera view and ask free-flowing follow-ups.

For brands, visual SEO can’t stop at filenames and alt text. The image or video must be useful. Clear angles, close-ups, scale, labels, captions, demonstrations, and transcripts can help a person and a search system understand what the visual is trying to prove.

A photograph showing exactly where a reset button sits on an appliance is more useful than a polished lifestyle image of the same appliance. 

Dig deeper: How multimodal discovery is redefining SEO in the AI era

Search is moving closer to action

Conversational systems increasingly connect research with execution. Search experiences can already assist with tickets, reservations, local appointments, shopping, and forms. 

As agentic capabilities develop, accurate availability, pricing, policies, product details, and accessible conversion paths become an increasingly important part of discoverability.

A great article can’t rescue inaccurate inventory or a broken booking flow.

More needs are being satisfied before a click

Visits containing a Google AI summary resulted in a traditional result click 8% of the time, compared with 15% when no AI summary appeared, per a Pew Research Center study. Links inside the summaries received clicks in only 1% of visits.

One study can’t provide a universal CTR forecast, but the direction matters. A brand can influence a decision without receiving a visit. The clicks that remain may also represent a later, more qualified need.

When an AI response covers the basics, people need a stronger reason to click. They may want to verify a claim, see the original demonstration, use a tool, join a community, or check current availability. The click increasingly means: “Give me the proof, experience, or utility the summary can’t.”

That dynamic elevates:

  • Original reporting, testing, research, and firsthand experience.
  • Transparent authorship, methodology, dates, sources, and corrections.
  • Calculators, datasets, templates, maps, and interactive tools.
  • Newsletters, saved lists, communities, and other reasons to return directly.
  • Clear next actions that respect the person’s stage and risk level.

If an AI answer can repeat everything on the page, the page needs to offer something more.

Trust becomes part of the conversion

Increased automation can make human judgment feel more valuable. People don’t require a person to handle every interaction, but they do want to feel understood rather than processed — especially when nuance, emotion, or risk enters the conversation.

Good self-service respects people’s time. It should answer routine questions clearly, admit uncertainty, and provide a direct path to a knowledgeable human when empathy or accountability matters.

Search isn’t only an acquisition channel delivering anonymous traffic. It can be the beginning of a relationship that continues through a useful tool, newsletter, expert response, or community. That relationship can begin before the click and deepen after it.

This principle has guided my work in news SEO. The first search often identifies the event, while the next wave of questions reveals what readers actually need. For instance:

  • A coaching announcement may begin with a name and team, then expand into contract terms, career history, replacement candidates, and what the move means for the upcoming season. 
  • A wildfire search sequence may begin with the fire’s name and location, then expand into evacuation zones, road closures, shelter information, containment levels, air quality, and the neighborhoods at greatest risk. 

The opportunity isn’t to publish a thin article for every variation. It’s to build a connected resource that can serve the evolving story.

Get the newsletter search marketers rely on.

See terms.

The strategic shift: Optimize the conversation, not just the keyword

This change can feel bigger than it needs to be. We must always remember: Simple isn’t stupid when it comes to audience engagement.

We don’t need to predict every prompt or rebuild an entire content program overnight. We need to listen more closely, connect related needs, and make our best information easier to find, understand, trust, and use.

Start with the audience journey, not the newest AI feature. Platforms will change. The need to be useful, credible, and genuinely responsive won’t.

Map follow-up paths around real decisions

Choose a high-value audience need and brainstorm what a person would naturally ask next. Map at least five types of follow-up:

  • Clarify: “What does that mean?”
  • Constrain: “What if I have a small budget?”
  • Compare: “How is option A different from option B?”
  • Validate: “What evidence supports that?”
  • Act: “What should I do, buy, book, or ask next?”

Use search data, support tickets, sales calls, reviews, community discussions, and prompt testing to your advantage, but keep the order straight. AI can suggest questions. Human behavior should tell us whether those questions matter.

Conversation maps should also reflect how different audiences express the same need. Test important journeys with native-language experts, accounting for regional phrasing, cultural context, and moments when people switch languages during an exchange. Translation alone may not reveal the same follow-up intent.

Not sure where to begin? Take five real user questions and write the most likely follow-up under each one. With that framework, you already have the foundation of a conversation map.

Five follow-up paths can reveal the full decision journey.
Five follow-up paths can reveal the full decision journey.

Build topic systems rather than prompt pages

Create a strong hub around the main decision and connect it to a small number of useful supporting resources. These might include:

  • A definitive overview.
  • A comparison or alternatives page.
  • A decision checklist.
  • A firsthand case study.
  • A visual demonstration.
  • A tool or calculator.
  • Product, location, policy, or service pages with current facts.

Don’t publish 30 weak articles that say nearly the same thing. Create one excellent explanation and support it with assets that serve distinct needs.

Internal links should follow the user’s questions, not just mirror the site’s organizational chart. On the anchor text front, “compare plans,” “check compatibility,” and “understand the risks” provide more direction than “learn more.”

Decide whether each follow-up requires a section, separate page, reusable component, or no new asset at all. Let the task (not a minor keyword variation) drive that decision. 

Dig deeper: How to build topic clusters that AI will cite

Make every key claim easy to verify

State the answer first, then provide the evidence behind it. When the answer depends on specific circumstances, explain those conditions and include the relevant source, date, test method, or limitation.

  • Weak example: “This laptop has all-day battery life.”
  • Stronger example: “In our battery test, this laptop lasted 14 hours while streaming video over Wi-Fi at 50% screen brightness. Battery life may be shorter when gaming or running power-intensive applications.”

The stronger version gives a search system a clear, well-supported answer to assess and potentially cite. More importantly, it tells a person what the claim means and when it may not apply.

Before publishing, always ask: “What evidence would a skeptical reader need to believe this?”

Design for retrieval and reading

Keep important information in crawlable text even when a video, graphic, or app provides the richer experience. Use descriptive headings, concise definitions, and logical sections to help users extract information more easily.

Maintain the fundamentals:

  • Allow intended crawling and indexing.
  • Use canonical URLs and descriptive internal links.
  • Keep important information out of images alone.
  • Match structured data to visible content.
  • Maintain Merchant Center, product feeds, and Business Profile data where relevant.
  • Provide a fast and accessible mobile experience.

There’s no secret AI markup. Google says AI Overviews and AI Mode don’t require special optimization. Search systems can’t confidently use what they can’t access or understand.

Test whether people can continue the journey across voice, keyboard, screen-reader, and visual interfaces. Clear headings, descriptive alt text, captions, transcripts, accessible forms, and understandable error messages do more than satisfy technical requirements. They determine who can participate in the conversation.

Treat images and video as answer assets

For every important visual, ask: “What can this answer that text can’t?”

A product image can show scale and fit. A repair video can demonstrate motion and sequence. A safety graphic can clarify a warning sign. Captions, transcripts, and expert reviews can provide context the visual can’t carry alone.

Avoid decorative imagery that adds no information. Use descriptive filenames and alt text for accessibility without keyword stuffing. Place visuals near the relevant explanation and provide stable pages where they can be discovered.

Could someone learn, compare, or complete a step from this visual? If not, reconsider whether it’s necessary.

Dig deeper: Your images have a new job in AI search

Create a strong next turn on owned surfaces

When someone arrives from an AI answer, assume they may already know the basics. Provide the next layer immediately:

  • A comparison after an overview.
  • A calculator after an explanation.
  • A “what changes for your situation?” breakdown.
  • Primary documents and methodology after a summary.
  • Availability and booking steps after local advice.

On-site search and chat should preserve useful context, cite source material, offer escalation routes, and admit uncertainty. Don’t send a well-informed visitor back to the beginning of the discovery process.

Preserve only the context needed to reduce repetition, not every detail a person shares. Sensitive information requires clear consent, limited retention, and an obvious path to human help.

Build authority people can remember

AI systems often synthesize multiple sources. Generic prose is easy to replace, but distinctive evidence and recognizable expertise aren’t.

Build assets people can associate with your brand: a named dataset, annual benchmark, expert rubric, original test, or trusted decision tool. Newsletters, alerts, memberships, proprietary tools, events, and communities can then reduce dependence on any single interface.

Search value increasingly incorporates visibility, trust, and direct audience relationships — not only the immediate click.

Dig deeper: Utility news content: How to win beyond clicks in AI search

Conversational search is a cross-functional responsibility

SEO and content teams can’t deliver the full journey alone. Conversational search touches:

  • Content and subject-matter experts, who provide accurate explanations and proof.
  • UX and product teams, who build usable paths and context-aware experiences.
  • Customer service and sales, who address real questions, objections, and anxieties.
  • Ecommerce and operations, who maintain prices, inventory, policies, and availability.
  • Legal, privacy, and accessibility partners, who establish responsible boundaries.
  • Analytics teams, who connect discovery with engagement and outcomes.

The strongest strategy synthesizes these functions around the audience’s decision-making process rather than forcing each individual to navigate the organization’s internal silos.

Follow this practical workflow for teams:

  • Choose one important journey: Define the audience, decision, risk, and desired outcome. Combine keyword data with support transcripts, sales objections, on-site searches, reviews, and expert interviews.
  • Map the conversation: Branch the initial need into clarification, constraints, comparison, validation, and action. Mark what already exists, pinpoint weaknesses, and establish what requires a new asset or data source.
  • Match the answer to the format: Use prose for explanation, tables for comparison, images for recognition, video for motion, tools for calculation, and feeds for changing facts. Add firsthand evidence, expert review, dates, and limitations.
  • Connect the journey: Align the hub, supporting pages, visuals, operational data, and conversion experience. Decide when self-service should become a context-aware human exchange.
  • Test and improve: Sample initial prompts, follow-ups, and image-based inputs across relevant interfaces. Track accuracy, citations, gaps, and competitors. Improve weak branches instead of aimlessly publishing new content.

Measurement: A scorecard for conversational search

No single metric tells the whole story of conversational search. Begin with four questions:

  • Can people find you? Track topic-level visibility, multimodal discovery, branded demand, and accurate presence within a fixed sample of AI answers.
  • Do the right people engage? Monitor tool use, video completion, evidence-module clicks, return visits, saved items, and newsletter sign-ups alongside pageviews.
  • Do they trust you? Monitor citation accuracy, corrections, earned references, direct traffic, freshness, and successful human interactions.
  • Do they act? Measure assisted conversions, qualified leads, bookings, purchases, support resolution, time to decision, and retention.

Treat AI citation testing as directional, not as definitive market share. Outputs vary by interface, conversation history, time, location, account context, and model. 

Document your definitions and maintain a consistent prompt sample before comparing periods. 

Measure the journey from discovery through action
Measure the journey from discovery through action.

Dig deeper: How to measure prompt-level visibility in AI search

What comes next

The next evolution is likely to move from conversational answers toward conversational action. A person may have a broken appliance, troubleshoot it, compare replacements, check local stock, and book installation within one search session.

As AI agents compare, filter, schedule, and transact under user direction, accurate prices, clear policies, consistent product information, accessible workflows, and source transparency will become even more important.

Original evidence, named expertise, clear methods, and current source pages will help people verify what they receive. Don’t chase every interface change or create thin content for every prompt variation. The goal is to add value, not volume.

Discovery will remain distributed across search engines, AI assistants, social video, communities, commerce platforms, and brand-owned experiences. The goal isn’t to put all of your search strategy eggs into one chatbot’s basket. It’s to become the clearest, most useful, and verifiable source for the decisions your audience needs to make.

If AI can’t find you, customers won’t either.

Track your visibility across AI search, uncover missed opportunities, and grow your presence where customers are asking questions.

See your AI visibility

Make the next question count

The follow-up query changes SEO because it exposes the full shape of intent. People don’t experience their needs as isolated keywords. They learn, revise, compare, validate, and act.

We don’t need to predict every sentence someone might type. We need to understand the decision journey well enough that our content, evidence, visuals, data, and experience remain useful as the conversation changes. That means creating useful content rather than commodity AI-generated drafts, keeping important information accessible, and never fabricating firsthand experience, reviews, or citations.

The brands that stand out won’t simply answer the first question. They’ll listen, adapt, and earn the next one — proving that even as search becomes more automated, lasting authority is still built through useful information, credible evidence, and genuine human connection.

Read more at Read More

Claude for SEO: 5 ways to automate repetitive work

/0 Comments/in , /by
Claude for SEO- 5 ways to automate repetitive work

After working on my hreflang XML sitemap project with Gemini, my company decided to invest in paid Claude accounts. I took the opportunity to see what else Claude could do through its desktop app and found some great uses for it.

Here are five ways I’ve been using it in my everyday SEO work.

1. Implement a daily briefing

The first thing I set up was a daily intelligence brief. Using lessons learned from other AI tools, I crafted a detailed outline of the types of news I wanted to see, including industry updates, competitor news, mergers, and acquisitions.

The key to getting more useful information and less fluff is to be specific and use the output as a guide for improvement. Don’t accept schlock — you’re paying for the tool, so teach it to do what you need.

For instance, I have Gemini checking atmospheric conditions for radio wave propagation in my amateur radio hobby, and it took several adjustments to get useful output.

Dig deeper: How to turn Claude Code into your SEO command center

Be the brand AI recommends.

See where your brand appears in AI search, where competitors are winning, and what it takes to become the answer AI recommends.

See your AI visibility

2. Connect to data sources for faster analysis

Yes, I can dig into Google Analytics and analyze the data, but how much faster would it be if I could just ask Claude to look at the data and answer questions for me? It turns out it’s much faster.

The folks at Rank Math shared instructions on how to make a direct connection without using third-party middleware. This tip has saved me a lot of time by allowing me to ask for what I need and quickly get an answer.

3. Create a better and faster hreflang XML sitemap

I decided to see how Claude would tackle the hreflang XML sitemap project. Like I did in my Gemini example, I told it what I wanted, fed it the latest versions of the sitemaps, and asked how it would approach the problem.

Claude surprised me by asking only for links to XML sitemaps as input. Using its own resources, Claude gathered all the pages and built the sitemap for me without any further intervention.

I double-checked the output and found it to be excellent on the first go, avoiding the need for the back and forth I experienced with Gemini — a huge time saver.

Get the newsletter search marketers rely on.

See terms.

4. Generate, translate, and localize content

Our team supports three main businesses across several regions. This requires content updates to be replicated across websites, translated, and localized as appropriate. We need to track this work as it goes through the pipeline so nothing falls through the cracks.

This is where Claude has been a tremendous benefit. I took an example of content that had been generated, vetted, and approved by one of our in-house experts and decided to see if Claude could handle automating some of the international SEO process. The results were far beyond my expectations.

As usual, beginning with the end in mind, I uploaded examples of previous translation and localization work to show what I needed done. Next, I uploaded a new item and gave it specific instructions:

  • Look at the content and identify the page on the main .com website to understand the service and its context.
  • Check the regional websites and determine if that same service is offered.
  • If it is, create a translated/regionalized version using our standard template.
  • Create tasks in our workflow tracker (we use Asana) and space out due dates according to a standard formula.
  • Upload the content docs to the created task.

I was pleasantly surprised by how quickly Claude was able to do this. It accomplished everything I needed except uploading the document to the ticket. The translations were on par with what we were getting with Google Translate, as verified by our in-country experts.

Strike another blow against inefficiency.

Dig deeper: Turn your SEO process into AI-powered tools

5. Replace an outdated WordPress plugin

While setting up a new WordPress website for one of our businesses, I found that one of the plugins we were using hadn’t been updated in several years and had known security issues. Thankfully, other security measures in place significantly mitigated the risk, but it’s always better to fix the problem than rely on other means.

I found a suitable substitute and thought about how long it would take my team and me to go through the 20+ websites we support and update the plugins. Even if it only took an hour per website, that’s more than 20 hours of dedicated, hands-on, fingers-to-keys work.

Before starting the process, I asked Claude if it could do it for me. Again, I was pleasantly surprised.

Using Claude’s Chrome extension, I logged into a staging environment for one of our smaller websites and told Claude what to do. It found all the places the plugin was being used, swapped it out with the new plugin, and did a quick visual check. There was one small discrepancy with the first attempt, which was quickly fixed when I pointed out the problem to Claude.

“Would you like me to check the other plugins for any security issues?” it asked when completed. Of course I would. It found an additional plugin that was unused and no longer supported. That was quickly deactivated and deleted.

My only time requirement was about five minutes per site of hands-on work as I:

  • Pulled a copy of each production website from staging.
  • Turned Claude loose.
  • Visually checked URLs that were updated.
  • Pushed the newly updated staging version to production.

Everything went on in the background as I worked on other tasks, and I didn’t have to get anyone else from my team involved.

I bet you’d be willing to trade getting the same result from five minutes of work vs. an hour any day of the week.

If AI can’t find you, customers won’t either.

Track your visibility across AI search, uncover missed opportunities, and grow your presence where customers are asking questions.

See your AI visibility

How to work effectively with AI

AI tools can feel overwhelming, but they’re an incredibly useful asset when used properly. Follow these practical guidelines when implementing AI, and you’ll be blazing through those tedious SEO tasks in no time.

Be the strategist and let AI be the doer

Don’t just demand a final product. Discuss the architecture, the edge cases, and the logic first, and be specific. Treat the AI like an intern by showing examples of final products when possible.

Use the proper tool for the proper job

Not all AIs are alike, and some are better at some tasks than others. Read, learn, and experiment. There are a lot of us out there using AI tools, and people have found some amazing results using one platform that completely failed on another.

Check, double-check, and correct when needed

Much like embracing the iterative process I referred to in my last article about using AI tools, don’t be afraid to check thoroughly and correct when needed.

The reality that AI is really still in its infancy becomes readily apparent when it fails. You may have to spoon-feed information at first to get the result you want. Remember, you don’t have to be polite — be direct and clear for faster results.

Read more at Read More

AI Overviews and paid search: When to fight, influence, or generate demand

/0 Comments/in , /by
AI Overviews and paid search- A three-bucket framework

AI Overviews reshaped the search results page, yet most paid search playbooks still run as if nothing happened.

