The 2026 link building statistics report: 180 sourced figures

This page is a collection of link building statistics on costs, link counts on ranking pages, time to effect, outreach reply rates and link retention. Each one is opened at its source and dated. It is built for anyone pricing a campaign or checking a figure before reusing it, and for anyone new to what link building is. Some figures come from large studies, others from one person tracking their own work, and the tables say which is which. The figures we could not trace to the page they are credited to are listed in their own section near the end, not repeated in the tables. Nothing on this page promises a ranking or a position. Each row gives the figure as its source prints it, the source and year, the sample it rests on, and the day we checked it.

What the ranking data shows about links

The largest studies find that top results hold more links, and links from more different sites, than the results below them. Backlinko's study, first published in 2016 and updated in April 2025, and Feed The Bot's September 2025 study both point the same way. Both are correlations, and they are the starting point for what backlinks do for a page.

What ranking pages hold

The headline correlation comes from Backlinko. In a study of 11.8 million Google search results, the #1 result in Google has an average of 3.8x more backlinks than positions #2-#10. Backlinko describes its work as a correlation study. The September 2025 study by Ihor Masalov at Feed The Bot covered 1,113 websites ranking in the top 10 for 200 commercial phrases. It found 96.3% of the sites that ranked in the top 10 have over 1,000 unique domain backlinks. "I was really hoping to find a unicorn of a site that had very few backlinks, yet ranked for commercial phrases. I have yet to find one," the author writes.

The odds with no links are long. In a December 2023 study of around 14 billion pages, updated in 2026, Ahrefs found that "96.55% of all pages in our index get zero traffic from Google". That is zero traffic, not zero links. On links, the same study puts the odds of real traffic with no referring domains at roughly 1 in every 6,671 pages. Murrough Foley scored 44,465 SERP results across 2,212 US keywords in a 2026 study: "Domain authority alone accounts for half the variation in rankings." Two limits apply. His "domain authority" is DataForSEO's domain-level score, not a link count, and his full model only explains 18% of what's going on. Digital Gratified's March 2026 study covered 420 SaaS sites, all already at DR 70+ with 20,000+ monthly organic traffic. It says nothing about small sites, and the page itself notes correlation is not causation.

The ranking figures, with their samples:

What practitioners report

In uSERP's State of Backlinks survey from September 2026, 62.1% still call backlinks a top-3 ranking factor. The survey covered 116 SEO professionals, most of them heads of SEO with 10+ years in the industry. And 87.1% would keep building links even if Google removed them as a ranking signal. Reporter Outreach's State of Link Building 2026 survey of 500 SEO pros, self-reported, found 58% increased their link building budget in 2026 compared to 2025. Aira's 2022 State of Link Building survey found 94% expect links to stay a ranking signal for the next five years. On paid links, the majority (69%) said that they believe they do positively influence rankings.

Ann Smarty put the long view in her September 2026 newsletter: "For decades, humanity hasn’t come up with better trust signals than citability and mentions." UK freelance SEO Dan Callis ran the test on his own site. He disavowed every link pointing to it in a 2020 test and wrote it up in August 2026. "My rankings (that drove little value and few conversions) dived off a cliff…"

The survey figures:

How many links ranking pages and sites hold

The honest answer to how many backlinks a page needs is that the median top-10 page holds far fewer than the average. In Semrush's Ranking Factors Study 2024, covering 16,298 English keywords, the top-10 average was 112 referring domains per URL against a median of 6. Backlinks ran 2,418 against a median of 13. The gap between average and median is the point: a few heavy pages pull the average up. Sakthivel Raju, a freelance SEO specialist who has been building links for clients for over 8 years, puts it this way: "Today 8–10 genuinely relevant placements will outperform 100 scraped ones, every time."

For local-service pages, Bogdan Krstic of Krstic SEO looked at 9,042 URLs ranking #1-#10 in their target markets. His pull ran over 20 days in May 2025, using the Moz index via Mangools. The average URL ranking on page one of Google in 2025 has 2,035 backlinks and 107 referring domains. "Backlinking is the most important aspect of local SEO. Over 90% of your time doing local SEO is likely going to be spent building backlinks," the local SEO expert writes.

WebFX's March 2026 study covered 1,462 page-one domains across 15 industries. It found a median of 907 referring domains, varying from 76 in Apparel to 3,027 in Finance & Insurance. The average link velocity was 48 new referring domains per month. And 92.2% of backlinks to top-ranking sites were classed as editorial. For SaaS link building, AutoBacklinks pulled the top-10 pages for 125 commercial SaaS keywords in July 2026. Medians ran from 23 referring domains at difficulty 0–10 to 80 at 11–30, 323 at 31–50 and 878 at 51–70. The brutal head terms needed 1,308.

