The Ten Blue Links Are Gone: We Dissected 30 SERPs

There is a picture in your head when someone says “Google results”: a tidy column of ten blue links, ranked top to bottom, and a race to be the first one. That mental model is doing a lot of quiet work in how the industry talks about rankings — “get to #1” only means “win the page” if the page is, in fact, ten blue links. So we checked whether it still is. We pulled the SERP-feature composition of 30 keywords across nine niches and counted how much of each results page is not an organic link.

The short answer: the plain ten-blue-links page is extinct in our sample. Zero of 30. Every single SERP we looked at carried multiple non-organic features stacked around and above the organic results — People Also Ask, video, image packs, an AI Overview, ads. “Position 1” still exists. It just no longer means what most decks imply it means.

This is a companion to our CTR-by-position study: that one showed rank and clicks have come apart on real data; this one shows a structural reason why. It’s the same discipline as measuring SEO like a psychometrician — before you trust a number, check that the thing it names still matches the thing it points at.

Method (reproducible)

• Data: Ahrefs keywords-explorer-overview serp_features for 30 keywords across nine niches — SEO/marketing, finance, health, software, home/DIY, travel, ecommerce, B2B, and general how-to. US, single snapshot pulled 2026-06-28. Same keyword set as our AI Overview study; this is the whole-page-composition cut of the same pull.
• What we counted: for each keyword, which SERP features Ahrefs flags as present — People Also Ask, video, image pack, AI Overview, news, paid ads, sitelinks, shopping, forum/discussion, local pack. We’re measuring presence, a binary per feature per SERP.
• Stat: prevalence — the share of the 30 SERPs carrying each feature.

The limits are real and stated up front, because they bound every claim below:

• n = 30, one snapshot, one country. SERP features are volatile; the same query an hour later, or from a different location, can render a different mix. This is a cross-section, not a longitudinal trend.
• Presence is not position. We report which features appear, not their measured screen height. We do not claim “the first organic result sits at pixel X.” Where we say a feature renders above the organic list, that’s how that feature type conventionally lays out — not a per-SERP pixel measurement.
• serp_features is Ahrefs’ reading of the page, not our own pixel-level scrape, and the AI Overview is counted once (Ahrefs’ paired ai_overview_sitelink flag is part of the same Overview, not a second feature).

None of that softens the headline, because the headline is a count of zero, and a sample of 30 is more than enough to notice that the “default” page never showed up once.

What’s actually on the page

Here is the prevalence table — the share of our 30 SERPs carrying each feature:

Read the top of that table as the new default furniture. People Also Ask and video are effectively universal — 97% each. Image packs, 90%. These aren’t garnishes that show up on commercial or media queries; they’re on nearly every page, across all nine niches. If your mental model of a SERP doesn’t include a People Also Ask block and a video carousel by default, it’s out of date.

Then the headline finding: zero of 30 SERPs were ten blue links. Not “rare.” Not “a minority.” None. The clean organic-only results page — the one every CTR forecast and every “just rank higher” pitch implicitly assumes — did not appear a single time in 30 tries. As a percentage of our sample, the canonical SERP is 0%.

77% carried an AI Overview, and half the SERPs (50%) carried paid ads. The AI Overview, by convention, renders above the organic list — so on a typical informational or commercial query, before a user reaches the first organic result they’ve already passed an AI-generated answer that’s trying to make the click unnecessary. (The ads in our sample were bottom-of-page blocks — they bracket the results rather than topping them, which is its own quiet tax on the page.)

It stacks. About 9 of 30 SERPs (~30%) carried an AI Overview and ads and People Also Ask all at once — three separate attention-grabbers on a single page. Call it the attention stack. On those pages, “organic #1” is one element competing among many for the click, not the headline the rank number quietly implies.

Put the whole table together and a typical SERP in our sample carried five or more distinct non-organic features (median around six, counting the AI Overview once). The ten blue links, where they survive at all, are now a minority region of the page — a column you scroll to, not the page you land on.

What this means: rank #1 isn’t the top of the page

Here’s the measurement point, and it’s the whole reason this study exists.

