Thomas Peham: “AI Search Isn’t Just Another Version of SEO” & Other Interview Highlights

These are highlights from our interview with Thomas Peham. You can learn about Thomas’ work on Otterly AI or connect with him on LinkedIn.

Why AI search is more than SEO

To optimize for AI visibility, monitoring language models alone isn't enough. The future is about optimizing for the interface, not the model itself. I don't think of ChatGPT, Claude, or Perplexity as traffic channels. They are more brand discovery channels.

Users don't open ChatGPT looking for ten blue links. Instead, they ask a question and receive an answer. That's why some people call it Answer Engine Optimization instead of Search Engine Optimization.

The biggest misconception is believing AI search is just another version of SEO.

Traffic, referral traffic, and conversions matter, of course. But we're gradually moving beyond optimizing purely for clicks.

SEO rarely became a board-level conversation because SEO teams often focused almost entirely on the website:

• More impressions.
• More clicks.
• More traffic.

Eventually, those clicks might generate leads or revenue. But AI search gives marketing a much broader opportunity.

Instead of arguing about SEO vs. GEO, we should be asking: “How do we make our brand discoverable across every search and AI interface?”

Website content, technical SEO, digital PR, earned media, brand marketing, and content marketing contribute to whether AI systems understand and recommend your brand.

AI search is fundamentally a marketing initiative, not simply an SEO initiative.

If you aren’t in the answer, are you missing the buying journey?

If your competitors consistently appear in AI answers while you don't, you'll almost certainly lose opportunities.

The simplest way to understand your current AI visibility is to ask your customers: "Where did you first hear about us?" That tells you whether AI platforms are already influencing discovery.

You can also monitor referral traffic from AI platforms and track website citations.

To a large extent, if your brand isn't mentioned in an AI-generated answer, we can say you're effectively absent from the buying journey. But…

Being remembered is often more valuable than simply being clicked. That’s why brands should optimize first for visibility and second for website citations.

The internet now defines your brand

There's often a disconnect between how a brand wants to be perceived and how the internet actually talks about it.

The biggest issue I've seen between how companies describe themselves and how AI describes them is consistency.

Another problem is simply outdated content.

Many organizations still have outdated pages somewhere on their website with old contact information or obsolete product descriptions.

With traditional search, outdated pages often disappeared onto page two or three. With AI search, those same pages may become part of the answer. That's why auditing and updating existing content is so important.

The larger your website is, the greater the chance that some of your information is no longer accurate or consistent.

Ultimately, AI visibility comes back to two principles: consistency and frequency. The brands succeeding today are doing the unglamorous work of making sure their messaging is accurate everywhere.

How to be more visible these days

Which platforms matter most for AI visibility

LinkedIn and YouTube are becoming significantly more influential for AI search than other platforms.

How people describe your company in YouTube videos increasingly affects AI-generated answers.

Editorial publications and earned media continue to play a major role as well.

We're no longer optimizing for one search engine. We're optimizing across multiple AI systems, each with slightly different behaviors.

Brand mentions vs. backlinks vs. technical SEO for AI visibility

Compared with Google, ChatGPT has often been more willing to reference smaller sites.

That means that even relatively small pieces of content can shape AI-generated perception.

If you were to choose, focus on brand mentions instead of technical SEO and backlinks, as they are the strongest drivers of AI visibility.

What metric matters more: Google rankings vs. AI citations vs. direct traffic

I'd choose direct traffic over Google rankings and AI citations because it remains one of the strongest indicators of brand awareness. It's also influenced by everything happening across AI search and traditional search.

Company blog vs. digital PR vs. community building for visibility

I'd probably choose digital PR over a company blog and community building because I've seen it create an enormous impact recently.

Reddit vs. Wikipedia vs. YouTube for mentions

From a purely quantitative perspective, I'd have to choose Reddit over Wikipedia and YouTube. Because it's currently one of the most-cited websites across AI search. But…

If we're thinking strategically rather than statistically, Wikipedia deserves much more attention than most companies give it. I think many brands still underestimate how valuable a strong Wikipedia presence can be.

Why everyone suddenly cares about Reddit

Don't ruin Reddit.Reddit is fundamentally a community platform. It always has been. It always will be. The biggest mistake brands can make is treating it like another advertising channel.

There are good strategies for Reddit promotion:

• Creating valuable discussions.
• Supporting communities.
• Helping users.
• Encouraging employees to participate authentically.

