Web Design and Development San Diego

How SEO leaders can explain agentic AI to ecommerce executives

/0 Comments/in , /by
How to communicate agentic AI to ecommerce leadership without the hype

Agentic AI is increasingly appearing in leadership conversations, often accompanied by big claims and unclear expectations. For SEO leaders working with ecommerce brands, this creates a familiar challenge.

Executives hear about autonomous agents, automated purchasing, and AI-led decisions, and they want to know what this really means for growth, risk, and competitiveness.

What they don’t need is more hype. They need clear explanations, grounded thinking, and practical guidance. 

This is where SEO leaders can add real value, not by predicting the future, but by helping leadership understand what is changing, what isn’t, and how to respond without overreacting. Here’s how.

Start by explaining what ‘agentic’ actually means

A useful first step is to remove the mystery from the term itself. Agentic systems don’t replace customers, they act on behalf of customers. The intent, preferences, and constraints still come from a person.

What changes is who does the work.

Discovery, comparison, filtering, and sometimes execution are handled by software that can move faster and process more information than a human can.

When speaking to executive teams, a simple framing works best:

  • “We’re not losing customers, we’re adding a new decision-maker into the journey. That decision-maker is software acting as a proxy for the customer.” 

Once this is clear, the conversation becomes calmer and more practical, and the focus moves away from fear and toward preparation.

Your customers search everywhere. Make sure your brand shows up.

The SEO toolkit you know, plus the AI visibility data you need.

Start Free Trial
Get started with

Semrush One Logo

Keep expectations realistic and avoid the hype

Another important role for SEO leaders is to slow the conversation down. Agentic behavior will not arrive everywhere at the same time. Its impact will be uneven and gradual.

Some categories will see change earlier because their products are standardized and data is already well structured. Others will move more slowly because trust, complexity, or regulation makes automation harder.

This matters because leadership teams often fall into one of two traps:

  1. Panic, where plans are rewritten too quickly, budgets move too fast, and teams chase futures that may still be some distance away. 
  2. Dismissal, where nothing changes until performance clearly drops, and by then the response is rushed.

SEO leaders can offer a steadier view. Agentic AI accelerates trends that already exist. Personalized discovery, fewer visible clicks, and more pressure on data quality are not new problems. 

Agents simply make them more obvious. Seen this way, agentic AI becomes a reason to improve foundations, not a reason to chase novelty.

Dig deeper: Are we ready for the agentic web?

Change the conversation from rankings to eligibility

One of the most helpful shifts in executive conversations is moving away from rankings as the main outcome of SEO. In an agent-led journey, the key question isn’t “do we rank well?” but “are we eligible to be chosen at all?”

Eligibility depends on clarity, consistency, and trust. An agent needs to understand what you sell, who it is for, how much it costs, whether it is available, and how risky it is to choose you on behalf of a user. This is a strong way to connect SEO to commercial reality.

Questions worth raising include whether product information is consistent across systems, whether pricing and availability are reliable, and whether policies reduce uncertainty or create it. Framed this way, SEO becomes less about chasing traffic and more about making the business easy to select.

Explain why SEO no longer sits only in marketing

Many executives still see SEO as a marketing channel, but agentic behavior challenges that view.

Selection by an agent depends on factors that sit well beyond marketing. Data quality, technical reliability, stock accuracy, delivery performance, and payment confidence all play a role.

SEO leaders should be clear about this. This isn’t about writing more content. It’s about making sure the business is understandable, reliable, and usable by machines.

Positioned correctly, SEO becomes a connecting function that helps leadership see where gaps in systems or data could prevent the brand from being selected. This often resonates because it links SEO to risk and operational health, not just growth.

Dig deeper: How to integrate SEO into your broader marketing strategy

Be clear that discovery will change first

For most ecommerce brands, the earliest impact of agentic systems will be at the top of the funnel. Discovery becomes more conversational and more personal.

Users describe situations, needs, and constraints instead of typing short search phrases, and the agent then turns that context into actions.

This reduces the value of simply owning category head terms. If an agent knows a user’s budget, preferences, delivery expectations, and past behavior, it doesn’t behave like a first-time visitor. It behaves like a well-informed repeat customer.

This creates a reporting challenge. Some SEO work will no longer look like direct demand creation, even though it still influences outcomes. Leadership teams need to be prepared for this shift.

Get the newsletter search marketers rely on.

See terms.

Reframe consideration as filtering, not persuasion

The middle of the funnel also changes shape. Today, consideration often involves reading reviews, comparing options, and seeking reassurance.

In an agent-led journey, consideration becomes a filtering process, where the agent removes options it believes the user would reject and keeps those that fit.

This has clear implications. Generic content becomes less effective as a traffic driver because agents can generate summaries and comparisons instantly. Trust signals become structural, meaning claims need to be backed by consistent and verifiable information.

In many cases, a brand may be chosen without the user being consciously aware of it. That can be positive for conversion, but risky for long-term brand strength if recognition isn’t built elsewhere.

Dig deeper: How to align your SEO strategy with the stages of buyer intent

Set honest expectations about measurement

Executives care about measurement, and agentic AI makes this harder. As more discovery and consideration happen inside AI systems, fewer interactions leave clean attribution trails. Some impact will show up as direct traffic, and some will not be visible at all.

SEO leaders should address this early. This isn’t a failure of optimization. It reflects the limits of today’s analytics in a more mediated world.

The conversation should move toward directional signals and blended performance views, rather than precise channel attribution that no longer reflects how decisions are made.

Promote a proactive, low-risk response

The most important part of the leadership discussion is what to do next. The good news is that most sensible responses to agentic AI are low risk.

Improving product data quality, reducing inconsistencies across platforms, strengthening reliability signals, and fixing technical weaknesses all help today, regardless of how quickly agents mature.

Investing in brand demand outside search also matters. If agents handle more of the comparison work, brands that users already trust by name are more likely to be selected.

This reassures leaders that action doesn’t require dramatic change, only disciplined improvement.

See the complete picture of your search visibility.

Track, optimize, and win in Google and AI search from one platform.

Start Free Trial
Get started with

Semrush One Logo

Agentic AI changes the focus, not the fundamentals

For SEO leaders, agentic AI changes the focus of the role. The work shifts from optimizing pages to protecting eligibility, from chasing visibility to reducing ambiguity, and from reporting clicks to explaining influence.

