ChatGPT and SEO: What Changes and What Doesn’t

ChatGPT affects SEO by shifting how people find information, not by making search irrelevant. Users who once typed queries into Google are increasingly asking conversational questions in AI interfaces and accepting synthesised answers without clicking through to source pages. For marketers, that changes the economics of organic traffic without eliminating the value of ranking well.

The structural question is not whether ChatGPT will replace search. It won’t, at least not entirely. The more useful question is which parts of your SEO investment are exposed to this shift and which are not.

What Is Actually Happening to Search Behaviour?

When I started in this industry around 2000, search was young and nobody really knew how to use it commercially. The idea that someone would type a question into a box and get a useful answer felt almost miraculous. Twenty-five years later, we have built entire business models around that behaviour, and now the behaviour is shifting again.

What ChatGPT and similar tools have done is make conversational, synthesised answers fast and accessible. For a large category of queries, particularly informational ones where someone wants to understand a concept or get a quick answer, an AI interface can now satisfy that need without the user visiting a website at all.

This is not a hypothetical future state. It is happening now. Queries like “what is a meta description” or “how does compound interest work” or “what are the symptoms of burnout” are being answered directly inside AI tools. The user gets what they need. Nobody gets the click. HubSpot has written about how marketers are beginning to factor this into their content strategies, and the conversation is moving fast.

The categories less affected are those where intent goes beyond information. If someone searches for a solicitor in Manchester, a hotel in Edinburgh, or a specific product they want to buy, they still need to go somewhere to complete that action. AI can assist the research phase, but the transaction still happens elsewhere.

Which SEO Traffic Is Most at Risk?

Not all organic traffic is equally exposed. The risk varies significantly depending on what type of content you have built and what intent it serves.

Informational content targeting broad, definitional queries is the most vulnerable. If your SEO strategy has been built around capturing volume on “what is” and “how does” queries with content that essentially restates publicly available information, that traffic is under real pressure. AI tools answer these questions competently and without requiring a click.

Transactional content is more durable. People buying something, booking something, or looking for a specific service provider still need a destination. The AI may help them narrow their choices, but it does not complete the purchase. This is one reason local SEO, which is fundamentally about connecting intent with a physical or specific provider, remains structurally sound. Moz has documented how local search continues to drive meaningful commercial outcomes even as the broader landscape shifts.

Opinion, analysis, and experience-based content is also more resilient. A well-argued point of view, a case study grounded in real data, or a piece that reflects genuine expertise in a niche area is harder for AI to replicate convincingly. This is partly a quality issue and partly a trust issue. People who want a genuine perspective from a credible source are not fully satisfied by a synthesised average of the internet.

I have judged the Effie Awards, which means I have spent time evaluating what marketing effectiveness actually looks like when it is measured rigorously. One thing that stands out consistently is that the work which endures is the work built on a genuine understanding of the audience, not the work that chased a format or a channel because it was fashionable. The same logic applies here. Content built on real expertise holds up. Content built on keyword volume alone does not.

If you are thinking about how this fits into a broader SEO framework, the Complete SEO Strategy hub on The Marketing Juice covers the interconnected decisions that determine whether an SEO programme holds its value over time.

What Does “Being Cited by AI” Actually Mean for Your Brand?

There is a new distribution dynamic worth understanding. When ChatGPT, Perplexity, or Google’s AI Overviews generate an answer, they sometimes cite sources. Being one of those cited sources is increasingly valuable, because it places your brand in front of a user who is actively seeking information on a topic you are authoritative on, without requiring them to have found you through a traditional search ranking.

The question is what determines whether AI tools cite you. The honest answer is that the mechanisms are not fully transparent, but the patterns are consistent with what has always driven authority in SEO: depth of coverage, quality of sourcing, clarity of structure, and the degree to which other credible sources reference you.

This is not a new game with new rules. It is an extension of the same game with a different scoreboard. The brands that built genuine topical authority through consistent, well-structured, well-sourced content are the ones appearing in AI-generated answers. The brands that bought links and published thin content to hit keyword targets are not.

Structured content matters here too. Clear headings, concise definitions, well-organised sections, and content that answers specific questions in a readable format are all signals that help AI tools identify and extract useful information from your pages. This is not about gaming the system. It is about writing clearly, which has always been the right approach.

How Should You Adapt Your Content Strategy?

The adaptation is less dramatic than some commentators suggest, but it is real. A few things are worth prioritising.

First, audit your informational content honestly. If you have pages ranking for high-volume informational queries that are essentially definitional, assess whether those pages are delivering commercial value or just traffic. Traffic that does not convert was never as valuable as it looked in the dashboard. This is a point I have made to clients for years. Vanity metrics are comfortable until the business asks what they produced.

Second, invest in content that reflects genuine expertise. This means first-hand experience, specific data, original analysis, and clear points of view. These are the things AI cannot easily replicate, because they require either lived experience or access to proprietary information. A piece that draws on your own client data, your own operational experience, or your own research is inherently more distinctive than a piece assembled from publicly available sources.

