There were so many search changes in 2025 that it was difficult to keep up with the industry.

Mark Williams-Cook has been closely tracking these changes. In this AMA, he answers burning questions about declining clicks, new Generative Engine Optimization tactics, and search skills that will be critical for success in 2026.

1.Google claims AI Overviews drive “higher quality” clicks. Is that just PR spin, or are you seeing evidence of something else in the data?

Google says searches are up, about 20% more queries from 2024 to 2025, but total organic clicks are “relatively stable.”  That means fewer clicks per search, especially on informational queries where AI Overviews answer the question directly. 

Google’s claim is that average click quality has improved, not that all clicks are better. In reality, we’re getting fewer clicks overall, but the ones that filter through may be more valuable. It’s a clever way of saying: “You’re not getting more traffic, but what you do get is higher quality.” So yes, there’s definitely some PR spin in how it’s presented.

2. How much does Chrome data support Google’s monopoly and profitability?

Well, the antitrust trial didn’t find it impactful enough to force them to drop it. I think Chrome data probably plays a bigger role in PPC than SEO, mainly in profiling user behavior.

What caught my attention was how the DOJ trial lined up with the content warehouse leak. There were mentions of Chrome passing URL popularity scores. However, Google insists they don’t use Chrome data in rankings, and I tend to believe they’re not lying, but they phrase things very carefully.

Still, it wouldn’t surprise me if Chrome data is used to determine which URLs should be crawled, indexed, or even re-ranked based on popularity. However, I don’t think it’s as powerful as SERP usage data, such as who clicks what, how long they stay, and whether they bounce or reform queries.

Looking ahead, browsers will matter a lot more. LLM search platforms often rely on web search in the background, which heavily depends on SERP usage data. If search moves into LLMs, we lose that data.

That’s why I think we’re seeing companies like Perplexity and ChatGPT building browsers as a way to track what users find helpful and to bring agentic AI experiences directly into the browser. 

So yes, browser wars are definitely coming!

3. What insights do you have about LLMs ranking for multilingual queries? 

This is a fundamental issue with a model based on language rather than on documents delivered in containers or websites that can provide important context. 

We’ve seen similar issues for years with features like People Also Ask because those suggestions are also language-model-driven. For example, if you do a search in the UK, you’d get answers like Walmart and U.S.-centric results that are completely irrelevant to local users.

If the AI system uses retrieval-augmented generation (RAG), standard optimizations still apply. But for the base model’s behavior, unless the user explicitly sets their region or context, there's not much you can do to control it.

Honestly, this feels like a platform problem, not a publisher problem. Until the AI search providers build in better contextual awareness, there’s only so much we can do.

4. What does “quality content” look like in 2026 when generative AI can produce endless copy?

That’s the billion-dollar question. I don’t think SEOs have ever fully nailed what quality content means. We’ve called it unique, helpful, or good, but those labels are vague.

At an SEO FOMO event, Gary Illyes said Google will be focusing more on original content. That’s not new, but what counts as original has changed. 

Ten years ago, it meant rewriting existing info. Then we discussed information gain, which involves adding something new. More recently, it’s about offering a personal point of view, because AI can’t replicate lived experience.

I think of content in two buckets:

• Solved information:  Slow-changing facts that AI can confidently generate. If your website relies on top-of-the-funnel information, expect it to get eaten up by AI answers. LLMs don’t need to perform any grounding or retrieval for these queries because they're so confident in their ability to generate the answer.
• Human content: Copy that reflects emotion, opinion, or storytelling. I've developed a sixth sense when reading something that's AI-related, and I immediately feel disappointed, disengaged, and don't want to continue reading it. People are craving something more human, which is why creators like Ed Zitron have thriving, paid newsletters.

Ultimately, it comes down to your business model. If you rely on cranking out high-volume, low-differentiation content, this change may be existential. But for everyone else, it’s an opportunity to do better work and stop rewarding fluff.

5. Is video now a required SEO tactic to increase visibility in AI and PAA results?

Video isn’t required, but it’s a great starting point, especially when working with subject matter experts who struggle with writing. 

We often record interviews with clients and then use AI to create content based on their unique perspectives. It’s often faster and more aligned than writing from a brief, because you’re capturing what they think, not just answering prompts.