This requires confidence, clear communication, and a willingness to challenge hype. Agentic AI makes SEO more strategic, not any less important.

Agentic AI should not be treated as an immediate threat or a guaranteed advantage. It’s a shift in how decisions are made.

For ecommerce brands, the winners will be those that stay calm, communicate clearly, and adapt their SEO thinking from driving clicks to earning selection.

That is the conversation SEO leaders should be having now.

Dig deeper: The future of search visibility: What 6 SEO leaders predict for 2026

Read more at Read More

Web Design and Development San Diego

What repeated ChatGPT runs reveal about brand visibility

/0 Comments/in , /by
What repeated ChatGPT runs reveal about brand visibility

We know AI responses are probabilistic – if you ask an AI the same question 10 times, you’ll get 10 different responses.

But how different are the responses?

That’s the question Rand Fishkin explored in some interesting research.

And it has big implications for how we should think about tracking AI visibility for brands.

In his research, he tested prompts asking for recommendations in all sorts of products and services, including everything from chef’s knives to cancer care hospitals and Volvo dealerships in Los Angeles.

Basically, he found that:

  • AIs rarely recommend the same list of brands in the same order twice.
  • For a given topic (e.g., running shoes), AIs recommend a certain handful of brands far more frequently than others.

For my research, as always, I’m focusing exclusively on B2B use cases. Plus, I’m building on Fishkin’s work by addressing these additional questions:

  • Does prompt complexity affect the consistency of AI recommendations?
  • Does the competitiveness of the category affect the consistency of recommendations?

Methodology

To explore those questions, I first designed 12 prompts:

  • Competitive vs. niche: Six of the prompts are about highly competitive B2B software categories (e.g., accounting software), and the other six are about less crowded categories (e.g., user entity behavior analytics (UEBA) software). I identified the categories using Contender’s database, which tracks how many brands ChatGPT associates with 1,775 different software categories.
  • Simple vs. nuanced prompts: Within both sets of “competitive” and “niche” prompts, half of the prompts are simple (“What’s the best accounting software?”) and the other half are nuanced prompts including a persona and use case (”For a Head of Finance focused on ensuring financial reporting accuracy and compliance, what’s the best accounting software?”)

I ran the 12 prompts 100 times, each, through the logged-out, free version of ChatGPT at chatgpt.com (i.e., not the API). I used a different IP address for each of the 1,200 interactions to simulate 1,200 different users starting new conversations.

Limitations: This research only covers responses from ChatGPT. But given the patterns in Fishkin’s results and the similar probabilistic nature of LLMs, you can probably generalize the directional (not absolute value) findings below to most/all AIs.

Your customers search everywhere. Make sure your brand shows up.

The SEO toolkit you know, plus the AI visibility data you need.

Start Free Trial
Get started with

Semrush One Logo

Findings

So what happens when 100 different people submit the same prompt to ChatGPT, asking for product recommendations?

How many ‘open slots’ in ChatGPT responses are available to brands?

On average, ChatGPT will mention 44 brands across 100 different responses. But one of the response sets included as many as 95 brands – it really depends on the category.

How many brands does ChatGPT draw from, on average?

Competitive vs. niche categories

On that note, for prompts covering competitive categories, ChatGPT mentions about twice as many brands per 100 responses compared to the responses to prompts covering “niche” categories. (This lines up with the criteria I used to select the categories I studied.)

Simple vs. nuanced prompts

On average, ChatGPT mentioned slightly fewer brands in response to nuanced prompts. But this wasn’t a consistent pattern – for any given software category, sometimes nuanced questions ended up with more brands mentioned, and sometimes simple questions did.

This was a bit surprising, since I expected more specific requests (e.g., “For a SOC analyst needing to triage security alerts from endpoints efficiently, what’s the best EDR software?”) to consistently yield a narrower set of potential solutions from ChatGPT.

I think ChatGPT might not be better at tailoring a list of solutions to a specific use case because it doesn’t have a deep understanding of most brands. (More on this data in an upcoming note.)

Return of the ’10 blue links’

In each individual response, ChatGPT will, on average, mention only 10 brands.

There’s quite a range, though – a minimum of 6 brands per response and a maximum of 15 when averaging across response sets.

How many brands per response, on average?

But a single response typically names about 10 brands regardless of category or prompt type.

The big difference is in how much the pool of brands rotates across responses – competitive categories draw from a much deeper bench, even though each individual response names a similar count.

Everything old (in SEO) truly is new again (in GEO/AEO). It reminds me of trying to get a placement in one of Google’s “10 blue links”.

Dig deeper: How to measure your AI search brand visibility and prove business impact

Get the newsletter search marketers rely on.

See terms.

How consistent are ChatGPT’s brand recommendations?

When you ask ChatGPT for a B2B software recommendation 100 different times, there are only ~5 brands, on average, that it’ll mention 80%+ of the time.

To put it in context, that’s just 11% of all the 44 brands it’ll mention at all across those 100 responses.

ChatGPT knows ~44 brands in your category

So it’s quite competitive to become one of the brands ChatGPT consistently mentions whenever someone asks for recommendations in your category.

As you’d expect, these “dominant” brands tend to be big, established brands with strong recognition. For example, the dominant brands in the accounting software category are QuickBooks, Xero, Wave, FreshBooks, Zoho, and Sage.

If you’re not a big brand, you’re better off being in a niche category:

It's easier to get good AI visibility in niche categories

When you operate in a niche category, not only are you literally competing with fewer companies, but there are also more “open slots” available to you to become a dominant brand in ChatGPT’s responses.

In niche categories, 21% of all the brands ChatGPT mentions are dominant brands, getting mentioned 80%+ of the time.

Compare this to just 7% of all brands being dominant in competitive categories, where the majority of brands (72%) are languishing in the long tail, getting mentioned less than 20% of the time.

The responses to nuanced prompts are harded to dominate

A nuanced prompt doesn’t dramatically change the long tail of little-seen brands (with <20% visibility), but it does change the “winner’s circle.” Adding persona context to a prompt makes it a bit more difficult to reach the dominant tier – you can see the steeper “cliff” a brand has to climb in the “nuanced prompts” graph above.