You still hear the same default rule: bid up where performance looks strong, bid down where it looks weak. That assumes every query is the same kind of opportunity, a small auction for the same kind of click. In B2B and industrial accounts, where sales cycles run for months, conversion data is thin, and much of the demand is informational, that assumption falls apart.

When Google answers a question directly on the results page, the value of a paid click varies across your keywords.

  • Some queries still lead to a real commercial decision, and winning the top spot above the answer is worth paying for.
  • Others get resolved inside the AI answer itself, so the goal shifts from buying a click to influencing the answer and building a brand strong enough to be included.
  • A third group has moved so far toward zero-click that paid search on its own can no longer rebuild the journey.

Those are three different problems, and a single bid lever can’t solve all three. Trying to is how accounts spend budget on the wrong surface. I separate them into three buckets: fight, influence, and generate demand.

The framework: Fight, influence, and generate demand

Each bucket describes a different mechanism on the results page, so each one calls for a different approach.

Bucket 1: Fight

Fight is for bottom-of-funnel queries where intent is clearly commercial, and a click still turns into direct leads and sales: product + modifier, supplier shortlists, “buy,” “quote,” “distributor,” and brand terms combined with buying intent. Here, the job hasn’t changed much. You want to win the auction that sits above the answer, and you want the ad itself to answer the buying question.

This is where paid competition concentrates, and the data show it. The likelihood of an ad appearing rose steadily with cost per click, according to a July SE Ranking study of commercial queries in Google’s U.S. AI Mode:

  • About 54% at $10 or more.
  • Around 24% for keywords under $2.
  • 32% between $2 and $10.

AI Mode isn’t the same surface as AI Overviews, but what they share is the pattern: paid inventory follows commercial weight. Where the market already pays more for a click, AI-era ad surfaces appear more often. You rarely compete alone, since about seven in 10 answers with an ad in that study showed two advertisers in the block.

Fight is expensive for a reason, so treat it that way.

  • Reserve the premium for terms where winning absolute top still beats your incremental cost per acquisition. Use bid simulators to estimate the cost of that jump in position first.
  • Watch Absolute Top impression share and impression share lost to rank, because that’s how you see the AI block pushing you down the page.

What you shouldn’t do is pour Fight budget into informational queries the Overview has already answered.

See exactly how your competitors win.

Uncover the keywords, ads, landing pages, and strategies driving your competitors’ paid search success—and find your next opportunity to outperform them.

Analyze your competitors

Bucket 2: Influence

Influence is for queries where the Overview, or a longer conversational answer, does the educating. Very often, the user won’t click.

The goal here isn’t to gain a cheap visit anymore. It’s to be part of how the answer gets built: cited as a source or presented as one of the brands named inside the answer.

As Shashi Thakur put it at Google Marketing Live EMEA 2026, the best ads must be answers. In an Influence query, your ad brief starts from the question the user is really asking, rather than from a keyword list, and it means accepting that SEO and PPC address the same challenge, not competing against each other.

In B2B and industrial accounts, the queries that trigger AI answer blocks are still mostly informational. In one of our accounts, the intent split of the terms surfacing these blocks looked like this:

Intent Share
Informational 83.3%
Learn and solve 10.0%
Commercial 6.7%

For “learn and solve” and some “informational” terms, plan for share of the answer instead of maximum CPC.

There is a budget consequence worth stating plainly. On influence queries, we work very closely with SEO, and we’re usually not interested in paying to appear where organic is already winning. Being visible twice, in paid and in the answer, barely happens anyway.

When the data pushes us to move money into the Fight and Generate demand buckets, we deliberately avoid being aggressive on terms where SEO already earns the visibility. That keeps paid spend pointed at the gaps we’re interested in covering.

Bucket 3: Generate demand

Generate demand is for queries where zero-click has already won and paid search alone won’t rebuild the top of the funnel. Informational demand that used to create assisted journeys now often ends on Google, with no visit to anyone.

You have two options. You can:

  • Keep bidding while your CPCs climb and volume falls.
  • Fund the work that happens before the search: Demand Gen, YouTube, first-party content, and the review and community presence that shows up later when the same person searches with commercial intent.

This work is about recognizing that some keywords have stopped being capture channels and have become a signal of a demand gap. The capture still happens later, in Fight, once the person is ready to act.

Generate demand fills the pool that Fight then fishes from. Measured only on last-click Search ROAS, this work always looks weak because its payoff appears as brand searches and better-qualified pipeline weeks later.

Bucket The job The paid response
Fight Clear purchase intent, where the click still feeds the sales pipeline directly Pay for absolute top when your brand isn’t already winning the answer box
Influence Top and mid-funnel, where the answer shapes consideration Win share of the answer through citation and SEO coordination, not raw CPC
Generate demand Zero-click is winning and Search alone can’t rebuild the funnel Build demand before the search, through Demand Gen, YouTube, and awareness

Dig deeper: What happens when AI Overviews contradict paid search ads?

Why one bid rule fails after AI Overviews

AI Overviews sit at the top of the page, sharing space that used to belong to ads and organic results. The answer often satisfies the user before they scroll, changing the economics of paid search in three ways.

1. Click-through falls on informational and mid-funnel queries

When users get a usable answer without leaving Google, both organic and paid lose clicks. The effect is stronger when your brand isn’t mentioned in the AI Overview.

The clicks that remain tend to be better qualified, but the drop in volume usually exceeds the lift in quality, so total conversions decline.

2. Cost per click goes up

There are fewer premium positions at the top. More advertisers are shifting their budgets to paid search to offset lost visibility.

The pressure is strongest in B2B and services, and on comparative queries such as “best X” or “X reviews.” Pure transactional queries hold up better because users still need to click through to buy or request a quote.

3. Position gets harder to read

There are often only a few useful slots above the AI Overview or AI Mode answer, and falling below the Overview can sharply reduce click-through, especially on mobile. Absolute Top impression share and impression share lost to rank become especially useful for understanding that pressure.

Underneath these changes is a shift in behavior. People can research and decide within a single answer experience, shortening the window to capture and persuade them. Intent still lives in your account, but the economics attached to each type of intent have changed. That’s why you should classify a query before you touch the bid.

Dig deeper: 4 strategic paid search pivots to survive Google’s AI Overviews

Get the newsletter search marketers rely on.

See terms.

How to classify your terms in practice

The framework only pays off if your account can distinguish between valuable and low-value traffic, so two foundations come first.

One is semantic coherence: keyword, ad, and landing page as a single unit of meaning, with each ad group built around one clear concept. Mixing concepts teaches the system the wrong matches, and that gets worse the moment you hand control to broad match, Performance Max, or AI Max.

The other is conversion quality: primary conversions that reflect real business value, lead scoring, and offline or CRM signals wherever you can pass them back. Smart Bidding and the AI surfaces amplify whatever you give them, so thin data makes every bucket noisier.

With those in place, classification becomes a routine you can repeat:

  • Pull your search terms and map intent, separating clear commercial intent from research and from purely informational queries.
  • Use SEO data, Search Console, AI Overview Reports, plus manual SERP checks as a proxy for where Overviews appear and whether your brand is cited.
  • Sort each term into Fight, Influence, or Generate demand, and revisit the sorting regularly because the SERP keeps changing.
Step Action Data Sources Output
1. Map intent Pull search terms and classify by intent type. Google Ads keywords and search terms (+ Search Console). Terms grouped by: commercial, research, informational.
2. Identify AI presence Check where AI Overviews appear and whether your brand is cited. AI Overview reports, manual SERP checks. Map of AI Overview presence by term.
3. Sort into buckets Assign each term to a strategic bucket. Intent mapping + AI presence data. Fight, Influence, or Generate demand list.
4. Revisit Regularly Re-sort terms as SERP landscape changes. Ongoing SERP monitoring, AI Overview reports. Update bucket assignments and review budget assigned to each.

On Fight terms, accept a position premium only where absolute top still wins your incremental cost per acquisition.

On Influence terms, align ad copy with the question the Overview answers, coordinate with SEO on content and citation, and pull back paid pressure where organic already wins.

On Generate demand terms, move budget into YouTube, Demand Gen or content when the Search clicks are gone but the query still matters and measure through brand search and assisted pipeline.

AI Max and broad match sit on top of this classification. They don’t replace it. Google needs broad match or keywordless campaigns for your ads to appear around AI answer surfaces, which expands reach while reducing control over what you show.

Use them where you already know the bucket and your tracking is clean, and introduce them through small, gradual tests.

Dig deeper: How to get your Google Ads seen in AI Overviews

Every click they win is a customer you lose.

See where competitors are investing, which keywords drive their results, and how to capture more of the market.

See who’s stealing your traffic

Put the budget where the opportunity is

Accounts that apply one rule to every query will keep funding the wrong mechanism, seeing it as rising costs they can’t explain. Accounts that re-segment and move money between buckets and into channels outside the auction spend less time arguing with a results page that has already answered the user.

In B2B and industrial search especially, shifting from optimizing bids to reallocating budget across the three buckets is the difference between defending last year’s setup and building for how people search now.

Read more at Read More

3 ways to make AI safer in a live ad account by Optmyzr

/0 Comments/in , /by

It’s been a pretty crazy few weeks in AI safety. We’ve watched agents go off the rails and do things nobody expected them to do, in production, on other people’s systems. So it’s a reasonable moment to ask how you use AI safely when the thing it’s touching is a live ad account with real money moving through it every hour.

Almost every conversation I have about agents in ad accounts opens the same way.

Do you trust the AI?

While it’s a good question, it’s often asked to shirk responsibility and conclude that AI shouldn’t be used. A better question is: How can we trust AI?

Then the conversation turns to what we can build around the AI model to make it safe for our business to use.

To explain, it’s always helpful to frame it the way you’re already evaluating collaborators. Nobody asks whether they trust a new PPC agency in the abstract. They ask what the team has access to, what they’re allowed to change without checking first, and who reviews the work. Three different questions with three different answers, and you’d never accept “I simply trust them” as a substitute.

Same three questions for an agent:

  • What can it see? An agent working off a thin data layer will give you a confident answer built on a third of your account. It won’t tell you it’s guessing when it doesn’t know it’s guessing.
  • What is it allowed to do? Not what you told it to do. What it is structurally prevented from doing, no matter what anyone tells it.
  • Who signs off? Not “we review the change history afterward.” Who has to say yes before?

Most agentic PPC setups I look at have a decent answer to the first question, but there’s a lot of benefit in becoming more stringent with the answers for the other two.

We’ve been thinking about safe AI for PPC a lot at Optmyzr, so let’s take a look at what we’ve learned so you can get yourself to safer AI faster.

Each layer pays for itself on its own

The reason I’d rather write this as three techniques than as one system: you don’t have to do all three to get value. Each one closes a different failure mode, and each one is useful the day you turn it on.

  • Layer 1: Better grounding means fewer confidently wrong answers. That’s worth having whether you ever let an agent write anything. Most people should start here.
  • Layer 2: A policy layer means the changes that do get made stay inside limits you set. That holds whether the change came from an agent, from a script, or from a person having a bad Tuesday.
  • Layer 3: A review step means nothing reaches the account without someone seeing it first. And as a nice side benefit, you end up with a record of why. That’s worth having even if your data layer is thin and you have no policies at all.

Then these layers compound, which is the part I find genuinely satisfying. Grounding makes the agent’s proposals worth reviewing, so the review step feels like leverage instead of homework. Policies filter the obvious non-starters before a human ever sees them, so the queue stays short enough that people keep opening it. Each layer makes the next one work better than it would on its own.

So, if you’re starting from zero, start with grounding. It has the fastest payoff and the least process to stand up. If you’re already letting an agent make changes, add policies this week. If you have both, the review queue is what turns it from something you use into something your team uses.

Order matters less than accumulation. Any one of these leaves you better off than you were yesterday, and each one you add compounds with the last.

Layer 1: Ground it

A blind agent is a dangerous agent.

Ask an agent connected to a thin data layer why your CPA went up last month. It will answer you. Fluently. Immediately. Based on whatever slice of your account it could actually reach.

What it couldn’t see might have been the entire answer. It’ll make do with what it has.

Realize that grounding is a safety feature, not a convenience feature. Every gap in what an agent can see is a place where it will make something up, and it will do it in exactly the same confident tone as everything else it says. There is no tonal tell.

So here’s what I think you should expect from an MCP, or any data layer, that you’re going to let an agent reason over for PPC.

The full query layer for Google Ads. Real GAQL. Any resource, field, segment or metric the API exposes, including the ones no packaged report covers. Not a curated summary of what a product manager thought was interesting. The moment your data layer is a curated subset, you’ve limited which questions the agent can answer well, and you haven’t told anyone which ones those are.

  • GA4 sitting alongside the ads data. So “what happened after the click” is part of the same question, instead of a second tool and a manual join. A lot of the diagnosis questions people actually ask are cross-boundary questions. If the boundary is still there, the agent guesses across it.
  • Complete change history, every actor. UI edits, scripts, Optmyzr, other tools. So “who made the change that moved our ROAS in March” is a question with an answer, rather than a group chat.
  • Negative keywords consolidated across all four levels. Account level, shared lists, campaign, ad group, plus a deterministic check of whether a given query is already blocked and by which negative, and a list of campaigns sitting with no negative protection at all. Negatives are a common place to watch an agent reason confidently and wrongly, because the true state is scattered across four places and nobody assembles it.
  • Auction Insights, with a drill-down into one competitor domain, shows your own performance on every keyword you share with them. Competitive questions are the ones where a hallucination is hardest to catch, because you have no independent read on the answer.
  • Vertical benchmarks. Your CTR, CPC, conversion rate, and impression share as a percentile against other accounts in your industry, rather than against a blog post average from three years ago that everyone quotes and nobody sources.
  • Multiple ad platforms. Google, Microsoft, Meta, Amazon, LinkedIn, OpenAI, TikTok, Yahoo. Budget questions are rarely single-platform questions, even when the person asking works mostly in one.
  • A stored profile of each account. Business model, economic posture, bid strategy mix, structure, budget behavior, what’s already been tried and what happened when you tried it. The agent reads this before it opens its mouth.

Optmyzr’s MCP has all of the above today, and it’s a one-click install from the Claude directory rather than an API console, a developer token, or an engineer on speed dial.

Grounding raises the quality of everything the agent says. The next layer decides what it’s allowed to do about it.

Layer 2: Gate it

How do you stop an AI from blowing through a month of ad budget by mistake?

Not by asking it nicely. Not in the prompt.

You use a policy layer that controls what an AI can and cannot do, and you separate it from the AI itself, so the AI can never change the rules.

Set the rules once. Enforce them everywhere, always.

This is automation layering, which I’ve been writing about across my books for years, applied to a new first layer. The original idea is simple: one system does the work, like Google’s own bidding and budget automations, and a second system, your own automations, scripts or rule engine strategies, validates that what the first system did actually makes sense for your business before it sticks.

AI just took over the first job. It’s now the thing making the recommendation. The second layer didn’t become less necessary; it became more necessary because the first layer is more creative and sometimes unpredictable.

Account policies are how you write down the “never do this.” Rules you set once, on the account itself, about what is allowed to happen there. No bid increase above 10% in a single move. No budget change beyond a set threshold. These campaigns don’t get touched. No competitor brand terms added. Whatever your version of “absolutely not” happens to be.

And the policy doesn’t care who’s asking.

An agent proposing a 20% bid increase gets blocked. A hallucination gets blocked. An instruction hidden in a document gets blocked. A junior with a misplaced decimal gets blocked. You, at 11pm on a Friday, in a hurry, on your phone, get blocked.

Same rule, same verdict, no exemption for good intentions or seniority.

Want it through anyway? Override it deliberately. It goes on the record with your name attached. An override you can perform without noticing is not a guardrail but a speed bump made of paint.

This structural detail matters: this isn’t a setting for the AI. It’s a setting on the account, and the AI is one more thing subject to it. Same as a script. Same as a person.

What PPC account policies look like. Screenshot from Optmyzr.com, September 2026.

A rule that lives in the prompt is a rule the model can be talked out of by a clever user, by an injected instruction sitting in a document it was asked to read, or by its own drift over a long session. A rule that lives on the account holds under every path into the account.

Layer 3: Keep a human in the loop

Everyone says they’re in the loop.

Few people can tell you which screen, which person, or which queue enforces being in the loop. Ask, and you usually get “we check the change history afterward,” which means nobody’s checking.

And good luck getting a quick answer when someone asks you to produce the reasoning behind an AI-assisted decision from three months ago.

So when we built the write path into our MCP, we didn’t leave the loop to good intentions. There is no route from the agent to your ad account that doesn’t first stop at a human. 

We took a pattern from engineering, where it’s been settled practice for decades: nobody pushes code to production without a change request that another person reviews. Why the account where you spend six figures a month deserves less process than a CSS tweak is a question our industry has never really answered.

Here’s the actual sequence.

  1. The agent proposes. Every write, whether a bid, a budget, a paused campaign, or a new negative, becomes a draft change request. Nothing reaches the ad platform yet. The agent’s job ends at “here’s what I think you should do, and why.”
  2. Policies evaluate it. Each row is checked against your account policies and carries its verdict with it, so a blocked row shows up as blocked, with the reason attached rather than in a log somewhere.
  3. A human opens the review. The rows, the stated reasoning for each one, the policy warnings, a timeline of what happened when, and the list of people eligible to approve it.
  4. You preview exactly what would go live. The exact deterministic changes, like your target ROAS for “Brand Campaign” will go from 200% to 220%, rather than a natural-language paraphrase of what the agent believes it’s about to do.
  5. You confirm. Only then does anything change in the ads account.
What change requests look like in a PPC account. Screenshot from Optmyzr.com, September 2026.

Whether it’s an AI or a colleague suggesting the change, everything goes through the same pipe. In the case of a colleague, the change request is a second pair of eyes. In the case of an AI, it’s the first pair.

An audit trail for AI in PPC

The change request queue turned out to be more valuable than just the safety it was built for.

Because it’s not just a list you go to in order to be the human in the loop. It’s a complete record of intent.

When a client asks in November why their target CPA was moved in March, you have the proposal, the rationale, the policy verdicts, who approved it, and when. Was it a human? Was it an AI? What was the data behind the recommendation? Who made the final call?

Try assembling that from a chat transcript six months later. Try assembling it from change history, which tells you what changed but never why.

