Does link building even matter for AI Overview citations? I checked 50 SERPs to find out

Spent the last few weeks pulling AI Overview citations across 50 different SERPs (mix of niches: health, finance, gaming, software) and comparing the cited pages against their backlink profiles.

Here is what stood out.

1. Citation ≠ ranking position
Plenty of pages got cited in the AI Overview while sitting at position 6, 8, even 11 in normal blue links. Traditional ranking signals clearly aren't the only thing driving who gets pulled into the summary.

2. Backlink count didn't correlate the way I expected
I compared referring domains on cited pages vs non cited pages on the same SERP. No clear pattern. Some cited pages had 200+ referring domains. Others had under 10. What every cited page DID have in common: a direct, extractable answer sitting in the first 100 to 150 words.

3. Format beat authority more often than not
Pages using clear headers, short direct answers, and simple structure got cited even when a higher DR competitor with better backlinks sat right above them in normal rankings.

4. The exception: YMYL niches
In health and finance, backlink heavy pages from known sites still dominated citations. Topical authority and trust signals seem to matter more there. In lower stakes niches (gaming, software, how to), structure won more often than authority.

What this means for link building going forward (my theory, open to pushback)
Links still matter for getting you into the consideration set, ranking high enough to be crawled and seen as relevant. But links alone don't seem to be deciding who gets quoted once you're in that set. Content structure and answer clarity are doing more of that work now.

Question for the thread
is anyone else tracking AI Overview citations against backlink data?
Curious if this holds in niches I haven't tested yet, especially casino/iGaming and crypto, since those are heavy YMYL adjacent spaces where I would expect links to matter more.

nice research, this matches what i've seen too, links get you in the room, structure gets you the mic.

Interesting findings. One suggestion for next steps is tracking user intent alignment, not just backlinks. I've noticed pages that answer the exact query quickly and clearly seem to get pulled more often, even when they're not the strongest page from a link perspective. Would be very interested to see whether crypto and iGaming follow the same pattern or lean more heavily on authority signals.

Yes, many SEO's are seeing the same pattern: AI Overviews often pick the clearest, most directly extractable answer rather than the strongest backlink profile, especially in non-YMYL niches. In sensitive areas like finance, crypto or health, authority and trusted links still heavily influence citations. Overall it looks like structure + clarity drives selection, while backlinks mainly decide eligibility.

I've seen similar patterns. Backlinks seem to help a page become eligible, but they don't guarantee an AI Overview citation. What consistently stands out is answer quality and structure—clear, concise responses near the top of the page, strong topical relevance, and content that's easy for AI to extract. For YMYL topics, authority and trust still appear to carry much more weight.
I'd also be interested to see data from iGaming and crypto. My guess is those niches require both strong authority signals and highly structured content to earn consistent citations.

I think the bigger split is page discovery vs paragraph selection. Links may help Google keep a URL in the mix, but the actual citation often feels like it’s choosing the cleanest extractable block on the page, not the strongest page overall. I’ve seen weaker articles get cited just because the answer sat right there in 2-3 lines and didn’t make the system work for it.

Good test. I also think links still have some role, but they are not everything. I have seen pages with less backlinks getting picked because the answer was simple and easy to understand. Would be nice to see same test on crypto niche.

Really appreciate the effort you put into this. Did you find that high-authority editorial backlinks had a stronger impact than niche-relevant links, or was relevance the bigger factor?

hvent tested crypto specifically but ran similar check on finance affiliate sites few months back. What I notice matches your YMYL pattern, citations stuck heavily to sites with established editorial history even when competing page had cleaner format.One thing additional I caught was schema markup presence, pages with FAQ or HowTo schema seemed to get cited more even outside YMYL, like structured data give extraction system clearer signal. Worth checking if your cited pages had schema vs ones that didnt, might be hidden variable behind that pattern you found

Great effort mate, and this is true.