This makes intuitive sense: when someone asks “best accounting software for a Head of Finance,” ChatGPT has a more specific answer in mind and commits a bit more strongly to fewer top picks.

Still, it’s worth noting that the overall pool doesn’t shrink much – ChatGPT mentions ~42 brands in 100 responses to nuanced prompts, just a handful fewer than the ~46 mentioned in response to simple prompts. If nuanced prompts make the winner’s circle a bit more exclusive, why don’t they also narrow the total field?

Partly, it could be that the “nuanced” questions we fed it weren’t meaningfully more narrow and specific than what was implied in the simple questions we asked.

But, based on other data I’m seeing, I think this is partly about ChatGPT not knowing enough about most brands to be more selective. I’ll share more on this in an upcoming note.

Dig deeper: 7 hard truths about measuring AI visibility and GEO performance

What does this mean for B2B marketers?

If you’re not a dominant brand, pick your battles – niche down

It’s never been more important to differentiate. 21% of mentioned brands reach dominant status in niche categories vs. 7% in competitive ones.

Without time and a lot of money for brand marketing, an upstart tech company isn’t going to become a dominant brand in a broad, established category like accounting software.

But the field is less competitive when you lean into your unique, differentiating strengths. ChatGPT is more likely to treat you like a dominant brand if you work to make your product known as “the best accounting software for commercial real estate companies in North America.”

Most AI visibility tracking tools are grossly misleading

Given the inconsistency of ChatGPT’s recommendations, a single spot-check for any given prompt is nearly meaningless. Unfortunately, checking each prompt just once per time period is exactly what most AI visibility tracking tools do.

If you want anything approaching a statistically-significant visibility score for any given prompt, you need to run the prompt at least dozens of times, even 100+ times, depending on how precise you need the data to be.

But that’s obviously not practical for most people, so my suggestion is: For the key, bottom-of-funnel prompts you’re tracking, run them each ~5 times whenever you pull data.

That’ll at least give you a reasonable sense of whether your brand tends to show up most of the time, some of the time, or never.

Your goal should be to have a confident sense of whether your brand is in the little-seen long tail, the visible middle, or the dominant top-tier for any given prompt. Whether you use my tiers of ‘under 20%’, ‘20–80%’, and ‘80%+’, or your own thresholds, this is the approach that follows the data and common sense.

See the complete picture of your search visibility.

Track, optimize, and win in Google and AI search from one platform.

Start Free Trial
Get started with

Semrush One Logo

What’s next?

In future newsletters and LinkedIn posts, I’m going to build on these findings with new research:

  • How does ChatGPT talk about the brands it consistently recommends? Is it indicative of how much ChatGPT “knows” about brands?
  • Do different prompts with the same search intent tend to produce the same set of recommendations?
  • How consistent is “rank” in the responses? Do dominant brands tend to get mentioned first?

This article was originally published on Visible on beehiiv (as Most AI visibility tracking is misleading (here’s my new data)) and is republished with permission.

Read more at Read More

Web Design and Development San Diego

Reddit says 80 million people now use its search weekly

/0 Comments/in , /by
Reddit search

Eighty million people use Reddit search every week, Reddit said on its Q4 2025 earnings call last week. The increase followed a major change: Reddit merged its core search with its AI-powered Reddit Answers and began positioning the platform as a place where users can start — and finish — their searches.

  • Executives framed the move as a response to changing behavior. People are increasingly researching products and making decisions by asking questions within communities rather than relying solely on traditional search engines.
  • Reddit is betting it can keep more of that intent on-platform, rather than acting mainly as a source of links for elsewhere.

Why we care. Reddit is becoming a place where people start — and complete — their searches without ever touching Google. For brands, that means visibility on Reddit now matters as much as ranking in traditional and AI search for many buying decisions.

Reddit’s search ambitions. CEO Steve Huffman said Reddit made “significant progress” in Q4 by unifying keyword search with Reddit Answers, its AI-driven Q&A experience. Users can now move between standard search results and AI answers in a single interface, with Answers also appearing directly inside search results.

  • “Reddit is already where people go to find things,” Huffman said, adding the company wants to become an “end-to-end search destination.”
  • More than 80 million people searched Reddit weekly in Q4, up from 60 million a year earlier, as users increasingly come to the platform to research topics — not just scroll feeds or click through from Google.

Reddit Answers is growing. Reddit Answers is driving much of that growth. Huffman said Answers queries jumped from about 1 million a year ago to 15 million in Q4, while overall search usage rose sharply in parallel.

  • He said Answers performs best for open-ended questions—what to buy, watch, or try—where people want multiple perspectives instead of a single factual answer. Those queries align naturally with Reddit’s community-driven discussions.
  • Reddit is also expanding Answers beyond text. Huffman said the company is piloting “dynamic agentic search results” that include media formats, signaling a more interactive and immersive search experience ahead.

Search is a ‘big one’ for Reddit. Huffman said the company is testing new app layouts that give search prominent placement, including versions with a large, always-visible search bar at the top of the home screen.

  • COO Jennifer Wong said search and Answers represent a major opportunity, even though monetization remains early on some surfaces.
  • Wong described Reddit search behavior as “incremental and additive” to existing engagement and often tied to high-intent moments, such as researching purchases or comparing options.

AI answers make Reddit more important. Huffman also linked Reddit’s search push to its partnerships with Google and OpenAI. He said Reddit content is now the most-cited source in AI-generated answers, highlighting the platform’s growing influence on how people find information.

  • Reddit sees AI summaries as an opportunity — to move users from AI answers into Reddit communities, where they can read discussions, ask follow-up questions, and participate.
  • If someone asks “what the best speaker is,” he said, Reddit wants users to discover not just a summary, but the community where real people are actively debating the topic.

Reddit earnings. Reddit Reports Fourth Quarter and Full Year 2025 Results; Announces $1 Billion Share Repurchase Program

Read more at Read More

Web Design and Development San Diego

OpenAI starts testing ChatGPT ads

/0 Comments/in , /by

OpenAI confirmed today that it’s rolling out its first live test of ads in ChatGPT, showing sponsored messages directly inside the app for select users.

The details. The ads will appear in a clearly labeled section beneath the chat interface, not inside responses, keeping them visually separate from ChatGPT’s answers.

  • OpenAI will show ads to logged-in users on the free tier and its lower-cost Go subscription.
  • Advertisers won’t see user conversations or influence ChatGPT’s responses, even though ads will be tailored based on what OpenAI believes will be helpful to each user, the company said.

How ads are selected. During the test, OpenAI matches ads to conversation topics, past chats, and prior ad interactions.

  • For example: A user researching recipes might see ads for meal kits or grocery delivery. If multiple advertisers qualify, OpenAI shows the most relevant option first.

User controls. Users get granular controls over the experience. They can dismiss ads, view and delete separate ad history and interest data, and toggle personalization on or off.

  • Turning personalization off limits ads to the current chat.
  • Free users can also opt out of ads in exchange for fewer daily messages or upgrade to a paid plan.

Why we care. ChatGPT is one of the world’s largest consumer AI platforms. Even a limited ad rollout could mark a major shift in how conversational AI gets monetized — and how brands reach users.

Bottom line. OpenAI is officially moving into ads inside ChatGPT, testing how sponsored content can coexist with conversational AI at massive scale.

OpenAI’s announcement.Testing ads in ChatGPT (OpenAI)

Read more at Read More

Web Design and Development San Diego

Google AI Mode doesn’t favor above-the-fold content: Study

/0 Comments/in , /by
AI Mode depth doesn't matter

Google’s AI Mode isn’t more likely to cite content that appears “above the fold,” according to a study from SALT.agency, a technical SEO and content agency.

  • After analyzing more than 2,000 URLs cited in AI Mode responses, researchers found no correlation between how high text appears on a page and whether Google’s AI selects it for citation.

Pixel depth doesn’t matter. AI Mode cited text from across entire pages, including content buried thousands of pixels down.

  • Citation depth showed no meaningful relationship to visibility.
  • Average depth varied by vertical, from about 2,400 pixels in travel to 4,600 pixels in SaaS, with many citations far below the traditional “above the fold” area.

Page layout affects depth, not visibility. Templates and design choices influenced how far down the cited text appeared, but not whether it was cited.

  • Pages with large hero images or narrative layouts pushed cited text deeper, while simpler blog or FAQ-style pages surfaced citations earlier.
  • No layout type showed a visibility advantage in AI Mode.

Descriptive subheadings matter. One consistent pattern emerged: AI Mode frequently highlighted a subheading and the sentence that followed it.

  • This suggests Google uses heading structures to navigate content, then samples opening lines to assess relevance, behavior consistent with long-standing search practices, according to SALT.

What Google is likely doing. SALT believes AI Mode relies on the same fragment indexing technology Google has used for years. Pages are broken into sections, and the most relevant fragment is retrieved regardless of where it appears on the page.

What they’re saying. While the study examined only one structural factor and one AI model, the takeaway is clear: there’s no magic formula for AI Mode visibility. Dan Taylor, partner and head of innovation (organic and AI) at SALT.agency, said:

  • “Our study confirms that there is no magic template or formula for increased visibility in AI Mode responses – and that AI Mode is not more likely to cite text from ‘above the fold.’ Instead, the best approach mirrors what’s worked in search for years: create well-structured, authoritative content that genuinely addresses the needs of your ideal customers.
  • “…the data clearly debunks the idea that where the information sits within a page has an impact on whether it will be cited.”

Why we care. The findings challenge the idea that AI-specific templates or rigid page structures drive better AI Mode visibility. Chasing “AI-optimized” layouts may distract from work that actually matters.

About the research. SALT analyzed 2,318 unique URLs cited in AI Mode responses for high-value queries across travel, ecommerce, and SaaS. Using a Chrome bookmarklet and a 1920×1080 viewport, researchers recorded the vertical pixel position of the first highlighted character in each AI-cited fragment. They also cataloged layouts and elements, such as hero sections, FAQs, accordions, and tables of contents.

The study. Research: Does Structuring Your Content Improve the Chances of AI Mode Surfacing?

Read more at Read More

Web Design and Development San Diego

Google’s February 2026 Discover Core Update

/0 Comments/in , /by

Today we’ve released
the February 2026 Discover core update. This is a broad update to our systems that surface
articles in Discover. Our testing shows that people find the Discover experience more useful and
worthwhile with this update.

Read more at Read More

Microsoft launches Publisher Content Marketplace for AI licensing

/0 Comments/in , /by
The future of remarketing? Microsoft bets on impressions, not clicks

Microsoft Advertising today launched the Publisher Content Marketplace (PCM), a system that lets publishers license premium content to AI products and get paid based on how that content is used.

How it works. PCM creates a direct value exchange. Publishers set licensing and usage terms, while AI builders discover and license content for specific grounding scenarios. The marketplace also includes usage-based reporting, giving publishers visibility into how their content performs and where it creates the most value.

Designed to scale. PCM is designed to avoid one-off licensing deals between individual publishers and AI providers. Participation is voluntary, ownership remains with publishers, and editorial independence stays intact. The marketplace supports everyone from global publishers to smaller, specialized outlets.

Why we care. As AI systems shift from answering questions to making decisions, content quality matters more than ever. As agents increasingly guide purchases, finance, and healthcare choices, ads and sponsored messages will sit alongside — or draw from — premium content rather than generic web signals. That raises the bar for credibility and points to a future where brand alignment with trusted publishers and AI ecosystems directly impacts performance.

Early traction. Microsoft Advertising co-designed PCM with major U.S. publishers, including Business Insider, Condé Nast, Hearst, The Associated Press, USA TODAY, and Vox Media. Early pilots grounded Microsoft Copilot responses in licensed content, with Yahoo among the first demand partners now onboarding.

What’s next. Microsoft plans to expand the pilot to more publishers and AI builders that share a core belief: as the AI web evolves, high-quality content should be respected, governed, and paid for.

The big picture. In an agentic web, AI tools increasingly summarize, reason, and recommend through conversation. Whether the topic is medical safety, financial eligibility, or a major purchase, outcomes depend on access to trusted, authoritative sources — many of which sit behind paywalls or in proprietary archives.