The biggest thing Thomas was wrong about

I've been wrong about plenty of things. These are some of the major ones:

First of all, llms.txt.

Early on, I believed llms.txt would become an important optimization factor. Later, after testing it ourselves, we found almost no measurable impact on AI visibility.

Besides, around late 2025, I said publicly that YouTube wasn't an important part of an AI search strategy, especially for ChatGPT.

With AI, something that's true today may no longer be true three months from now. Marketing strategies (and especially AI search strategies) need to adapt much faster than annual business planning cycles usually allow.

LinkedIn: Useful platform or marketing theater?

The biggest problem isn't LinkedIn itself, but the hype.

LinkedIn has become one of those places where marketers try very hard to position themselves as experts.

LinkedIn is incredibly valuable for AI visibility because it’s an open ecosystem. Most profiles, posts, and pulse articles are public. Meaning, search engines and AI systems can crawl and reference the content quite easily.

My advice is simple:

• Encourage employees to share authentic experiences.
• Share useful knowledge.
• Cut out the fluff.

People (and increasingly AI systems) respond much better to genuine expertise than to manufactured thought leadership.

How to make your testing effective

Whenever we test anything, we change one factor at a time. Everything else stays constant.

The better the experiment is designed, the more trustworthy the results become.

If we discover that an experiment doesn't support a clear conclusion, we simply don't publish it.

I'd much rather publish fewer studies than publish something we can't confidently stand behind.

AI biases

The challenge is that every LLM already reflects certain biases. These systems are built by companies with their own histories, cultures, and assumptions.

Before asking whether AI is biased, we also have to ask whether the information it's learning from is biased.

What should you measure in the AI era?

There are really two metrics to care about most: brand visibility across AI search platforms and self-reported attribution. Together, they tell me two different stories. One measures brand awareness. The other measures business impact.

Accompanied, they're much more meaningful than rankings alone.

How to measure value without a click

That's one of the biggest challenges today.

There are really only two direct ways to prove the value of AI visibility: self-reported attribution and referral traffic. Everything else becomes indirect.

Why AI data shouldn't stay in dashboards

AI visibility data should stop living inside dashboards and instead become part of everyday workflows and even AI agents.

Historically, marketing data has been fragmented. SEO data lived in SEO tools. CRM data lived inside the CRM. Analytics lived somewhere else.

Now, your AI assistant can analyze information across all of the channels and platforms simultaneously. That's much more powerful than looking at isolated reports.

What’s the future of analytics tools?

I think the next generation of search tools will feel less like dashboards and more like conversations because that’s already happening.

Analytics is becoming conversational. Instead of searching through charts, you simply ask questions. Because all the data is connected, the system can answer immediately.

But I think both models will continue to exist:

• Static reporting for structured business reviews.
• Conversational analytics for exploration and decision-making.

Building a product in the age of AI

The AI market is changing incredibly quickly. For example, when Thomas was building Otterly AI:

• ChatGPT introduced web search.
• Google launched AI Overviews.
• Then Gemini evolved.
• Then AI Mode appeared.

On top of that, every country experienced these changes differently.

Building a global product in the AI age means constantly balancing product development with customer education.

Giving people immediate access to your software product

As a software buyer, the biggest frustration for me was always being unable to evaluate a product independently. I still believe giving people immediate access to your software product creates a much better buying experience.

Who should own AI visibility?

Ultimately, AI visibility is a marketing channel. That means accountability belongs with marketing leadership, typically the CMO or Head of Marketing.

That said, implementation usually starts with SEO.

SEO teams already understand search, content structure, technical optimization, and discoverability, so they're often in the best position to lead the execution.

An SEO team can't succeed alone when it comes to AI visibility. You also need content marketing, PR, brand marketing, technical teams, and product marketing.

AI visibility isn't a one-person job. I like to compare it to a road trip. The Head of Marketing is driving the vehicle, but everyone else is sitting alongside them.

Rankings can’t be the main KPI

It’s a dead idea that rankings are still the main KPI.Traditional rankings don't matter in the same way inside ChatGPT.

GEO vs. SEO: Are they the same?

GEO is not just SEO with a new name. The interfaces, user behavior, and the customer journey are different. GEO is much broader than simply renaming SEO.

These are highlights from our Adsy Talk podcast with Thomas Peham. For more insights, watch the full interview on our YouTube channel: