Google AI Overviews answer many queries directly on the results page, so fewer people click through to websites. Pew Research Center (July 2025) found users clicked a link on 8% of searches that showed an AI Overview, against 15% without one. Informational and top-of-funnel pages lose the most traffic; high-intent and cited pages hold up.
Most B2B SEO strategies were built for a results page that no longer exists. For 20 years the deal was simple: rank in the top three, earn the click, capture the visit. AI Overviews break that deal for a large share of informational searches. Google writes the answer at the top of the page and the user often gets what they came for without leaving. The loss isn’t evenly spread. Ranking still matters, and being quoted inside the AI Overview now matters as much.
Key takeaways
- AI Overviews cut click-through on informational and top-of-funnel queries. Pew Research Center (July 2025) measured an 8% link-click rate on searches with an AI Overview, against 15% without.
- The traffic hit is uneven. “What is” and “how to” pages are most exposed. Comparison, pricing, product, and bottom-of-funnel pages hold up.
- Getting cited inside an AI Overview is the new visibility. It rewards a clear answer near the top of the page, clean markup, existing top-5 rankings, and mentions elsewhere on the web.
- Keyword volume is now a weak proxy for value. A query with an AI Overview can have high search volume and almost no clicks.
- The first move is an audit of your top traffic pages, not a rebuild. Most sites need answer-structure fixes on 10–15 pages.
What actually changes when an AI Overview appears
When an AI Overview appears, Google generates a written answer above the organic results and cites a few sources inside it. Your organic listing gets pushed down the page, and a share of users get their answer without clicking anything.
Google launched AI Overviews at I/O in May 2024 and expanded them through 2025. The mechanism has two effects. First, position: even at #1, the Overview sits above you, so your listing drops below the fold on most mobile layouts. Second, substitution: when the Overview answers the question fully, the click never happens. That’s what Pew Research Center measured in July 2025: an 8% link-click rate with an Overview present, against 15% without. [DATA — VERIFY: confirm Pew July 2025 figures before publishing.] Pew also found users clicked a source link inside the Overview about 1% of the time, so “we’ll win the citation referral” is weaker consolation than it sounds.
Not every query triggers an Overview. Google shows them most on informational, research-style searches and rarely on transactional or brand queries. [DATA — VERIFY: current AI Overview trigger rate on B2B queries.] That split decides where your traffic is safe. For the wider picture, see our roundup of B2B digital marketing trends.
Which pages lose traffic, and which hold up
Exposure to AI Overviews tracks intent. The more generic and informational the query, the more likely Google answers it in the Overview and the more traffic you lose. The more specific or transactional the query, the safer the page.
| Page / query type | Exposure to AI Overviews | Why |
|---|---|---|
| “What is” / definition pages | High | Google answers in two sentences. No reason to click. |
| “How to” / basic guides | High | The Overview summarises the steps inline. |
| Listicles / “best X” roundups | Medium–High | Often synthesised into the answer, sometimes cited. |
| Comparison (“X vs Y”) pages | Medium | Buyers want detail and proof, so many still click. |
| Original data / research | Low–Medium | Frequently the cited source, which sends branded referral. |
| Product / solution pages | Low | Bottom-of-funnel intent, not informational. |
| Pricing pages | Low | Specific, transactional, and brand-owned. |
This is uncomfortable for a lot of B2B content plans, because the high-volume informational keywords those plans are built around sit in the top rows. A page targeting “what is marketing automation” is far more exposed than one targeting “marketing automation setup for [specific use case]”. That is why keyword volume has become a poor way to prioritise. A keyword with 12,000 monthly searches and a permanent Overview can send fewer real visits than one with 300 searches and none. The number that matters now is clicks you can win, not impressions Google might show. Pressure-test which target keywords still return a clickable results page before commissioning more posts.
Getting cited is the new ranking
Being cited inside an AI Overview is a distinct goal from ranking, and it’s earned differently. Google’s model picks sources that answer the specific question clearly, come from a trusted domain, and are structured so the answer can be lifted cleanly.
From what’s visible so far, a few things correlate with citation. [DATA — VERIFY: Google has not published official citation criteria; treat these as observed patterns.]
- A direct answer to the query in the first 100 words, written as a standalone statement a model can quote.
- Pages already ranking in positions 1–5, since Overviews pull from sources Google trusts.
- FAQ and how-to schema, and clean semantic markup.
- Third-party mentions, so the model sees corroboration beyond your own site.
None of that is new discipline. It’s the same relevance-and-authority work, aimed at extraction instead of the blue link. The practical shift is structural: stop burying the answer under 400 words of preamble. Say it in the first two sentences, then earn the rest of the read. This is where SEO and AEO overlap. AEO (Answer Engine Optimisation) applies the same idea to ChatGPT, Perplexity, and Google’s AI Mode, where buyers now research vendors before they reach your website. We treat this as one connected job rather than a separate line item, which is how we’ve built our services.
What this means for B2B specifically
B2B feels the shift harder than most sectors, because so much of the B2B funnel starts with informational research. The “how does X work” reading that used to land on your blog now often resolves inside an Overview or an AI chat tool, before a prospect knows your name.