The tension. The traditional web bargain was simple: publishers shared content, and platforms sent traffic back. That model breaks down when AI delivers answers directly, cutting clicks while still depending on premium content to perform well.

Bottom line. If AI is going to make better decisions, it needs better inputs — and PCM is Microsoft’s bet that a sustainable content economy can power the next phase of the agentic web.

Microsoft’s announcement. Building Toward a Sustainable Content Economy for the Agentic Web

Read more at Read More

Web Design and Development San Diego

Inspiring examples of responsible and realistic vibe coding for SEO

/0 Comments/in , /by

Vibe coding is a new way to create software using AI tools such as ChatGPT, Cursor, Replit, and Gemini. It works by describing to the tool what you want in plain language and receiving written code in return. You can then simply paste the code into an environment (such as Google Colab), run it, and test the results, all without ever actually programming a single line of code.

Collins Dictionary named “vibe coding” word of the year in 2025, defining it as “the use of artificial intelligence prompted by natural language to write computer code.”

In this guide, you’ll understand how to start vibe coding, learn its limitations and risks, and see examples of great tools created by SEOs to inspire you to vibe code your own projects.

Vibe coding variations

While “vibe coding” is used as an umbrella term, there are subsets of coding with support or AI, including the following:

Type Description Tools
AI-assisted coding  AI helps write, refactor, explain, or debug code. Used by actual developers or engineers to support their complex work. GitHub Copilot, Cursor, Claude, Google AI Studio
Vibe coding Platforms that handle everything except the prompt/idea. AI does most of the work. ChatGPT, Replit, Gemini, Google AI Studio
No-code platforms Platforms that handle everything you ask (“drag and drop” visual updates while the code happens in the background). They tend to use AI but existed long before AI became mainstream. Notion, Zapier, Wix

We’ll focus exclusively on vibe coding in this guide. 

With vibe coding, while there’s a bit of manual work to be done, the barrier is still low — you basically need a ChatGPT account (free or paid) and access to a Google account (free). Depending on your use case, you might also need access to APIs or SEO tools subscriptions such as Semrush or Screaming Frog.

Your customers search everywhere. Make sure your brand shows up.

The SEO toolkit you know, plus the AI visibility data you need.

Start Free Trial
Get started with

Semrush One Logo

To set expectations, by the end of this guide, you’ll know how to run a small program on the cloud. If you expect to build a SaaS or software to sell, AI-assisted coding is a more reasonable option to take, which will involve costs and deeper coding knowledge.

Vibe coding use cases

Vibe coding is great when you’re trying to find outcomes for specific buckets of data, such as finding related links, adding pre-selected tags to articles, or doing something fun where the outcome doesn’t need to be exact.

For example, I’ve built an app to create a daily drawing for my daughter. I type a phrase about something that she told me about her day (e.g., “I had carrot cake at daycare”). The app has some examples of drawing styles I like and some pictures of her. The outputs (drawings) are the final work as they come from AI.

When I ask for specific changes, however, the program tends to worsen and redraw things I didn’t ask for. I once asked to remove a mustache and it recolored the image instead. 

If my daughter were a client who’d scrutinize the output and require very specific changes, I’d need someone who knows Photoshop or similar tools to make specific improvements. In this case, though, the results are good enough. 

Building commercial applications solely on vibe coding may require a company to hire vibe coding cleaners. However, for a demo, MVP (minimum viable product), or internal applications, vibe coding can be a useful, effective shortcut. 

How to create your SEO tools with vibe coding

Using vibe coding to create your own SEO tools require three steps:

  1. Write a prompt describing your code
  2. Paste the code into a tool such as Google Colab
  3. Run the code and analyze the results

Here’s a prompt example for a tool I built to map related links at scale. After crawling a website using Screaming Frog and extracting vector embeddings (using the crawler’s integration with OpenAI), I vibe coded a tool that would compare the topical distance between the vectors in each URL.

This is exactly what I wrote on ChatGPT:

I need a Google Colab code that will use OpenAI to:

Check the vector embeddings existing in column C. Use cosine similarity to match with two suggestions from each locale (locale identified in Column A). 

The goal is to find which pages from each locale are the most similar to each other, so we can add hreflang between these pages.

I’ll upload a CSV with these columns and expect a CSV in return with the answers.

Then I pasted the code that ChatGPT created on Google Colab, a free Jupyter Notebook environment that allows users to write and execute Python code in a web browser. It’s important to run your program by clicking on “Run all” in Google Colab to test if the output does what you expected.

This is how the process works on paper. Like everything in AI, it may look perfect, but it’s not always functioning exactly how you want it. 

You’ll likely encounter issues along the way — luckily, they’re simple to troubleshoot.

First, be explicit about the platform you’re using in your prompt. If it’s Google Colab, say the code is for Google Colab. 

You might still end up with code that requires packages that aren’t installed. In this case, just paste the error into ChatGPT and it’ll likely regenerate the code or find an alternative. You don’t even need to know what the package is, just show the error and use the new code. Alternatively, you can ask Gemini directly in your Google Colab to fix the issue and update your code directly.

AI tends to be very confident about anything and could return completely made-up outputs. One time I forgot to say the source data would come from a CSV file, so it simply created fake URLs, traffic, and graphs. Always check and recheck the output because “it looks good” can sometimes be wrong.

If you’re connecting to an API, especially a paid API (e.g., from Semrush, OpenAI, Google Cloud, or other tools), you’ll need to request your own API key and keep in mind usage costs. 

Should you want an even lower execution barrier than Google Colab, you can try using Replit. 

Simply prompt your request and the software will create the code, design, and allow testing all on the same screen. This means a lower chance of coding errors, no copy and paste, and a URL you can share right away with anyone to see your project built with a nice design. (You should still check for poor outputs and iterate with prompts until your final app is built.)

Keep in mind that while Google Colab is free (you’ll only spend if you use API keys), Replit charges a monthly subscription and per-usage fee on APIs. So the more you use an app, the more expensive it gets.

Inspiring examples of SEO vibe-coded tools

While Google Colab is the most basic (and easy) way to vibe code a small program, some SEOs are taking vibe coding even further by creating programs that are turned into Chrome extensions, Google Sheets automation, and even browser games.