As for People Also Ask, I don’t focus much on trying to rank there. I use it more for understanding search intent. It tells me what Google expects to see near a topic, and that’s helpful for content creation.

6. Are we overestimating the threat of answer engines, or is complacency dangerous?

Complacency is always dangerous. Regardless of the industry, the same money flows from customers to companies online. So, when we say “threat,” we must define what it means.

In terms of LLM search versus traditional search, tactics such as digital PR and community building are becoming increasingly important, but they’ve always been core to SEO.

I think the thing we might be complacent about is how we measure success. You can succeed in AI search without generating any traffic. 

For example, I recently chose a conveyancer to move house with by using LLMs to summarize reviews of specific companies near me and get an overview of what they were good at, what they weren't so good at, and the prices people had discussed.

 Then, I Googled the company, went to their website, and converted there. I didn’t drive any LLM traffic for them, but they still won my business.

So, if you're only tracking traffic from tools like ChatGPT, you're missing the bigger picture. AI is more of an influencer-type lever than a traffic channel, and that’s a critical distinction.

7. What would you emphasize when communicating SEO value to C-suites with all these changes?

At Candour, we’ve started sending short weekly explainer videos to clients, about 2 to 4 minutes each, and designed specifically for stakeholders. The videos cover key topics like platform shifts, usage statistics, and risks like hallucinations that could impact their brand.

The goal is to drip-feed useful information over time. You’re not going to change someone’s mind in a single meeting. You need to break it down, tailor the message for each audience, and be consistent.  It’s about planting seeds so they come to the right conclusions themselves, at their own pace. 

8. What’s one 2025 search trend you think was total BS, and what should we be paying attention to instead?

For me, it’s the rebranding of basic SEO under fancy GEO terminology. Many of the tactics being touted as new are things we’ve always done, just renamed.

Take “chunking” as an example. It’s framed as a technical breakthrough where you structure content so LLMs can digest it more effectively. 

But really, it’s just writing clearly, breaking up topics, using headers, and not dumping massive paragraphs into a wall of text. That’s how we’ve been writing for people for years.

Instead of obsessing over buzzwords, focus on the fundamentals. Make content better, clearer, and more useful. That’s what moves the needle.

9. Is building communities (forums, Discord, Slack) a smarter hedge than competing head-to-head?

Yeah, for sure. When Reddit started dominating visibility, first in Google and then in LLMs, I saw many people saying, “We need to be posting there.” That always makes me cringe a bit, because most of those efforts tend to go poorly. I’ve been a Redditor for 19 years, and I’ve seen how that usually plays out.

With clients, we always start by identifying where their audience already spends time. There’s no point launching a Discord if your customers aren’t there. But when you get it right, community marketing becomes an engine for everything else, aligning with where search is headed.

Historically, search has been drifting away from webmaster-controlled spaces for years. PageRank was one shift, and AI is another. Now, it's more about what the broader internet is saying about you, such as your sentiment, visibility, and reputation, than about your site.

Communities help shape that narrative. If your product is good, a community often forms anyway. However, investing in, growing, and managing it gives you influence. It’s a smart hedge, because even if search disappeared tomorrow, a strong community would still be incredibly valuable.

10. Are there any search skills or tools SEOs should start learning now to get ahead in 2026?

Here are a few things to focus on:

Learn how large language models work

SEOs often focus on tactics, usually due to pressure from managers or stakeholders. That’s understandable, but knowing how a language model works, how pretraining happens, how text is generated, and what influences it offers a competitive advantage.

When you understand the technology, you can combine your LLM and marketing knowledge to develop your own strategies, rather than relying on recycled advice.

It also provides protection; understanding the system helps you avoid tactics that may seem effective in the short term but are risky in the long term. Most clients want sustainable success, not short spikes followed by penalties.

Follow thought leaders

Follow thought leaders like Dawn Anderson and Britney Muller. Both were central to the SEO world and then transitioned into fields like information retrieval and machine learning. That technical perspective is incredibly valuable and will become more important in the future.

LLMs are influence channels, not traffic sources, but they are still important for brand visibility. Take time to learn how they work under the hood so you can adapt your SEO strategy to optimize visibility. Finally, collaborate with thought leaders to create content that AI can’t replicate as a way to grow your own channels.