6 min read
Is SEO dead now that everyone uses AI? What the data says
Every few years something is declared the death of SEO. This time the claim has more substance than usual, because AI-generated answers really do satisfy some searches without a click. But the conclusion people draw from it - stop investing in search - is the wrong one.
Here is a more useful framing: search is not shrinking, the click-through rate on certain query types is. Knowing which types lets you shift effort instead of abandoning the channel.
What actually changed
Two things. First, AI Overviews and assistant answers now handle a large share of simple informational queries - definitions, conversions, quick facts. Those searches used to produce a click to a page that gave a one-line answer. Many of them no longer do.
Second, the entry point moved for some people. Rather than searching and scanning, they ask an assistant and read a synthesised answer with a few citations. That is a smaller, more competitive shortlist than page one of Google.
Which traffic is genuinely at risk
- •Thin definitional content - 'what is X' pages with a paragraph of answer and nothing else. This is the most exposed category.
- •Simple how-to content that can be summarised in three steps.
- •Aggregated list posts that mostly restate what is already on other sites.
- •Basic comparison content with no original data or firsthand testing.
Which traffic is holding up
- •Commercial and transactional queries, where people want to evaluate and choose a provider. These still convert through a click.
- •Local and 'near me' intent, where the outcome is a visit or a call.
- •Deep, specific content with original data, firsthand experience, or proprietary examples that a model cannot synthesise from elsewhere.
- •Branded search, which grows as awareness grows and is unaffected by AI summarisation.
- •Interactive utility - calculators, assessments, configurators - which cannot be answered inside a chat window.
The practical adjustment
The strategy shift is from volume to depth and from information to decision support. Fewer, better pages that help someone choose, backed by things only you have: your data, your process, your client outcomes, your firsthand judgement.
- •Audit which of your pages are purely definitional and either deepen them or fold them into stronger pages.
- •Shift budget toward content that supports a buying decision rather than a curiosity.
- •Add interactive tools where they genuinely help - they earn engagement and links that static pages do not.
- •Optimise for citation as well as ranking, so the queries that end in an answer still surface your name.
- •Measure branded search and direct traffic alongside organic clicks, because AI visibility often shows up there first.
The uncomfortable part
If a large share of your organic traffic came from thin informational pages, some of that is not coming back, and no amount of optimisation will restore it. The honest response is to reallocate rather than to keep topping up a declining asset.
The businesses doing well in this shift are the ones that treated search as a way to demonstrate expertise rather than a way to accumulate pageviews. That approach was always more durable. AI just made the difference obvious sooner.
Frequently asked questions
Want this done for you?
Sapphire Digital runs AI SEO for founders and growing teams - technical foundations, structured data, answer engine optimisation, and AI-scaled content production.
Keep reading
What is AI SEO? A plain-English guide for founders
AI SEO means two things at once - using AI to do search work faster, and optimising so AI assistants cite you. Here is how both work in practice.
How to get your business cited by ChatGPT and Perplexity
Assistants name a handful of sources per answer. Here is the practical work that decides whether yours is one of them.