The goal behind highlighting these tools is not only to showcase great work by the community, but also to inspire, build, and adapt to your specific needs. Do you wish any of these tools had different features? Perhaps you can build them for yourself — or for the world.

GBP Reviews Sentiment Analyzer (Celeste Gonzalez)

After vibe coding some SEO tools on Google Colab, Celeste Gonzalez, Director of SEO Testing at RicketyRoo Inc, took her vibing skills a step further and created a Chrome extension. “I realized that I don’t need to build something big, just something useful,” she explained.

Her browser extension, the GBP Reviews Sentiment Analyzer, summarizes sentiment analysis for reviews over the last 30 days and review velocity. It also allows the information to be exported into a CSV. The extension works on Google Maps and Google Business Profile pages.

Instead of ChatGPT, Celeste used a combination of Claude (to create high-quality prompts) and Cursor (to paste the created prompts and generate the code).

AI tools used: Claude (Sunner 4.5 model) and Cursor 

APIs used: Google Business Profile API (free)

Platform hosting: Chrome Extension

Knowledge Panel Tracker (Gus Pelogia)

I became obsessed with the Knowledge Graph in 2022, when I learned how to create and manage my own knowledge panel. Since then, I found out that Google has a Knowledge Graph Search API that allows you to check the confidence score for any entity.

This vibe-coded tool checks the score for your entities daily (or at any frequency you want) and returns it in a sheet. You can track multiple entities at once and just add new ones to the list at any time.

The Knowledge Panel Tracker runs completely on Google Sheets, and the Knowledge Graph Search API is free to use. This guide shows how to create and run it in your own Google account, or you can see the spreadsheet here and just update the API key under Extensions > App Scripts. 

AI models used: ChatGPT 5.1

APIs used: Google Knowledge Graph API (free)

Platform hosting: Google Sheets

Inbox Hero Game (Vince Nero)

How about vibe coding a link building asset? That’s what Vince Nero from BuzzStream did when creating the Inbox Hero Game. It requires you to use your keyboard to accept or reject a pitch within seconds. The game is over if you accept too many bad pitches.

Inbox Hero Game is certainly more complex than running a piece of code on Google Colab, and it took Vince about 20 hours to build it all from scratch. “I learned you have to build things in pieces. Design the guy first, then the backgrounds, then one aspect of the game mechanics, etc.,” he said.

The game was coded in HTML, CSS, and JavaScript. “I uploaded the files to GitHub to make it work. ChatGPT walked me through everything,” Vince explained.

According to him, the longer the prompt continued, the less effective ChatGPT became, “to the point where [he’d] have to restart in a new chat.” 

This issue was one of the hardest and most frustrating parts of creating the game. Vince would add a new feature (e.g., score), and ChatGPT would “guarantee” it found the error, update the file, but still return with the same error. 

See the complete picture of your search visibility.

Track, optimize, and win in Google and AI search from one platform.

Start Free Trial
Get started with

Semrush One Logo

In the end, Inbox Hero Game is a fun game that demonstrates it’s possible to create a simple game without coding knowledge, yet taking steps to perfect it would be more feasible with a developer.

AI models used: ChatGPT

APIs used: None

Platform hosting: Webpage

Vibe coding with intent

Vibe coding won’t replace developers, and it shouldn’t. But as these examples show, it can responsibly unlock new ways for SEOs to prototype ideas, automate repetitive tasks, and explore creative experiments without heavy technical lift. 

The key is realism: Use vibe coding where precision isn’t mission-critical, validate outputs carefully, and understand when a project has outgrown “good enough” and needs additional resources and human intervention.

When approached thoughtfully, vibe coding becomes less about shipping perfect software and more about expanding what’s possible — faster testing, sharper insights, and more room for experimentation. Whether you’re building an internal tool, a proof of concept, or a fun SEO side project, the best results come from pairing curiosity with restraint.

Read more at Read More

Web Design and Development San Diego

LinkedIn: AI-powered search cut traffic by up to 60%

/0 Comments/in , /by
AEO playbook

AI-powered search gutted LinkedIn’s B2B awareness traffic. Across a subset of topics, non-brand organic visits fell by as much as 60% even while rankings stayed stable, the company said.

  • LinkedIn is moving past the old “search, click, website” model and adopting a new framework: “Be seen, be mentioned, be considered, be chosen.”

By the numbers. In a new article, LinkedIn said its B2B organic growth team started researching Google’s Search Generative Experience (SGE) in early 2024. By early 2025, when SGE evolved into AI Overviews, the impact became significant.

  • Non-brand, awareness-driven traffic declined by up to 60% across a subset of B2B topics.
  • Rankings stayed stable, but click-through rates fell (by an undisclosed amount).

Yes, but. LinkedIn’s “new learnings” are more like a rehash of established SEO/AEO best practices. Here’s what LinkedIn’s content-level guidance consists of:

  • Use strong headings and a clear information hierarchy.
  • Improve semantic structure and content accessibility.
  • Publish authoritative, fresh content written by experts.
  • Move fast, because early movers get an edge.

Why we care. These tactics should all sound familiar. These are technical SEO and content-quality fundamentals. LinkedIn’s article offers little new in terms of tactics. It’s just updated packaging for modern SEO/AEO and AI visibility.

Dig deeper. How to optimize for AI search: 12 proven LLM visibility tactics

Measurement is broken. LinkedIn said its big challenge is the “dark” funnel. It can’t quantify how visibility in LLM answers impacts the bottom line, especially when discovery happens without a click.

  • LinkedIn’s B2B marketing websites saw triple-digit growth in LLM-driven traffic and that it can track conversion from those visits.
    • Yes, but: Many websites are also seeing triple-digit (or more) growth in LLM-driven traffic. Because it’s an emerging channel. That said, this is still a tiny amount of overall traffic right now (1% or less for most sites).

What LinkedIn is doing. LinkedIn created an AI Search Taskforce spanning SEO, PR, editorial, product marketing, product, paid media, social, and brand. Key actions included:

  • Correcting misinformation that showed up in AI responses.
  • Publishing new owned content optimized for generative visibility.
  • Testing LinkedIn (social) content to validate its strength in AI discovery.