We built this for safety, and it turned into the best account documentation we’ve ever had. If you’re an agency, that’s a credibility argument as much as a safety one.

What good looks like

Put the three layers together, and you get something I’d describe, approvingly, as boring.

Not underpowered. Boring. 

As in: you know what it can see, you know what it structurally cannot do, and you know nothing reaches the account without you. The excitement belongs in the findings, not in wondering what it got up to while you were at lunch.

That’s the bar I’d hold any agentic PPC setup to, ours included, and it’s a bar you climb one rung at a time:

  • It sees the whole account because gaps in what an agent can see are where it starts inventing, confidently.
  • It’s bounded by rules you wrote, which live on the account rather than in the prompt, and which apply to everyone identically.
  • It can’t act alone, because every write becomes a change request with policy verdicts attached, reviewers named, and a confirmation step no agent can fabricate.

You don’t have to arrive there in one move. Pick the layer that closes your biggest gap, ship it, then add the next one.

We’ve been building this against real accounts and genuinely stupid edge cases since well before MCP was a term marketers used. It’s a one-click install from the Claude directory now.

If you tried agentic PPC once, got a confidently wrong answer, and quietly shelved it, that’s the failure I’d most like you to come back and retest.

Having safe AI for our ad accounts can’t be something we expect to get just from picking the right model. It’s something we achieve by layering in processes and technologies we control.

Try Optmyzr’s safe AI for PPC free for 14 days.

Read more at Read More

How to Create A Landing Page That Actually Converts

/0 Comments/in /by

Key Takeaways

  • A high-converting landing page drives one specific action. Every element on the page should serve that single goal.
  • The median landing page converts at 6.6 percent across all industries. Your target should reflect your vertical, since averages range from under 2 percent in ecommerce to over 10 percent in other industries.
  • Tools like Unbounce’s Smart Copy speed up copy iteration, and its Smart Traffic feature dynamically routes visitors to the variant most likely to convert. Human review is still what catches poor-quality output, however.
  • Cart abandonment averages over 70 percent industry-wide. Landing page sign-up and checkout flows suffer from similar friction.
  • High-converting landing pages follow six core formats: text-only, lead generation, click-through, long-form, video, and SEO. Each format serves different goals, so choose yours based on the action you want visitors to take.

Ever wondered why your landing page pulls in traffic but barely converts anyone?

You’re not alone. 

The median landing page converts about 6.6 percent of visitors, and plenty of pages fall well below that. 

The problem is rarely traffic or budget. It’s the page.

Effective landing pages follow a proven structure, with every element earning its place through testing. I’ll walk you through how that process works so you can stop guessing and start optimizing with a clear plan.

What Is a High-Converting Landing Page?

A high-converting landing page is a webpage that excels at getting visitors to take one specific action. That could be signing up for a newsletter, registering for a webinar, claiming a discount, making a purchase, or downloading your app. 

Here’s an example from Revolut that directs visitors to download the digital banking company’s app: 

Revolut’s landing page displays their headline, “Change the way you money,” in bright white text on a blue sky background. A woman stands inside a square “smartphone” frame overlaid with widgets from the app’s dashboard, displaying how you can move and track your money.

A strong landing page turns visitors into new subscribers and fresh leads, which you can nurture through your sales funnel until they’re ready to buy.

Core elements include a compelling call to action (CTA), a sign-up form, an attention-grabbing headline, and subheadings. It also contains easy-to-read content, images, and perhaps a video.

High-priority elements like your sign-up form should go above the fold (the visible part of the webpage when you first see it) for maximum visibility. I’ll cover placement strategy in more detail later. 

Types of High-Converting Landing Pages

High-converting landing pages come in several formats. Here are the ones worth knowing about.

Text-Only Landing Pages

These skip the videos and heavy graphics, relying on copy and a few supporting images to keep the focus on the message.

Copyblogger is a strong example that pairs minimal visuals with a clear CTA button.

Copyblogger webpage featuring a bold headline stating, “The most important skill in business is the ability to move people with words.” Below the headline is a subscription form titled “Build a Real Business With Content.”

Lead Generation Landing Pages

Lead generation landing pages collect details such as a name, email, and phone number. The idea is to turn prospects into leads by offering a free lead magnet, such as templates, a short course, eBooks, or access to a webinar.

This is how Workhuman uses them:

Workhuman’s landing page for their webinar, “The Power of Beliefs: How 7 Core Beliefs Shape Performance, Resilience, and Wellbeing at Work,” showcases how landing pages can generate leads through a capture form and engaging copy.

Although the form is relatively extensive (with several additional fields out of view), it makes segmenting and targeting easier. Its basic but striking design keeps it clutter-free while relaying the essential information.

Click-Through Landing Pages

Click-through landing page conversions aim to get people to take an action, such as signing up for a freebie, requesting a product demo, registering for an event, or buying a product. Typical features include a clear CTA, an engaging headline, and magnetic content designed to persuade the reader.

Shopify uses an enticing discount offer for new users and features a single-field sign-up form and a simple yet tempting CTA button.

Shopify entices users to sign up for a free trial by touting the millions of merchants who have “made it” on its platform. They also reference a $ 1-per-month subscription offer and $15,000 in credits for users who sell on their platform.

Long-Form Landing Pages

A long-copy landing page is common in the digital marketing industry. It’s a go-to format for showcasing the benefits of your products or services to close a sale.

These longer pages use strong headlines and subheadings, features and benefits, images, social proof (e.g., reviews), and CTAs throughout. They might also include case studies, trust badges, and a video chat option, as SuperOffice does.

SuperOffice’s landing page uses white text on a green background, as well as several visual elements like trust badges, client logos, and dashboard screenshots to show potential customers the benefits of their platform.

Video Landing Pages

Video landing pages feature video as the primary content element. 

Why? Video can improve engagement and conversions and reduce bounce rate by keeping visitors on the site longer. It also builds trust with prospects before they commit. 

This format works well for products or services that require setup or configuration. When you visit Slack, for example, you’ll find a video that walks you through what the messaging software does and how to use it.

Slack’s landing page uses an embedded video, which they invite visitors to watch with the headline, “Discover a better way of working.”

Video also brings reach. Digital video has a worldwide audience of 3.88 billion, just over 63 percent of whom watch on smartphones. 

Demand keeps growing, too.

A Wyzowl survey finds that 84 percent of consumers want more videos, and 96 percent have watched an explainer video to learn about a product or service.

Campaign vs. SEO Landing Pages

You can build landing pages for two broad purposes: long-term organic visibility or short-term campaign performance.

An SEO landing page targets organic traffic by answering specific search queries. These pages need ongoing optimization to maintain rankings. 

Zendesk, for example, uses a clean, structured features page with short text paragraphs and updates it regularly as new product features roll out.

Zendesk’s example of an SEO landing provides answers for people looking to learn more about product features. 

A campaign landing page supports a specific product launch or marketing push. It relies on paid advertising to drive targeted traffic and uses frequent A/B testing to find the best-performing version. 

Here’s an example from Canva, which uses a campaign landing page to promote a free graphic design course and drive sign-ups through its Design School.

Canva’s landing page shows elements like headlines, graphics, and CTA buttons that are typically A/B tested to optimize performance and push visitors to take action. 

How to Create a High-Converting Landing Page

Here’s my five-step process to build a strong landing page.

1. Choose Your Tools

Several tools can help you start building landing pages:

  1. Unbounce: A drag-and-drop landing page builder with no-code A/B testing, AI-powered copywriting, and real-time analytics. Free 14-day trial.
  2. OptimizePress: A WordPress-based builder for conversion-optimized landing pages, funnels, and membership sites. 14-day money-back guarantee.
  3. Instapage: A landing page platform for building, optimizing, and scaling personalized pages. Free 14-day trial on Create and Optimize plans.
  4. Leadpages: A builder with optimized templates, lead magnet delivery, and built-in analytics. Free seven-day trial.
  5. GetResponse: A marketing platform with a drag-and-drop landing page creator and AI content generator. Free 14-day trial.
  6. PopUp Domination: A pop-up tool that adds email opt-in forms to your landing pages to capture more subscribers. Works with every web host and email provider. Free 14-day trial.
  7. OptinMonster: A lead capture tool that adds pop-ups, slide-ins, and other opt-in forms to your pages to grow your email list. 14-day money-back guarantee.

2. Conduct Market Research

Every good landing page starts with a thorough understanding of your target market and customers, so you can create value and provide a desirable customer experience. 

Research your target market to understand what your customers need, the language they use, and what might prevent them from converting. That might be price concerns or unclear value, for example.

From there, dig into keyword research to find the search terms your audience uses, including long-tail keywords that signal high intent. Run a competitive analysis to see what’s already ranking and where the content gaps are. 

My free tool, Ubersuggest, can help with all of this. Here, it shows keyword ideas for “Starbucks” broken into suggestions, questions, and prepositions, each with search volume data. 

An Ubersuggest keyword report for “Starbucks” shows the information available to you when conducting keyword research.

These categories help you understand how your audience searches for a topic so you can match your landing page copy to real user intent.

3. Design With Conversions in Mind

A high-converting landing page brings together several design elements that work toward a single goal. Here’s what the anatomy looks like:

"The Anatomy of a Perfect Landing Page" illustrating key design elements for effective user engagement. It highlights numbered components, including clear headlines, concise body copy, trust indicators, standout buttons, easy navigation, relatable images, and a strong call-to-action, with color-coded labels and brief descriptions for each element.

Some elements deserve extra attention:

  • Headline: It’s the first thing people read, so make it bold, clear, and benefit-driven. It also needs to be creative and to the point while creating a sense of urgency. Alternatively, use it to target specific customers and offer a solution to their problems.
  • Subheadlines: Take the same approach with subheadings and keep them short and direct. Ultimately, subheadings should reinforce the headline’s message and keep readers moving down the page.
  • Call-to-action button: Your CTA button needs to stand out visually. Use a contrasting color and keep the text simple, with wording like “Download now” or “Buy now.”
  • Videos and images: Showcase your product and feature testimonials with videos and imagery. Consider including a hero image (a large, eye-catching image that visitors will see first). Headshots also work well.
  • Trust badges: Awards, trade badges, company logos, and testimonials are trust indicators that help reduce hesitation for visitors who are close to converting.

I cover more of these principles in my guide to what high-converting websites do differently.

4. Ensure Mobile-Friendliness

Mobile devices account for more than 50 percent of global web traffic and generate 70 percent of online shopping orders.

That’s why your landing page needs to work just as well on a phone as it does on a desktop.

Google recommends a few core practices for mobile landing pages:

  • Keep the layout clutter-free, easy to read, and fast-loading. Google advises avoiding pop-ups.
  • Include intuitive navigation like a “home” page button on every page and a drop-down menu.
  • Create punchy, relevant copy with a clear tone.
  • Consider using Google Sign-In to simplify data collection.

Google has more tips on optimizing your mobile landing page.

5. Test and Optimize

Testing page variations is how you go from guessing to knowing what works. It’s how you know which version converts best so you can use the results to inform future landing page designs. My A/B testing introduction covers the fundamentals. 

Start by testing one variable at a time:

  • Button colors or page backgrounds
  • Long versus short sign-up forms
  • Content order, images, or copy

Run each test for at least two weeks, then compare conversion and engagement rates to make data-driven decisions. Optimizely, Crazy Egg, and Hotjar all offer A/B testing tools. Check how your page renders across devices with MobileTest.me.

Here are some best practices for landing page optimization:

  • Use a clear user interface: For your landing page to work, ensure the user interface uses simple navigation and other elements, nudging users toward one clear action. The focus is on your visitor, not you. Align every element on your landing page to appeal to the end user. David Risley’s landing page for Blog Marketing Academy is a great example.
  • Use SEO: Optimize your landing page with keywords by performing keyword research and understanding search intent.

Keep a narrow focus as you optimize. The clearer and simpler your page, the more likely you’ll get someone to take the action you want.

Ryan McHugh, the director of CRO at NP Digital, puts it this way:

“One of the biggest mistakes site owners make when building landing pages is not having a clear, single focus for the page. Landing pages should have one primary objective, whether it’s capturing email addresses, getting sign-ups, or making a sale. When a page has too many distractions, including links, navigation options, or too much information, it can confuse visitors and decrease conversions. Keep the content blocks concise and relevant. Focus on a clear value proposition and be sure to place a consistent and action-oriented call to action prominently on your pages.”

A few ways to put that into practice:

  • Move non-essential information to the “About” section. 
  • Keep forms short. 
  • Make the CTA the focal point and place secondary links below the fold.

Design Best Practices

Your color palette, typography, and visual hierarchy all affect whether visitors convert or bounce. Here are the design details to get right.

  • Think about color psychology: Colors shape brand perception and influence buying decisions. If you’re using a solid background color, make sure the text contrasts enough to stay readable. A dark background needs light text, and vice versa.
  • Use clean and legible fonts: The most popular fonts for a landing page are PP Neue Montreal, GT Super, Aeonik, and Inter, all chosen for readability and impact. 
  • Be selective with links: Every link on your landing page is a potential exit. If your goal is to capture leads, minimize outbound links and keep the focus on your CTA.

One more thing: Keep your visuals tidy and polished.

If you’ve spent more than 30 minutes on the internet, chances are you’ve seen one of those ads with an annoying GIF or clickbait headline like “Lose 50 lbs with this one weird trick.” Avoid distracting elements on your website that may sidetrack visitors from the main message.

Wix does this by using a limited but eye-catching color palette, a contrasting CTA button with an enticing offer, and inspirational language. 

Wix’s landing page uses a minimal yet eye-catching color palette, paired with an inspirational headline, “The new way to create a website,” and a bold “Get Started” CTA button.

Finally, be mindful of mobile users. I can’t tell you how discouraging it is to visit an online store on my tablet only to find that the CTA button overlaps with the pricing.

Copywriting Best Practices

Good design gets visitors to your landing page. Good copy gets them to convert. Here are the copywriting best practices for high-converting landing pages. My copywriting guide goes deeper on these concepts, but these fundamentals will get you started.

Include Very Important Attributes (VIAs):

Showcase the VIAs that appeal to visitors. Identify two to five key features, benefits, or pain points, and highlight them on the front page rather than overwhelming visitors with a long list.

Verint’s VP of Content, Jasen Williams, shared a testimonial on how NP Digital’s SEO strategies have been transformative to their website performance. 

Use the AIDA Model

AIDA stands for attention, interest, desire, and action. Apply this practice to content creation by using captivating graphics to drive attention, hold interest with valuable context, stimulate desire by connecting it to product benefits, and lead to action with a CTA. 

Additionally:

  • Format your copy so it’s easy to read: Use an organized structure full of headings, bullets, and quotes, and don’t be afraid of white space. This makes your site look less cluttered.
  • Cut your fluff: Are you familiar with the phrase “kill your darlings”? You’re about to do that here by cutting anything that doesn’t add value. Use simple language and keep paragraphs short.
  • Social proof: You can add things like media mentions, accreditations and awards, and usage stats.
  • Don’t forget large language models (LLMs): You have to consider AI platforms and the LLMs that power them. LLMs tend to surface content that answers questions directly and quickly. Structure your copy around longer-form, question-based queries, and lead each section with a clear answer to improve your chances of appearing in AI-generated results.

Landing Page SEO Best Practices

SEO for landing pages comes down to a few essentials:

  • Headings: Use H1, H2, H3, and H4 tags to give your page a clear structure for both readers and search engines. Include your main keyword in the title and H1, and work secondary keywords into your other headings.
  • Meta tags: Your title tag and meta description tell search engines what your page covers and shape how it appears in search results. Keep them concise and keyword-relevant.
  • Alt text: Alt text describes your images for search engines and screen readers. It supports both SEO and accessibility, so treat it as a requirement.
  • Local SEO: If your landing page targets local customers, develop localized content that drives traffic to your door.
Ubersuggest’s Google snippet contains the fundamentals of landing page SEO best practices, showcasing an optimized page title and meta tags. 

Search intent ties all this together. Every element on your page should reflect why someone is searching, not just what they’re searching for.

LLMs and AI search features reinforce this principle. Follow SEO fundamentals and focus on answering question-based queries directly to keep your landing page visible in AI results alongside traditional search.

How AI Is Changing Landing Page Optimization

While AI is rooted in traditional SEO concepts, it’s significantly changing the creative side of landing page workflows. 

Copy that once took a week to draft can now be generated and tested in an afternoon. A/B testing now happens dynamically, with visitor traffic routed to the best-performing test variant in real time.

Additionally, search itself has shifted, with AI Overviews providing zero-click information that used to require a click-through. 

Each change reshapes what you should build. 

AI Tools for Copy and Page Creation

AI writing tools have dramatically sped up the timeline for drafting landing page copy. What used to take a week of copywriter iteration can now be a starting draft in minutes, with multiple variants ready to test by the end of the day. 

Tools like Unbounce’s Smart Copy do this directly inside its page builder.

Fast doesn’t necessarily mean good, however. 

Unbounce’s Conversion Benchmark Report shows that the correlation between difficult words on a page (three-plus syllables) and its conversion rate is 62 percent stronger than it was in 2020, and pages written at a fifth- to seventh-grade reading level convert more than two times higher than professional-level writing. 

Raw AI output tends to be wordier and more formal than what converts. It’s great for accelerating the creation of variants, but think twice before skipping a human read that catches corporate voice creeping back in.

AI-Powered Testing and Traffic Optimization

Traditional A/B testing splits traffic 50/50 and waits for statistical significance before declaring a winner, which usually takes weeks and burns through budget on the losing variant the whole time. 

Thankfully, we can avoid some of that loss now with AI-powered traffic optimization. Tools like Unbounce’s Smart Traffic analyze each visitor’s attributes in real time (device, referrer, campaign source, and browsing behavior) and route them to the variant most likely to convert for that specific profile.

That means you can run multiple variants without paying the full traffic cost of a losing test, allowing you to optimize continuously rather than in discrete two-week cycles.

<h3> AI Search and AEO Implications for SEO Landing Pages </h3>

Traditional SEO best practices still apply in AI search, but AI Overviews and answer engines create some nuance. 

