The future of search and getting found

AI overviews are killing organic search. Here’s what’s left.

Key takeaway: Google’s AI answers have already killed traffic to basic how-to and definition content, and comparison searches are next over the coming 12 to 24 months. Transactional searches survive longest, since an AI can recommend a hotel but can’t take payment or log someone into their account. The bigger shift is that search is moving from ranking URLs to citing trusted sources, and standard SEO only gets you crawled, not cited, which means entity clarity and schema now matter more than rankings.

The 10 blue links aren’t dead yet, but they’re on borrowed time.

Google isn’t switching off organic search overnight. It’s doing something slower and more annoying: quietly starving it, one query type at a time.

If you run a website that relies on search traffic, you need to know exactly which searches are already gone, which ones are going next, and which ones will still send you visitors in five years. Get this wrong and you’ll spend next year’s marketing budget fighting a battle that’s already over.

• Informational searches are already gone.
• Comparison searches are next.
• Transactional searches will survive.
• Search is turning into a citation game, not a ranking game.
• What to actually do about it.

Informational searches are already gone.

If someone searches “how do I fix a leaky faucet” or “what’s the current inflation rate”, Google answers them on the results page. No click. No visit. No chance for your website to say hello.

This isn’t a maybe. It’s already happened. Sites built on basic how-to content and definition pages have watched organic traffic fall by anywhere from 15% to over 60%. People got their answer and moved on with their day.

If your content strategy is “write a blog post that answers a simple question”, that strategy stopped working. The question gets answered before your page even loads.

Comparison searches are next.

Give it 12 to 24 months. Searches like “best CRM for estate agents” or “product x vs product y” will get the same treatment. Google’s AI will pull specs, reviews and forum opinions straight into a chat window, and organic listings will get shoved below the fold.

This is bad news if your business model depends on listicles, affiliate content or review directories. Positions four through ten on a results page are about to become invisible. Nobody scrolls that far when the AI has already told them the answer.

Transactional searches will survive.

Here’s the bit people miss when they panic about AI killing search: transactions still need a destination.

An AI can tell someone which hotel to book. It can’t take their payment. It can’t log them into their bank account. It can’t complete the purchase for them. At some point, the person has to land on a real website and do the thing.

That’s why bottom-funnel searches, the ones with clear buying intent, will hold their value longer than anything else. Organic search doesn’t disappear here. It just gets smaller and more valuable.

There’s also a simple reason AI search can’t kill the open web completely: it needs the open web to work. If publishers stop writing because nobody visits their sites, the AI has nothing left to learn from or cite. The two need each other, whether either side likes it or not.

Search is turning into a citation game, not a ranking game.

The real shift isn’t traffic. It’s what search is actually for.

Search used to be an index of URLs. You wrote a page, Google ranked it, people clicked it. Now search is turning into a database of citations. The AI picks a small number of sources to name and trust inside its answer, and everyone else gets ignored, regardless of where they’d have ranked in the old system.

Standard SEO gets your page found by a crawler. It doesn’t get your brand named inside an AI answer. Those are two different jobs now, and most agencies are still only doing the first one.

What to actually do about it.

Stop writing content that only answers a simple question.

If the answer fits in a sentence, an AI will give it away for free before anyone reaches your site. Write about the stuff that needs your judgement, not the stuff a bot can summarise from three other websites.

Get your entity clarity sorted.

AI models need to know who you are, what you do and where you do it, consistently, across every page. Vague “about us” copy and mismatched business details across your site make you harder to cite, not easier.

Fix your schema.

Organisation, person and service schema aren’t optional extras any more. They’re how you tell an AI crawler exactly what it’s looking at, without making it guess.

Protect the bottom-funnel pages properly.

Your service pages and product pages are where the real value now sits. Don’t bury them under a pile of thin blog posts that AI overviews have already made pointless.

None of this is about chasing rankings any more. It’s about being the source an AI decides to trust. That’s a different job, and it needs a different plan, not a rehash of what worked in 2019.

Editorial note: I haven’t included a specific client case study in this post as none was supplied. If you want this backed up with real before-and-after data from one of our own AI visibility projects, let me know and I’ll pull one together.