Growth has a pace, too. Ahrefs's 2018 study covered 10,000 non-branded keywords. In it, most of the #1 ranking pages tend to get followed backlinks (from new referring domains) at a pace of between +5% and +14.5% per month. That is a rate of each page's own count, not a target. And one person can add a useful datapoint. Marko Tanaskovic wrote in January 2026: "I recently completed an exhaustive backlink analysis of 500 other websites, spanning five specific categories: news, e-commerce, education, tech, and health." His headline figure: "On average, the news websites I studied have 222,268,926 backlinks." He checked the sites in Ahrefs and kept only the 416 with a DA of 70 or more, so the figure is one person's pull of very large sites.

The link counts, with samples:

How long links take to move a ranking, and to be seen

Two clocks matter, and they are separate. Practitioner tracking puts most ranking movement between 2 weeks and 3 months. Moz measured 10 weeks per rank jump back in 2016. And one study of 16 million pages puts the average page at 27.4 days to be indexed. Our own page on how long link building takes to work covers the campaign view.

Time to a ranking move

Andy Zhao tracked 50 backlink campaigns across 5 different industries and published in April 2026 on his own site. Most backlinks take between 2 weeks and 3 months to show measurable ranking impact. The peak impact window is actually 4-8 weeks: 38% of effective links showed movement during this window. Only 6% of the links he tracked had zero impact after 6 months. One case he reports: "One niche edit I placed on a DR 65 blog’s existing article moved my client’s keyword from #22 to #9 in 11 days."

The oldest careful measurement is still the most cited. In 2016, Kristina Kledzik picked out 76 links pointing to pages similar in content and tracked them for Moz. The result: it takes 10 weeks on average to see 1 rank jump. Pages ranking on the second or deeper pages of SERPs took off after 8-9 weeks. "It seems that each link has a small to medium effect initially, but that effect increases over time," she wrote. Moz's own caveat: 76 links is a small sample.

Two practitioner surveys put the wait in months: the 2022 Authority Hacker survey of 755 link builders, kept in an archived copy after the live page began redirecting, and Aira's 2022 survey. Their figures are in the table below, and they are practitioner estimates, not promises.

Hamza Siddiqui wrote up a client case in September 2026 on his blog. A B2B SaaS comparison page stuck at position 34 for five months got six links over three weeks. The page did nothing visible for the first month. His timing rule: "I expect the first visible movement around week 6 to 10 after a decent link goes live." His warning: "If the client had pulled the budget at week four, they would have walked away with the conclusion that links do not work, and that would have been wrong."

The timing figures:

Time to be indexed

IndexCheckr's February 2025 study of 16 million pages found pages take an average of 27.4 days to be indexed. Within the first week, 14.00% were indexed; within the first 30 days, 64.86%; within the first six months, 93.2%. The pages were tracked by IndexCheckr users, and the averages cover pages that did get indexed. The study's own caveat: the time measured reflects only the period since a page was tracked in IndexCheckr. IndexVero's checker, across 737,787 URLs checked between 3 July and 15 September 2026, recorded 55.51% as indexed; the denominator is observations, not unique websites. Andy Zhao tracked 1,247 links for 6 months. Only 64% of all backlinks were indexed within 90 days, and the slowest link that did eventually get indexed took 137 days.

Google's Gary Illyes put typical timings on a conference slide, relayed by Barry Schwartz at Search Engine Roundtable in October 2026. Discovery of a new URL runs about 20 hours; indexing end to end runs about 1.5 hours. After the post went up, Illyes told Schwartz to "mind that this was an exercise to see if the audience can relate to the numbers we pulled internally and put in those slides."

Our own practice sits on this clock. We use two indexing services as standard for all clients, and between them we see roughly a 95% indexing rate. Because we place links on real sites owned by real people, the pages would most likely be indexed without a service. Left alone, though, that can take up to 4 weeks. That figure is about indexing time only, never about when a link moves a page. More on getting backlinks indexed on our site.

The indexing figures:

What links cost

The short answer: it depends who sells and where. BuzzStream's June 2026 data puts a guest post at $295 on average bought directly from a site and $461 through a vendor, while marketplace and publisher medians run from $30 to $570 for a sponsored article. For the breakdown by product and tier, see the full link building cost guide.

Measured prices: datasets and marketplaces

BuzzStream's guest post data comes from over 257k sites that appear on link vendor databases. A guest post costs $295 on average if you buy directly from a site, while the average price a vendor charges is $461. At the top end, a high-quality guest post costs $3,130 on average through a vendor, and a top-tier site costs $6,018. Our own prices sit inside the same market. A Rhino Rank Guest Post runs from $75 (DA10+) to $400 (DA60+) per placement, each a 750+ word article with one contextual dofollow link.