According to HubSpot, nearly 30 percent of marketers report decreased search traffic as consumers turn to AI tools, and over 92 percent are already optimizing for both traditional and AI-powered search.

For SEO landing pages, answer engine optimization (AEO) is now as impactful as classic on-page SEO. 

Pages that answer specific questions directly and carry visible experience, expertise, authoritativeness, and trustworthiness (E-E-A-T) signals are what AI systems draw on when generating answers. The signals are concrete: author bylines, credentials, client results, and publish dates.

Clean heading hierarchy and structured data make it easier for AI to parse your content and pull from it accurately. Pages built on keyword density without substance get passed over.

What’s a Good Landing Page Conversion Rate?

A “high-converting landing page” only means something once you have a benchmark to measure against. That benchmark varies by industry and traffic sources, and it shifts depending on where the page sits in your marketing funnel. 

Three data points can help you set realistic targets.

Average Conversion Rate by Industry and Channel

Unbounce research puts the median at 6.6 percent across all industries. 

Email is the highest-converting traffic source at 19.3 percent on average, well ahead of paid social (12 percent) and paid search (10.9 percent). Google leads among paid search channels.

The device people use to find you plays a role, too. Unbounce found 83 percent of landing page visits happen on mobile, yet desktop still converts about 8 percent better on average. 

The same report links simpler, shorter copy to higher conversion rates, which is why readable pages consistently outperform corporate-voice ones.

Conversion Rate by Industry: Ecommerce, SaaS, and More

Industry averages swing hard from the overall median. Statista puts the average conversion rate across all ecommerce sites at under 2 percent, a fraction of the 6.6 percent Unbounce reports across landing pages more broadly.

In ecommerce, skincare pages convert the highest at 2.4 percent (above average), while luxury apparel pages bottom out at just 0.3 percent.

The takeaway here is that chasing a single universal benchmark is a mistake. Compare your ecommerce conversion optimization against the tightest peer set you can find, then work on the specific friction points holding them back. 

Cart Abandonment and Checkout-Stage Conversion Loss

According to the Baymard Institute, the average cart abandonment rate is 70.2 percent, based on 50 different studies. 

Excluding pure window-shopping, the top reasons shoppers bail are:

  • Extra costs like shipping and fees (40 percent)
  • Slow delivery (20 percent)
  • Distrust of payment security (19 percent)
  • Forced account creation (18 percent)
  • Overly long checkout flows (17 percent)

Baymard estimates that a better checkout design alone could recover $260 billion in lost ecommerce orders in the U.S. and the European Union. 

The same friction points hit landing page sign-up and checkout flows, too. Anywhere you ask a visitor to take an action, reasons like those above are what keep them from following through. 

Audit your page against that list.

Want to See These Best Practices in Action?

Reading about what works only gets you so far. To see how real brands apply these principles, check out my breakdown of over 30 landing page examples that convert, along with the specific design choices that make them work.

FAQs

What are high-converting landing pages?

High-converting landing pages are web pages designed to encourage visitors to take a specific action, such as signing up for a newsletter or making a purchase. These pages are optimized for conversion rates, designed to maximize the number of visitors who take the desired action.

What are the key components of a landing page?

Every landing page needs a clear headline, benefit-focused copy, a compelling CTA, and social proof. Trust signals and mobile-friendly design round out the essentials.

How do you create a high-converting landing page?

Start with a clear, compelling headline that highlights your product’s main benefit, and make sure your copy focuses on your target audience’s needs and pain points.

It’s also important to include high-quality images and videos that showcase your product in action and give visitors a sense of its capabilities, include a CTA, and keep text brief. Always A/B test content to see which option works best.

Finally, be sure to include testimonials, simplify navigation, and clearly state your value proposition.

What is a good landing page conversion rate?

Unbounce’s Conversion Benchmark Report puts the average conversion rate at 6.6 percent across all industries, but you shouldn’t be chasing a generic number. Your target should reflect your industry, since individual averages range from under 2 percent in ecommerce to over 10 percent in other verticals.

How can you improve your landing page conversion rate?

Start with a message match between your ads and page copy, since a mismatch kills conversions faster than any design flaw. Simplify your forms, then A/B test your headline and other elements.

Do landing pages help SEO?

Yes, when built for it. Landing pages that match specific search intent and load quickly can rank organically and drive long-term traffic beyond your paid campaigns.

Can you have too many landing pages?

The consensus is no. Digital marketing experts agree that it makes sense to have multiple landing pages. For instance, you might want to target different demographics.

Conclusion

There’s no single formula for a high-converting landing page. Don’t be afraid to test different approaches and see what works for your company and your offer.

Start by getting to know your buyers and what appeals to them. Get a sense for their pain points, and then build pages that solve them. Pair that with a CTA that compels them to take action and watch your conversions grow. 

Apply the optimization elements in this guide, then let A/B testing strategies show you what converts best.

Read more at Read More

How to Audit Any Page in Under 3 Minutes with SEOquake

/0 Comments/in /by

You’re looking at a page that should be performing better, and you want answers fast.

Maybe it’s a product page that slipped in rankings, a competitor’s post outranking yours, or a client page that isn’t converting.

You don’t need a full site crawl to start diagnosing.

SEOquake’s free Chrome extension audits any page and tells you what’s working (and what isn’t), all without leaving the page.

In this tutorial, I’ll walk you through an SEOquake workflow that takes you from triage to diagnosis in just three minutes:

  • Triage: Open SEOquake’s Quick View pop-up for an instant SEO snapshot (1 minute)
  • Diagnose & Decide: Switch to the report for deeper analysis and exports. Then, decide whether it’s a page edit or a wider issue that needs a full site audit (2 minutes).

What Is SEOquake?

SEOquake is Semrush’s free SEO Chrome extension for auditing any webpage’s technical signals, on-page SEO, linking strategy, and core Semrush metrics.

Install the extension, open any page, and click the SEOquake icon in your browser for faster, easier SEO research and analysis.

Chrome Web Store – SEOquake

The SEOquake extension provides two ways to audit:

  • The Quick View pop-up provides both page- and domain-level metrics for a fast snapshot of performance and on-page and technical issues
  • The Full Report goes deeper with 28 on-page SEO audit checks and guidance on what to do for each one

Results are color-coded throughout every report, so you can quickly diagnose pages and prioritize fixes.

SEOQuake – Color coded report

Here’s what each color tells you:

  • Green is a pass
  • Orange is a warning — worth a look, but not an emergency
  • Red is a failure — fix these issues first
  • Blue provides useful context for checks that don’t have a pass/fail answer

We’ll use these color cues to stay focused throughout the audit.

Note: To show you how SEOquake’s three-minute page audit works, I’ll be analyzing a real product page from UnderFit, a men’s undershirt brand. Want to audit along with me? Install SEOquake for free.

Triage Any Page’s SEO Issues with the Quick View Pop-Up (Minute One)

Triage is the first minute of SEOquake’s three-minute page audit.

In 60 seconds or fewer, you’ll answer this question: Is anything obviously wrong with this page?

The Quick View pop-up has five tabs:

  • Page Info: Title tag, meta description, canonical, and robots.txt status
  • Content: Heading structure, keyword density, and image alt text
  • Audit: Technical and page-level checks color-coded by severity
  • Schema: Detected structured data types and social preview information
  • Semrush: Domain-level organic search traffic, traffic cost, and backlinks, plus page-level URL backlinks

SEOQuake – Quick View – Tabs

I’ll focus on three tabs: Page Info, Audit, and Content (about 20 seconds per tab) to get a fast read on UnderFit’s product page.

Start with Page Info for Core SEO Elements

SEOquake’s Page Info tab is a great starting point because it provides a quick overview of key SEO elements.

I open the pop-up on UnderFit’s page and assess the results.

SEOQuake – Quick View – UnderFit

This is a quick scan, so I don’t overthink it.

The five things I prioritize:

  • Title tag and meta description: Are both present? Are the character counts within range?
  • Canonical: Is it pointing to the URL I want indexed?
  • Robots.txt: Does the file allow crawlers to access the page?
  • Schema.org markup: Is structured data detected and valid?
  • Link, image, and heading counts: Do the numbers fit the page type?

Then, I quickly scan for anything marked red — or anything that doesn’t line up with on-page SEO best practices.

On UnderFit’s page, SEOquake flagged the Schema.org markup in red.

The reason points to JSON-LD errors, which means search engines may not parse the structured data correctly.

SEOQuake – Quick View – UnderFit – Schema markup

For a product page, that’s a real concern.

Invalid product schema can make the page ineligible for product rich results, including price, availability, and ratings in search.

So, that’s going straight onto my list to check in the Full Report.

The rest is mostly clean aside from two elements:

  • Title tag (65 chars): Present but flagged orange — slightly outside the ideal range
  • Meta description (182 chars): Present but flagged orange — outside the ideal range

These aren’t blockers, but I note them down.

SEOQuake – Quick View – UnderFit – Title and meta tags

Scan the Audit Tab for Technical Problems

Some issues don’t surface in Page Info, so I go to the Audit tab next.

SEOquake’s Audit checks for technical SEO essentials, including:

  • Crawling & Indexation
  • Security & Protocol
  • Page Configuration

SEOQuake – Quick View – UnderFit – Audit

On the UnderFit page, everything is green or orange.