Early in my career, when I was told there was no budget for a new website, I taught myself to code and built it myself. That experience taught me something about the relationship between constraint and resourcefulness, but it also taught me that the best solutions often come from understanding the problem at a technical level rather than delegating it entirely. The same instinct applies to AI and SEO. You do not need to understand the architecture of a large language model, but you do need to understand what it is optimising for and what it cannot do.

Third, think about your content in terms of topical authority rather than individual keyword rankings. If you want to be cited as a credible source on a subject, you need to cover that subject comprehensively and consistently over time. A single well-optimised page is less powerful than a body of interconnected content that demonstrates sustained expertise. Search Engine Journal has tracked how Google’s approach to authority has evolved over time, and the direction of travel has always been toward depth and credibility rather than volume and optimisation tricks.

Fourth, do not abandon SEO in favour of AI optimisation as if they are separate disciplines. They are not. The signals that make content rank well in traditional search, authority, relevance, structure, and trust, are largely the same signals that make content likely to be cited by AI tools. The investment is not wasted. The measurement model is changing.

What Happens to Keyword Research and On-Page SEO?

Keyword research does not become irrelevant, but the framing shifts. The most useful question is no longer simply “what are people searching for” but “what are people trying to accomplish, and at what point in that process do they need a human source rather than a synthesised answer.”

Queries with high commercial intent, specific geographic context, or a need for personalised advice are still best served by a real source. Queries that are purely definitional or factual are increasingly being absorbed by AI interfaces before they reach a search engine at all.

On-page SEO remains important, but the emphasis shifts toward clarity and structure. Well-organised content with clear headings, concise answers to specific questions, and logical information architecture is both more readable for humans and more parseable for AI systems. These are not competing objectives. Writing clearly for a human audience produces content that performs well across both contexts.

The integration of SEO with other channels also becomes more important as the informational traffic layer gets compressed. The case for integrating SEO and paid search strategy has always been sound, and it becomes more compelling when organic informational traffic is under pressure. Paid search can cover gaps. Retargeting can capture users who engaged with AI-generated answers but did not convert. The channels work together, and marketers who manage them in silos will miss the compounding value.

Is This the End of SEO as a Channel?

No. And I say that not as reassurance but as a commercial observation. Search engines still process an enormous volume of queries every day. The proportion being intercepted by AI tools is growing, but it is not the whole picture, and it is not uniform across intent types.

What is changing is the mix. The informational layer of organic search, which was always the largest by volume and the weakest by commercial intent, is under the most pressure. The transactional and navigational layers are more durable. This means the overall traffic numbers may decline for many sites, but the quality of the remaining traffic may actually improve if the measurement is done honestly.

I have managed significant ad budgets across a wide range of industries, and one thing I have seen repeatedly is that marketers confuse volume with value. A smaller number of highly qualified visitors who convert is worth more than a large number of informational visitors who do not. If AI tools absorb the latter category, the impact on actual business outcomes may be smaller than the headline traffic numbers suggest.

The organisations that will struggle are those that built their entire content strategy around capturing informational volume and never developed a clear model for converting that traffic into commercial outcomes. The organisations that will adapt well are those that treated content as a means to build authority and trust, not just a mechanism to generate impressions.

The broader forces reshaping how organisations use technology and information are significant, and BCG has written thoughtfully about how structural change affects organisations over time. The pattern in marketing is consistent with what happens in other functions: the tools change faster than the strategic thinking does, and the organisations that hold their nerve and keep their focus on outcomes rather than channels tend to come out ahead.

The complete picture of how SEO fits into a sustainable acquisition strategy, including how to think about content depth, authority building, and channel integration, is covered in the Complete SEO Strategy hub. If you are re-evaluating your approach in light of AI, that is a useful place to think through the interconnected decisions.

The Measurement Problem Nobody Is Talking About

Here is something that deserves more attention than it gets. If AI tools are answering queries without generating clicks, your existing measurement infrastructure will not capture the influence those interactions have on your audience. A user who encounters your brand name in an AI-generated answer, then searches for you directly, then converts, will look like a direct or branded search conversion in your analytics. The AI-assisted touchpoint is invisible.

This is not a new problem. Attribution in marketing has always been an approximation rather than a precise science. Analytics tools show you a perspective on reality, not reality itself. But the AI layer makes the gap between what is measured and what actually happened wider, and marketers who do not account for that will systematically undervalue their content investment.

The practical implication is that brand search volume, direct traffic trends, and conversion rates from branded queries become more important as leading indicators of content effectiveness. If your content is being cited in AI answers and building awareness, you should see it in those signals before you see it in traditional organic traffic metrics. Watching those numbers alongside your organic rankings gives you a more complete picture of whether your authority-building is working.

Forrester has written about the challenge of aligning different parts of an organisation around shared metrics, and the tension between sales and marketing measurement is a version of the same problem: when the measurement model does not reflect how value is actually created, decisions get distorted. The same risk applies to SEO measurement in the AI era.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions