Yes, but less than before and more on some searches. What Google says, what ranking studies show, how links relate to AI answers, and when links are your gap.
Yes, backlinks still matter in 2026, but what kind of links you have matters as well as how many. Google still lists PageRank, one of its link analysis systems, among its core ranking systems, and a large correlation study by Backlinko found that top results have more links than the pages below them. Google has also said links carry less weight than they did in its early years. They also count for more on some searches than others, and so far they line up only weakly with being named in AI answers. Here is what Google says, what the data shows, and how to decide what to do on your own site.
What Google says about links today
Google's guide to its ranking systems has a section on link analysis. It says Google has "various systems that understand how pages link to each other as a way to determine what pages are about and which might be most helpful in response to a query." On PageRank, it adds: "How PageRank works has evolved a lot since then, and it continues to be part of our core ranking systems."
Google has also been clear that links are one signal among many. In its November 2022 office hours, Duy Nguyen of the Google Search team said "backlinks as a signal has a lot less significant impact compared to when Google Search first started out many years ago", and that Google has "hundreds of them" when it comes to ranking signals. In the same answer, he said Google has "many algorithms capable of detecting unnatural links at scale and nullify them."
Put together: links still count, they count for less than they once did, and the ones built to game rankings may count for nothing.
What the ranking data shows
Backlinko's analysis of 11.8 million Google search results, built with Ahrefs data (page last updated Apr. 14, 2025), found that "the #1 result in Google has an average of 3.8x more backlinks than positions #2-#10." It also found that a site's overall link authority "strongly correlates with higher rankings." This is a correlation study. It shows that well-linked pages tend to rank higher, not that links alone put them there.
The same study found that "approximately 95% of all pages have zero backlinks." If your page has none, you have not yet tested whether links help it.
When links matter more, and when they matter less
A study from Ahrefs splits the picture by type of search. Patrick Stox and data scientist Xibeijia Guan took the top 1,000,000 US keywords by search volume and compared link counts with the top 20 results for each. The Ahrefs SERP study found a Spearman correlation of 0.255 for referring domains and 0.248 for backlinks. Ahrefs calls these weak correlations, and says that does not make links unimportant.
The split is more useful than the overall number:
- Search volume. Links lined up better with rankings on searches with higher volume. Stox gives two possible reasons: those searches are more competitive, and pages that rank for them are seen more, so they pick up links on their own.
- Local searches. Links and referring domains correlated "at .33 for local queries", which Stox says is a lot higher than what Ahrefs had seen before. Stox suspects that local service sites have very similar content, so links help Google tell them apart.
- Informational searches. Backlinks lined up more with rankings for informational queries than for commercial and transactional ones.
- Branded searches. Links correlated better here too, but Stox reads that as big brands simply having more links, not as links mattering more for brands.
- Over time. In a rough comparison with older Ahrefs data, links "correlate to rankings slightly less than they used to", which fits what Google has said.
Stox has also tested the question directly. He reports that when he disavowed all links to a few blogs, traffic and rankings fell. His view on what to do with the data is that "companies, especially local companies, probably should focus on at least foundational links", while links should not be your whole strategy.
How many links you need depends on your industry
There is no single number. WebFX's 2026 study of 1,462 domains across 15 industries found a median of 907 referring domains for page-one sites, but the median ran "from 76 in Apparel to 3,027 in Finance & Insurance." WebFX also found that on average "92.2% of backlinks pointing to top-ranking sites are editorial links", with directory links at 6.8%.
So compare yourself with the sites ranking in your own field, not with a number from a blog post. And look at what kind of links they have, not only how many.
What practitioners are doing
Budgets have not dried up. In Reporter Outreach's survey of 500 SEO professionals, 58% said they increased their link building budget in 2026 compared to 2025, and 62% said they prioritize quality over volume. Our recalculation of that survey's tactic data looks at which methods respondents rate highest compared with how many use them.
Do backlinks help you show up in AI answers?
The same survey found 74% believe backlinks affect AI search visibility, but only 24% are tracking it. Here is what is known so far.
Google says its AI features in Search, AI Overviews and AI Mode, have "no additional requirements" beyond normal SEO. To be shown as a supporting link there, a page "must be indexed and eligible to be shown in Google Search with a snippet". Google does not mention backlinks on that page, but it points site owners back to the same SEO basics that apply to normal results.
On AI mentions, brand mentions line up far more strongly than links do. An Ahrefs study of 75,000 brands found that branded web mentions correlated 0.664 with how often a brand appeared in AI Overviews, against 0.218 for the number of backlinks. Domain Rating came in at 0.326 and referring domains at 0.295. A later Ahrefs study across ChatGPT, AI Mode and AI Overviews again found "very weak correlations" between link metrics and brand mentions. Both are correlation studies, and Ahrefs says so.
Ahrefs counts "mentions of the brand name anywhere across the web" as branded web mentions, and in the same study branded anchors, where the brand name is the linked text, correlated 0.527 with AI Overview mentions. A link that names your brand registers on both measures. A bare link adds only to the backlink count. Our post on nofollow links and AI search covers one study of that relationship, and our four-week AI visibility log shows how to check whether your own brand gets named.
Should you keep investing in links?
A rule of thumb we use, depending on where you stand:
| Your situation | What to do |
|---|---|
| New page with no links in a competitive field | Earn a few relevant editorial links before judging the content. |
| Ranking on page one but below sites with far more referring domains | Links are a likely gap. Close it with editorial links, not volume. |
| Targeting local service searches | Links lined up with rankings more on local searches than overall in Ahrefs' data. Earn a few relevant local links before rewriting pages again. |
| Ranking well with fewer links than competitors | Keep the links you have and put most effort into content. |
| Most of your links came from directories, swaps or paid posts | Stop adding more of the same. Google's spam policies list low-quality directory links, excessive link exchanges, and buying links for ranking purposes without a sponsored or nofollow tag as link spam. |
Before you decide, check where you rank. SerpCrest's Keyword Rank Checker shows where a site ranks on Google right now for a set of keywords, by country, with the ranking page and who beats it. Compare those competitors' link profiles with yours, and you will have a clearer idea whether links are the gap.