Is it working? LinkedIn said early tests produced a meaningful lift in visibility and citations, especially from owned content. At least one external datapoint (Semrush, Nov. 10, 2025) suggested that LinkedIn has a structural advantage in AI search:

  • Google AI Mode cited LinkedIn in roughly 15% of responses.
  • LinkedIn was the #2 most-cited domain in that dataset, behind YouTube.

Incomplete story. LinkedIn’s article is an interesting read, but it’s light on specifics. Missing details include:

  • The exact topic set behind the “up to 60%” decline.
  • Exactly how much click-through rates “softened.”
  • Sample size and timeframe.
  • How “industry-wide” comparisons were calculated.
  • What tests were run, what moved citation share, and by how much.

Bottom line. LinkedIn is right that visibility is the new currency. However, it hasn’t shown enough detail to prove its new playbook is meaningfully different from doing some SEO (yes, SEO) fundamentals.

LinkedIn’s article. How LinkedIn Marketing Is Adapting to AI-Led Discovery

Read more at Read More

Web Design and Development San Diego

Are we ready for the agentic web?

/0 Comments/in , /by
Are we ready for the agentic web?

Innovations are coming at marketers and consumers faster than before, raising the question: Are we actually ready for the agentic web?

To answer that question, it’s important to unpack a few supporting ones:

  • What’s the agentic web?
  • How can the agentic web be used?
  • What are the pros and cons of the agentic web?

It’s important to note that this article isn’t a mandate for AI skeptics to abandon the rational questions they have about the agentic web. 

Nor is it intended to place any judgment on how you, as a consumer or professional, engage with the agentic web.

LinkedIn poll on Copilot Checkout

With thoughts and feelings so divided on the agentic web, this article aims to provide clear insight into how to think about it in earnest, without the branding or marketing fluff.

Disclosure: I am a Microsoft employee and believe in the path Microsoft’s taking with the agentic web. However, this article will attempt to be as platform-agnostic as possible.

What’s the agentic web? 

The agentic web refers to sophisticated tools, or agents, trained on our preferences that act with our consent to accomplish time-consuming tasks.

In simple terms, when I use one-click checkout, I allow my saved payment information to be passed to the merchant’s accounts receivable systems. 

Neither the merchant nor I must write down all the details or be involved beyond consenting to send and receive payment.

For fun, I put this question to four different AI models, and the responses are telling: 

  • Copilot: “The agentic web is an emerging layer of the internet where AI agents can reason and take actions on behalf of people or organizations, turning expressed intent into outcomes while preserving user choice, transparency, and control.” 
  • Gemini: “The agentic web is an evolution of the internet where autonomous AI agents – rather than just human users – interact with websites, APIs, and each other to perform complex tasks and make decisions on our behalf. “
  • Perplexity: “The agentic web is an emerging phase of the internet where autonomous AI agents act as the primary interface, navigating, interpreting, and taking actions across websites and services on behalf of users with minimal human involvement.” 
  • Claude: “The agentic web is an evolution of the internet where AI agents autonomously navigate, interact with, and complete tasks across websites and services on behalf of users.” 

I begin with how different models answer the question because it’s important to understand that each one is trained on different information, and outcomes will inevitably vary.

It’s worth noting that with the same prompt, defining the agentic web in one sentence, three out of four models focus on diminishing the human role in navigating the web, while one makes a point to emphasize the significance of human involvement, preserving user choice, transparency, and control.

Two out of four refer to the agentic web as a layer or phase rather than an outright evolution of the web. 

This is likely where the sentiment divide on the agentic web stems from.

Some see it as a consent-driven layer designed to make life easier, while others see it as a behemoth that consumes content, critical thinking, and choice.

It’s noteworthy that one model, Gemini, calls out APIs as a means of communication in the agentic web. APIs are essentially libraries of information that can be referenced, or called, based on the task you are attempting to accomplish. 

This matters because APIs will become increasingly relevant in the agentic web, as saved preferences must be organized in ways that are easily understood and acted upon.

Defining the agentic web requires spending some time digging into two important protocols – ACP and UCP.

Dig deeper: AI agents in SEO: What you need to know

Agentic Commerce Protocol: Optimized for action inside conversational AI 

The Agentic Commerce Protocol, or ACP, is designed around a specific moment: when a user has already expressed intent and wants the AI to act.

The core idea behind ACP is simple. If a user tells an AI assistant to buy something, the assistant should be able to do so safely, transparently, and without forcing the user to leave the conversation to complete the transaction.

ACP enables this by standardizing how an AI agent can:

  • Access merchant product data.
  • Confirm availability and price.
  • Initiate checkout using delegated, revocable payment authorization.

The experience is intentionally streamlined. The user stays in the conversation. The AI handles the mechanics. The merchant still fulfills the order.

This approach is tightly aligned with conversational AI platforms, particularly environments where users are already asking questions, refining preferences, and making decisions in real time. It prioritizes speed, clarity, and minimal friction.

Universal Commerce Protocol: Built for discovery, comparison, and lifecycle commerce 

The Universal Commerce Protocol, or UCP, takes a broader view of agentic commerce.

Rather than focusing solely on checkout, UCP is designed to support the entire shopping journey on the agentic web, from discovery through post-purchase interactions. It provides a common language that allows AI agents to interact with commerce systems across different platforms, surfaces, and payment providers. 

That includes: 

  • Product discovery and comparison.
  • Cart creation and updates.
  • Checkout and payment handling.
  • Order tracking and support workflows.

UCP is designed with scale and interoperability in mind. It assumes users will encounter agentic shopping experiences in many places, not just within a single assistant, and that merchants will want to participate without locking themselves into a single AI platform.

It’s tempting to frame ACP and UCP as competing solutions. In practice, they address different moments of the same user journey.

ACP is typically strongest when intent is explicit and the user wants something done now. UCP is generally strongest when intent is still forming and discovery, comparison, and context matter.

So what’s the agentic web? Is it an army of autonomous bots acting on past preferences to shape future needs? Is it the web as we know it, with fewer steps driven by consent-based signals? Or is it something else entirely?

The frustrating answer is that the agentic web is still being defined by human behavior, so there’s no clear answer yet. However, we have the power to determine what form the agentic web takes. To better understand how to participate, we now move to how the agentic web can be used, along with the pros and cons.