The consequence is specific. Awareness-stage traffic that fed retargeting pools and email lists thins out, so you have fewer people to nurture toward a demo. For any company that built its pipeline on top-of-funnel content, that’s a real gap. [DATA — VERIFY: your informational vs commercial organic split from Search Console.] The offset is that bottom-of-funnel intent is largely untouched. Someone searching “[your category] pricing” wants to act, and Google mostly leaves those results alone. So the sensible response is to rebalance: protect and expand the high-intent commercial pages that convert, and stop grading content on informational volume that AI now absorbs. There’s a credibility angle too. When a model chooses who to cite, it favours brands with third-party coverage and review-site presence, so off-site work (guest content, PR, getting quoted, G2 or Capterra listings) now feeds both classic search and AI answers.
Should you stop investing in SEO?
No. SEO still works for B2B, but the return has moved. Value is draining out of high-volume informational rankings and concentrating in high-intent commercial pages and in getting cited by AI. Cut the budget entirely and you lose the bottom-of-funnel demand capture that Overviews barely touch.
The converting queries are the specific, commercial ones that were always your best traffic. Those buyers still search, compare, and click, because an AI summary doesn’t settle who to trust with a contract. What should change is the ratio. If your plan is 80% top-of-funnel pillar content and 20% commercial pages, that split now works against you. Weighting it toward the pages that either convert or get cited is a better use of the same budget. SEO isn’t the problem. A content strategy built for the 2020 results page is.
What this looks like in practice
For a UAE trade finance advisory, Express Trade Finance, we built the pipeline entirely on high-intent search. The programme ran on Google Ads and produced 100 high-intent leads a month, converting to high-value deals. [DATA — VERIFY: budget scale and campaign timeframe before publishing.]
The relevant part here is why that structure holds up under AI Overviews. The queries we targeted were bottom-of-funnel and commercial: people looking for a specific financial service, ready to enquire. Those are the searches Google is least likely to answer inside an Overview, because a generated summary doesn’t help someone choose who to trust with a trade finance deal. The demand sat in the part of search AI isn’t absorbing.
What we wouldn’t have recommended is chasing top-of-funnel informational keywords to “build awareness” first. It would have been slower, harder to attribute, and most exposed to substitution. Full write-up in the Express Trade Finance case study.
Worth saying plainly: this was a paid-search programme, not an organic one. The principle transfers, since it’s about targeting buying intent, but we’re not claiming an organic ranking result we didn’t run.
What most people get wrong
The common reaction is to keep publishing informational content at the same rate and hope citation makes up for the lost clicks. It looks reasonable, because the pages still get impressions and the rankings still show. It costs you twice: you keep paying to produce pages that convert less, and you starve the commercial pages that would.
The other error is treating AEO as a new service to buy later, separate from SEO. It isn’t a separate discipline. The work that gets you cited in an Overview (clear answers, clean structure, topical authority, off-site mentions) is the same work that ranks a page. Splitting them into two projects means paying twice for one job.
For most B2B sites the fix is boring. Audit the top 10–15 traffic pages, find the ones targeting informational queries that now trigger Overviews, and either rebuild them around a direct answer or move that effort to commercial intent. No rebuild. No new retainer line. Just a harder look at what each page earns.
Frequently asked questions
Will AI Overviews kill my organic traffic?
No, but they will reduce it on informational queries. Pew Research Center (July 2025) found link clicks roughly halved when an AI Overview was present (8% versus 15%). Commercial and bottom-of-funnel pages are largely unaffected, so your impact depends on how much traffic comes from “what is” and “how to” searches.
Which of my pages are most at risk?
Definition pages, basic how-to guides, and generic informational posts, because Google can answer those queries in a sentence or two. Comparison pages, pricing pages, product pages, and original research hold up, because they serve intent an AI summary can’t satisfy. Check your top traffic pages in Search Console against that pattern.
How do I get my content cited in an AI Overview?
Answer the query directly in the first 100 words, in a self-contained statement a model can lift. Add FAQ and how-to schema, keep the page technically clean, and build the topical authority that gets you into positions 1–5, since Overviews tend to cite pages that already rank. Third-party mentions help.
Is AEO different from SEO?
Not really. Answer Engine Optimisation aims the same relevance and authority work at AI answers instead of blue links. The techniques overlap almost completely: clear answers, structured markup, trusted domains, off-site citations. Treating them as one job is cheaper and more coherent than buying AEO as a separate service later.
Do AI Overviews appear on B2B searches?
Yes, and more often on the informational, research-style queries at the top of the B2B funnel. They appear far less on transactional and brand queries. The exact trigger rate for your category shifts over time, so check a sample of your own target keywords rather than trusting a blanket figure. [DATA — VERIFY: current B2B trigger rate.]
How do I track traffic lost to AI Overviews?
Google doesn’t label AI Overview impressions separately in Search Console yet, so it’s imperfect. Watch for pages holding their ranking position while click-through rate drops. That gap is the usual signature of an Overview appearing above you. Segment by query type to see whether informational pages fall faster than commercial ones.
Where to start
Pick your 10 highest-traffic informational pages this week. For each, rewrite the first 100 words so it answers the target query directly, in a statement a model could quote, then check whether the keyword still returns a clickable results page. That single pass tells you which pages to defend, rebuild, or retire.
If you want the wider system this fits into, read our guide to building a full-funnel B2B strategy. And if you’d rather we ran the audit with you, book a growth audit.