“Rank #1” is quietly equivocating. The phrase used to mean two things at once that were the same thing: top of the organic list and top of the page. When the page was ten blue links, those were identical, and the industry built its entire intuition — its forecasts, its pricing, its “we’ll get you to position 1” promises — on the assumption that they’d stay identical.

They didn’t. The metric’s name never changed. The thing it points at did. Today “position 1” means top of the organic list, and on most of our SERPs that list begins below an AI Overview and a People Also Ask block (with image and video panels often wedged in too). You can hold rank #1 and still be the third or fourth distinct thing a searcher encounters. Same label, different object — the kind of silent construct drift that, in psychometrics, quietly invalidates a measure while everyone keeps reporting it with a straight face.

This is the structural companion to what we found in the CTR-by-position study. There, rank and clicks had visibly decoupled on real data — positions 11–20 out-clicking 4–10, the smooth curve falling apart below the top three. This is one of the mechanisms behind that decoupling. If the SERP feeds the answer to the user inside an AI Overview, or satisfies them inside a People Also Ask expansion, or sends the click to a video, then the organic CTR attached to “position 1” is no longer a function of the rank — it’s a function of how much page the non-organic features ate first. Forecasting traffic from rank alone treats the page as if it were still ten blue links. Our sample says it never is.

What to do instead

You don’t need to panic about this. You need to stop reading one number as if it were a different number.

1. Stop reading rank as traffic. A ranking is a position in the organic list, not a guarantee of attention or clicks. The gap between “we rank #1” and “we get the clicks” is exactly the space the SERP features occupy. Track actual clicks and visibility, not position deltas multiplied by a borrowed curve — the CTR study shows what happens when you do the latter.
2. Check what owns the SERP before you value a keyword. Before you write the brief or price the project, look at the feature mix. A keyword whose SERP is an AI Overview plus ads plus a video carousel is worth less per organic position than its search volume implies, because organic is a thin slice of that page. Two keywords with identical volume and identical difficulty can be completely different opportunities once you account for the SERP features crowding the organic results.
3. On feature-heavy SERPs, the win may not be organic #1 at all. If the page is dominated by an AI Overview, People Also Ask, or video, then being cited inside those features can matter more than the top organic slot. The lever shifts from “rank the page” to “be the source the feature pulls from” — a different game with different inputs, and the one worth fighting on a page where organic is the fourth thing down.

A rank is a coordinate, not an outcome. Read the page it sits on, not the cell in your tool.

Frequently Asked Questions

Are the ten blue links really gone?

In our sample, yes. Across 30 keywords in nine niches, zero SERPs were a plain organic-results page — every one carried multiple SERP features like People Also Ask, video, image packs, or an AI Overview. It’s a single US snapshot, not a census, but a default page that appears 0 times in 30 is, for practical purposes, extinct.

What are the most common SERP features now?

In our 30-SERP sample, People Also Ask and video appeared on 97% of pages, image packs on 90%, and AI Overviews on 77%. Paid ads showed on 50%. These SERP features are the default page furniture across niches now, not occasional extras — most pages carried five or more distinct non-organic features at once.

Does ranking #1 still mean being at the top of the page?

No, and that’s the core finding. “Position 1” means top of the organic list, but on most SERPs that list begins below an AI Overview and a People Also Ask block. The label didn’t change; what it points at did. You can rank #1 organically and still be the third or fourth thing a searcher encounters on the page.

How should SERP features change how I value a keyword?

Treat search volume and difficulty as necessary but not sufficient. A keyword whose SERP features eat the top of the page leaves organic results a smaller share of available attention, so an organic position there is worth less than the raw volume suggests. Check the feature mix before committing budget — two same-volume keywords can be very different bets.

If you want this kind of analysis applied to your own keywords — what features actually own the SERPs you’re chasing, and whether ranking is even the right goal on them before you spend a quarter on it — that’s the work. Our fractional SEO practice reads the whole page, not just the position.

Method note: Ahrefs serp_features via keywords-explorer-overview, US, single snapshot 2026-06-28, n = 30 keywords across 9 niches. Figures are feature presence across the 30 SERPs — which features appear, not their measured screen position. The AI Overview is counted once.