10 keyword research shifts since AI Overviews launched

Google's AI Overviews changed what "ranking" means. A page in position 1 can still get cited by the Overview above it — or completely bypassed. Here are 10 shifts that change how to pick keywords.

1. Informational queries lose half their organic clicks

When Google synthesises an answer at the top of the SERP, the click-through to position 1 drops by 40-60%. Pure informational keywords are no longer worth the effort they used to be.

2. Transactional intent gained value

Queries with clear buying intent ("best X", "X review", "X pricing") still drive clicks because AI Overviews handle them cautiously. Shift budget toward commercial keywords.

3. Long-tail wins bigger now

AI Overviews appear less frequently on long-tail queries. The further you go from common phrasings, the more likely the organic top 10 keeps its full traffic share.

4. Citation in the Overview replaces position 1

Being cited in the AI Overview now drives more attributable visits than ranking #1 with no citation. Optimise for Overview citability — clear answers, schema markup, factual claims.

5. Brand-name queries became more defensible

AI Overviews avoid hallucinating brand details. Branded queries route to your own site reliably. Investing in brand-search demand pays out cleaner than it used to.

6. "How to" content gets summarised, not clicked

Step-by-step content is the easiest for the Overview to compress. If your strategy depended on "how to X" listicles, expect a 50%+ traffic drop on those pages.

7. Comparison content is more durable

"X vs Y" keywords still drive clicks because the Overview can't easily make commercial recommendations. Comparison pages have become disproportionately valuable.

8. Featured-snippet optimisation became Overview optimisation

The structure that won featured snippets — concise answer at the top, supporting detail below — is what wins citations in AI Overviews now. The same playbook, higher stakes.

9. Schema markup graduated from optional to required

Without structured data (Product, FAQPage, HowTo, Article), your content is invisible to the model that picks Overview citations. Pages without schema rank in Overviews at a tiny fraction of the rate of pages with it.

10. Search volume tools became less accurate

The traffic numbers in keyword tools are extrapolated from Google's old click-distribution model. AI Overviews break that model. Treat search volumes as upper bounds, not expected clicks.

The new keyword research playbook

Shift your filtering criteria:

1. Down-weight pure informational keywords with high SERP-feature competition.
2. Up-weight commercial / transactional / comparison terms.
3. Look at the SERP as it actually displays today — including Overviews, video carousels, knowledge panels — not the metric tool's idealised version.
4. Optimise for citability, not just ranking.

SERPTool flags AI Overview occupancy + SERP feature density per keyword, so you can see which targets will keep their clicks and which won't.

Run a free analysis — 40 credits on signup.