Dig deeper: The Great Decoupling of search and the birth of the agentic web

How can the agentic web be used? 

Working from the common theme across all definitions, autonomous action, we can move to applications.

Elmer Boutin has written a thoughtful technical view on how schema will impact agentic web compatibility. Benjamin Wenner has explored how PPC management might evolve in a fully agentic web. Both are worth reading.

Here, I want to focus on consumer-facing applications of the agentic web and how to think about them in relation to the tasks you already perform today.

Here are five applications of the agentic web that are live today or in active development.

1. Intent-driven commerce  

A user states a goal, such as “Find me the best running shoes under $150,” and an agent handles discovery, comparison, and checkout without requiring the user to manually browse multiple sites. 

How it works 

Rather than returning a list of links, the agent interprets user intent, including budget, category, and preferences. 

It pulls structured product information from participating merchants, applies reasoning logic to compare options, and moves toward checkout only after explicit user confirmation. 

The agent operates on approved product data and defined rules, with clear handoffs that keep the user in control. 

Implications for consumers and professionals 

Reducing decision fatigue without removing choice is a clear benefit for consumers. For brands, this turns discovery into high-intent engagement rather than anonymous clicks with unclear attribution. 

Strategically, it shifts competition away from who shouts the loudest toward who provides the clearest and most trusted product signals to agents. These agents can act as trusted guides, offering consumers third-party verification that a merchant is as reliable as it claims to be.

2. Brand-owned AI assistants 

A brand deploys its own AI agent to answer questions, recommend products, and support customers using the brand’s data, tone, and business rules.

How it works 

The agent uses first-party information, such as product catalogs, policies, and FAQs. 

Guardrails define what it can say or do, preventing inferences that could lead to hallucinations. 

Responses are generated by retrieving and reasoning over approved context within the prompt.

Implications for consumers and professionals 

Customers get faster and more consistent responses. Brands retain voice, accountability, and ownership of the experience. 

Strategically, this allows companies to participate in the agentic web without ceding their identity to a platform or intermediary. It also enables participation in global commerce without relying on native speakers to verify language.

3. Autonomous task completion 

Users delegate outcomes rather than steps, such as “Prepare a weekly performance summary” or “Reorder inventory when stock is low.” 

How it works 

The agent breaks the goal into subtasks, determines which systems or tools are needed, and executes actions sequentially. It pauses when permissions or human approvals are required. 

These can be provided in bulk upfront or step by step. How this works ultimately depends on how the agent is built. 

Implications for consumers and marketers 

We’re used to treating AI like interns, relying on micromanaged task lists and detailed prompts. As agents become more sophisticated, it becomes possible to treat them more like senior employees, oriented around outcomes and process improvement. 

That makes it reasonable to ask an agent to identify action items in email or send templates in your voice when active engagement isn’t required. Human choice comes down to how much you delegate to agents versus how much you ask them to assist.

Dig deeper: The future of search visibility: What 6 SEO leaders predict for 2026

Get the newsletter search marketers rely on.

See terms.

4. Agent-to-agent coordination and negotiation 

Agents communicate with other agents on behalf of people or organizations, such as a buyer agent comparing offers with multiple seller agents. 

How it works 

Agents exchange structured information, including pricing, availability, and constraints. 

They apply predefined rules, such as budgets or policies, and surface recommended outcomes for human approval. 

Implications for consumers and marketers 

Consumers may see faster and more transparent comparisons without needing to manually negotiate or cross-check options. 

For professionals, this introduces new efficiencies in areas like procurement, media buying, or logistics, where structured negotiation can occur at scale while humans retain oversight.

5. Continuous optimization over time 

Agents don’t just act once. They improve as they observe outcomes.

How it works 

After each action, the agent evaluates what happened, such as engagement, conversion, or satisfaction. It updates its internal weighting and applies those learnings to future decisions.

Why people should care 

Consumers experience increasingly relevant interactions over time without repeatedly restating preferences. 

Professionals gain systems that improve continuously, shifting optimization from one-off efforts to long-term, adaptive performance. 

What are the pros and cons of the agentic web? 

Life is a series of choices, and leaning into or away from the agentic web comes with clear pros and cons.

Pros of leaning into the agentic web 

The strongest argument for leaning into the agentic web is behavioral. People have already been trained to prioritize convenience over process. 

Saved payment methods, password managers, autofill, and one-click checkout normalized the idea that software can complete tasks on your behalf once trust is established.

Agentic experiences follow the same trajectory. Rather than requiring users to manually navigate systems, they interpret intent and reduce the number of steps needed to reach an outcome. 

Cons of leaning into the agentic web 

Many brands will need to rethink how their content, data, and experiences are structured so they can be interpreted by automated systems and humans. What works for visual scanning or brand storytelling doesn’t always map cleanly to machine-readable signals.

There’s also a legitimate risk of overoptimization. Designing primarily for AI ingestion can unintentionally degrade human usability or accessibility if not handled carefully. 

Dig deeper: The enterprise blueprint for winning visibility in AI search

Pros of leaning away from the agentic web 

Choosing to lean away from the agentic web can offer clarity of stance. There’s a visible segment of users skeptical of AI-mediated experiences, whether due to privacy concerns, automation fatigue, or a loss of human control. 

Aligning with that perspective can strengthen trust with audiences who value deliberate, hands-on interaction.

Cons of leaning away from the agentic web 

If agentic interfaces become a primary way people discover information, compare options, or complete tasks, opting out entirely may limit visibility or participation. 

The longer an organization waits to adapt, the more expensive and disruptive that transition can become.

What’s notable across the ecosystem is that agentic systems are increasingly designed to sit on top of existing infrastructure rather than replace it outright. 

Avoiding engagement with these patterns may not be sustainable over time. If interaction norms shift and systems aren’t prepared, the combination of technical debt and lost opportunity may be harder to overcome later.

Where the agentic web stands today

The agentic web is still taking form, shaped largely by how people choose to use it. Some organizations are already applying agentic systems to reduce friction and improve outcomes. Others are waiting for stronger trust signals and clearer consent models.

Either approach is valid. What matters is understanding how agentic systems work, where they add value, and how emerging protocols are shaping participation. That understanding is the foundation for deciding when, where, and how to engage with the agentic web.

Read more at Read More