For insertions, the October 2026 update of BuzzStream's pricing data gives an average price for a link insertion of about $179. Only about 14% of the sites in the entire database offered link insertions. If the term is new to you, here is what a niche edit is. Our equivalent: a Rhino Rank Curated Link is "From just $60 per link", tiered by the placement site's referring domains.

Marketplace and publisher prices spread widely. ESBO Ltd's dataset, published on Ranktracker's blog in September 2026, covered 16,625 publishers across 53 languages and 33 categories. "The median publisher in our dataset charges $570 for a sponsored article," with the mean considerably higher at $1,233. Adsy's listings, February 2026 data across 52,671 websites: on average, you’d have to pay $459 for a guest post; the average price website owners list is $929 per backlink, but the average that people typically pay is $207 per link. SaaSLinks analysed 105,258 websites that sell links and found "the median backlink — a guest-post placement — costs about $140 in 2026". Adbassador's 808 publisher listings sit at the cheap end with a median guest post price of $30. iCopify puts marketing-niche placements at a $275 median. And when BacklinkGPT's AI agent pitched 510 sites for three of its own websites, "Sites we pitched mostly asked $80–250 per link."

The measured prices:

What surveys report people pay and budget

The $83 average in the table is just the price paid to a site for placing the link; it does not include the cost of the labour that goes into outreach or content. The same 2022 Authority Hacker survey of 755 professional link builders found in-house SEOs pay 75% more for links than niche site owners.

Editorial.Link's March 2026 survey of 518 SEO professionals found 91.89% of respondents believe their competitors buy links (476 yes, 42 no). That is a belief about competitors. The same survey calculated the average acceptable cost of acquiring one high-quality link at $508.95 from 412 responses, an opinion, not a price paid. Reporter Outreach's Q1 2026 survey of 500 SEO professionals, all self-reported, found 76% of buyers will pay $300 or more per link. Linkody, tracking more than 865,000 real backlinks for its users, notes that "About 1 in 10 of the links we monitor has a cost attached."

For wider SEO work, Ahrefs polled 439 SEO service providers in August 2024: freelancers $71.59/hour, agencies $98.90/hour, consultancies $171.18/hour on average. SE Ranking's December 2024 survey found 40% of agencies in the US and Canada charge over $125 per hour, against only 6% of European agencies. Both cover all SEO services, not links alone.

The survey figures:

What individuals paid

Frey Chu, who writes the Ship Your Directory newsletter, compared link quotes for his directory in August 2026 and picked one: "If you want my answer: the clear winner is the $75 one. DR 61, and real referring page traffic." Hamza Siddiqui logged a month of in-house building in September 2026. 42 hours and $1,849 produced 5 live links, so the real in-house link building cost was about $370 per link. "I have been building links since 2019, and the honest answer is that a single good link costs somewhere between $150 and $600 in 2026," he writes. Spencer Haws published the cost of a whole niche-site project at Niche Pursuits, a 2019 project last updated in March 2022. "I've spent about $1,850 total here," covering a VA's blog comments, an outreach specialist's guest posts, and links from services like The Hoth and Authority Builders.

Daniel Stanica spent his own money testing the market in August 2026: "I placed real, paid orders with 13 different link-building services and documented what each one actually delivered." His review is his own, not ours, and one line in it names us: "RhinoRank was one of the strongest experiences. I ordered a high-authority guest post, and the link landed within about ten days."

What three individuals paid:

Outreach: reply rates and what a campaign returns

Reply rates depend on the sample and the year. Hunter's State of Email Outreach 2026 report is built on 31 million emails sent by Hunter users in 2025. It puts the average cold email sequence reply rate at 4.5%. Hunter's digital PR / link building category sits at a 13% average response rate. The 2019 Backlinko and Pitchbox study of 12 million outreach emails found only 8.5% receive a response. Mentiohunt's October 2026 platform logs show 12.6% of prospects contacted reply: 259 of 2,056.

Reply rates by sample and year

The larger the sample, the lower the rate tends to be. The table below carries the rates by source and year, from platform logs and millions of emails down to single campaigns.

Single authors report more. Nisha Bharti, an SEO specialist, tracked one project in January 2026: "I tracked every outreach email sent during this project. The best-performing pitch had a 34% response rate." The project size is not stated. Janardan Das, a consultant, remembers the grind: "I was sending 200 emails a week, getting maybe three replies, and celebrating a 1.5% response rate like it was some kind of win." And Mia Elvasia at Blog Recode tracked every pitch for 12 months: 47 pitches, 23 responses (49%), 14 guest posts published, a 30% success rate. The page does not say how many of the 23 responses were yeses; 14 became published posts.