That usually means there’s no obvious technical emergency. But orange warnings still require a judgment call.

I’ll evaluate anything in orange against what I know about the site and the SEO risk.

For this page:

  • XML sitemap: Flagged as missing, so I’ll verify the sitemap exists and includes this URL
  • Meta robots: Flagged orange for review. A conflicting directive can block indexation, so it’s worth a closer look.
  • Hreflang: Shows zero tags, but I wouldn’t worry about that for a U.S.-only store

SEOQuake – Quick View – UnderFit – Hreflang

With technical issues noted, I move on to the Content tab.

Review the Content Tab for Structure and Keywords

Next, I open the Content tab to check whether the page is structured clearly.

Content structure helps users navigate the page and can make it easier for search engines and AI systems to interpret it, so this is worth double-checking.

SEOQuake – Quick View – UnderFit – Content

In the Content tab, I check:

  • Title tag and meta description: Are both present, and are the character counts within range?
  • Heading and word counts: Do the H1/H2/H3 distribution and total word count fit what I’d expect for a product page?
  • Heading Structure: Is there one clear H1, and do the headings follow a logical hierarchy?
  • Keyword Density: Do the top terms match the page topic, and do important terms appear in the title, meta description, and H1?
  • Images: How many images are missing alt text or flagged as oversized?

Seeing UnderFit’s heading structure laid out makes me pause.

The H1 appears way down the page, and the hierarchy jumps levels throughout — going from H1 straight to H4, for example.

SEOQuake – Quick View – UnderFit – Headings

That’s worth reviewing for accessibility, reader navigation, and overall content clarity.

The image summary also flags that 36 images are missing alt text.

SEOQuake – Quick View – UnderFit – Images

Since images do a lot of selling on a product page, I’ll look into this further.

Missing alt text gives search engines less image information and makes the page less accessible to shoppers.

Neither is a blocker, but I’ll check both in the next step of the audit: Diagnose.

Related reading: 11 Chrome Extensions for SEO That Will Streamline Your Workflow

Diagnose Issues with the SEOquake Full Report (Minutes 2–3)

In the first minute of the audit, I found a few issues on UnderFit’s product page that I want to investigate in SEOquake’s Full Report.

In the next two minutes, I’ll diagnose the cause of each flag. Then I’ll decide whether I can fix the issues at the page level or need to run a site-wide audit.

I click “View Full Report” at the top of SEOquake’s Quick View pop-up to open the panels.

SEOQuake – View full report

The Full Report has six panels:

  • Page Info: Basic page signals, technical snapshot, and public Semrush metrics
  • Audit: Technical SEO checks with fix guidance
  • Keyword Density: Key terms by phrase, length, density, and prominence
  • Links: Internal and external links metrics
  • Compare: URL or domain comparison
  • Semrush: Domain and page data and metrics

SEOQuake – Full report – Tabs

I’ll walk you through a few of the Full Report panels to give you a quick overview.

But when you run your own audit, go directly to the reports that match the flags you found.

Dig Deeper into Technical Issues in the Full Audit Panel

Since SEOquake identified a few technical issues, I’ll start with the Audit panel.

The full Audit covers crawlability, indexation, meta robots, mobile compliance, sitemaps, and schema.

It gives me a clearer sense of how serious each flag is and what to do about it.

SEOQuake – Full report – Audit

I use the menu to jump straight to the flags rather than scrolling through everything:

SEOQuake – Full report – Audit – Sections

For each flag, I’ll do the following:

  • Click the “i” to see what the check measures
  • Check the color-coded status to see what SEOquake found
  • Expand “Tips” for guidance on what to do next

SEOQuake – Full report – Audit – Info tips

For UnderFit, I check the XML sitemap warning first.

Under Technical Compliance, SEOquake confirms that “https://underfit.com/sitemap.xml” exists but is not clearly linked.

SEOQuake – Full report – Audit – XML sitemap

That’s definitely going on my fix list, as it could affect this page’s discoverability.

Triage also flagged UnderFit’s Schema.org markup as invalid, so I’ll check the Schema section next.

SEOquake shows green, meaning it’s detected markup on the page.

SEOQuake – Full report – Audit – Schema

I need to see if that’s valid, so I click “Test in Rich Results.”

SEOQuake – Full report – Audit – Schema – Test in Rich Results

This opens Google’s Rich Results Test, which runs the page automatically.

It detects 3 invalid items.

That confirms what triage suggested, so it goes onto my fix list, too.

After that, I work through the remaining flags the same way — including the heading hierarchy issues and missing alt text from triage — adding actionable items to my list.

Pro tip: Once you’ve worked through the flags, export the full audit as a PDF to share with a developer or client, or to keep a record before you make changes.

SEOQuake – Full report – Audit – Export as PDF

Spot Topic Coverage Gaps in the Keyword Density Panel

If a page isn’t performing the way you’d expect, Keyword Density is worth a quick look.

SEOQuake – Full report – Keyword Density

It tells you which terms appear most frequently in the page’s visible content and whether the topic coverage looks right for the page type.

For UnderFit, I start with the Keyword Cloud to get a quick read on what the page is about.

SEOQuake – Full report – Keyword Density – Keyword Cloud

The core product terms are the most prominent: undershirt, fit, fabric, v-neck, cotton. That’s exactly what I’d expect for this page type.

Side note: If you’re auditing a competitor, SEOquake’s Keyword Density report is a fast way to see which topics they’re covering that you aren’t.

Then I dig into the keyword tables for more detail.

The “Found in” column confirms those terms are showing up in the right places — “undershirt” appears in the meta description and H1, while “fit” and “underfit” both appear in the title tag.

SEOQuake – Full report – Keyword Density – 1-word phrases

Next, I scan the multi-word phrase tables for any missing or off-topic phrases.

Most reinforce the page’s topic: “the invisible undershirt,” “v-neck undershirt,” and “built to last.”

Some odd phrases, like “zoom zoom,” come from interactive elements and image labels in the HTML rather than the main product copy.

SEOQuake – Full report – Keyword Density – 2-word phrases

That’s normal — tools like SEOquake can pick up text from buttons, image labels, and other page elements, so widget text sometimes appears in the data.

Check Link Health in the Links Panel

Broken, vague, or unhelpful links can make it harder for search engines and users to navigate pages effectively.

The Links panel shows the status of all internal and external links on the page, including dofollow/nofollow, anchor text, and HTTP response codes.

SEOQuake – Full report – Links – Save as CSV

On UnderFit’s page, the report shows one broken internal link and 37 broken external links.

SEOQuake – Full report – Links – External links – Broken

So I go to the External Links table to dig deeper — they’re mostly JavaScript links.

I’ll add those to the fix list to confirm whether they’re functional or actually problematic.

Pro tip: If you’re sharing findings with a developer or content team, export the table as a CSV. It provides them with the affected URLs, anchor text, follow status, and response codes in a single file.

SEOQuake – Full report – Links – Save as CSV

Decide on a Page Fix or Site Audit Using the Semrush Panel

I’ve now worked through the triage flags and have a list of issues to address on UnderFit’s product page.

But before I fix anything, I need to decide whether each issue is specific to this page or points to a broader site problem. The Semrush panel gives me the additional context I need to make that call.

By default, it shows domain- and page-level metrics, including:

  • Semrush Rank: The domain’s organic visibility ranking
  • SE Traffic: Estimated monthly organic search traffic for the domain
  • Root Domain Backlinks: The domain’s total backlink profile
  • URL backlinks: Total backlinks to the page

SEOQuake – Full report – Semrush

Optionally, I can connect SEOquake to my Semrush account to unlock deeper data.

This includes Traffic Analytics data (engagement metrics such as bounce rate, visit duration, and channel-level traffic) and detailed backlink reports.

SEOQuake – Full report – Semrush – Connected data

Pro tip: To connect your Semrush account to SEOquake, click the gear icon in the top-right corner of the Quick View pop-up, then click “Connect to Semrush” or access the Integrations tab in Settings.

A weak page on a strong domain — like UnderFit, where the domain has solid traffic and a reasonable backlink profile — usually points to a content or page-level problem.

But some flags aren’t cleanly at the page level.

The XML sitemap warning, for example, is almost always site-wide, and JSON-LD errors might repeat across all product templates.

Those aren’t problems I can confirm or fix with SEOquake alone, so I’ll move to a full site crawl with a tool like Semrush’s Site Audit.

Diagnose Any Page in 3 Minutes with SEOquake

As you’ve seen, the SEOquake Chrome extension lets you audit any page in three minutes and walk away knowing exactly what’s working, what’s broken, and what to do next.

Install SEOquake from the Chrome Web Store and run your first audit today:

  • Triage using the Quick View pop-up to catch obvious problems
  • Diagnose the flags in the Full Report to understand what’s wrong
  • Decide whether it’s a page fix or a site-wide issue
  • Escalate to Semrush Site Audit when the problem is bigger than the page

To crawl your full site, start a free 14-day Semrush trial and get Site Audit access for 100 URLs per month.

The post How to Audit Any Page in Under 3 Minutes with SEOquake appeared first on Backlinko.

Read more at Read More