The reply rates:

What follow-ups and personalisation add

Follow-ups pay. Hunter's 2025 data: using three messages instead of one increases total replies by +106% (6.8% vs 3.3%), and sequences with no personalisation show an average reply rate of 3.6%, compared to 5.6% with two custom attributes in the email body. Smaller lists win: sequences with 21–50 recipients outperform by +158% compared to 500+ recipients (6.2% vs 2.4%), and contacting one or two people per company yields a +46% higher reply rate than emailing 3+ people per company (5.1% vs 3.5%).

The 2019 Backlinko and Pitchbox study tells a similar story on follow-ups: attempt 1 response rate 8.5%, attempts 1+2+3 17.1%, and 3 attempts to multiple contacts 22.15%. Note the tension: the 2019 study found multiple contacts per site helped, while Hunter's 2025 data found one or two contacts per company replied better than three or more. Different years, different samples; both are in the table. By topic, guest post emails drew 24.70%, roundups 23.80%, link building 21.90%, sponsorships 21.80%. A personalised subject line drew 21.8% against 16.7% without.

BuzzStream's May 2025 digital PR study found lower reply rates (0.59%) for the second email compared to the first follow-up (1.74%), so the second chase earns less. Its State of Digital PR 2026 report found 48% of respondents always personalise their email pitches to journalists. Authority Hacker's 2022 survey found 61.7% of link builders used social media for some of their outreach, and this group built 22% more links than those who didn’t. Authority Hacker's study of over 600,000 outreach emails, as reported by Ahrefs in 2020, found sending three follow-ups at least doubled their results. The primary could not be found, so we cite it as reported. Mentiohunt's logs put the first follow-up at +2.3 points on top of a 9.2% initial-email reply rate. "Shorter pitches worked better. My 3-sentence emails got more responses than my novel-length proposals," adds Mia Elvasia of Blog Recode.

What follow-ups and personalisation add:

What one campaign returns

Single campaigns return links in single or low double figures. Ahrefs's 2020 test sent 515 emails (473 delivered, 42 bounced) and won a grand total of 36 editorial links from 32 unique websites. That is a 5.71% conversion rate against a 17.55% reply rate. The team thinks it could set up the whole campaign in one working day. In 2022, Irina Maltseva, former head of marketing at Hunter, reported 96 new links from 54 domains in less than three months, with only one person involved in the process. "Sometimes, it took up to 30 minutes to write an email. Yes, it’s time consuming, but I found the response rate to be incredibly high (more than 50%)," she wrote on the Ahrefs blog.

Jakub Chutnak's campaign on the EmailListVerify blog (page last updated June 2023): 127 people emailed, 83% opened, about 20% replied, 10 backlinks in approximately one week. A Mailshake customer case study by Micheal Brennan carries no data year we can confirm. It reports 284 outreach emails in about one and a half hours, and 13 links in the 3 week period. The two follow-up sequences added 5 and 2 replies. Mia Elvasia's guest post year cost 156 hours in total, an average of 11 hours per successful guest post. And Mentiohunt's customers have marked 12 opportunities as won so far; the page stresses these are replies, not links.

What one campaign returns:

What prospects ask for, and who does the work

In Mentiohunt's October 2026 logs, 159 of 259 replying prospects asked for payment; only 20 proposed a link exchange, and 4 asked for a free product, licence or test access. The page's own caveat: many replies come from sites that sell links and quote a price to anyone who emails them.

BuzzStream's Link Building Trends Report was run in partnership with Citation Labs and is quoted on the BuzzStream blog (we read the quote, not the report itself). It found most organisations (61%) manage their link building entirely in-house. Weighing that against outsourcing link building is a separate decision, and we have a page on it.

Guest posts and digital PR

Guest posting was the most popular link building tactic for 64.9% of the 755 link builders in Authority Hacker's 2022 survey. In 2026, Editorial.Link's survey of 518 SEO experts found digital PR is seen as the most effective link-building tactic for 2026, chosen by 48.6% of respondents. Digital PR also costs more, at $597 per link on average across BuzzStream's 2025 survey.

Guest posts

Respona's June 2026 test reached out to exactly 1,000 guest posting sites, split over 4 campaigns, and got 205 replies, a reply rate just above 20%. The breakdown: sure, you can write a guest post for us (55/205); sure, but you have to pay us (137/205); no, absolutely not (13/205).

Mia Elvasia's year of pitches found industry blogs of 5K-50K monthly traffic harder to get accepted (25% success rate), while Medium publications had an easy acceptance rate (about 60%). The industry blogs gave her "quality backlinks that actually moved my SEO needle". Her plan for the next year: "Instead of 47 pitches, I’m targeting 12 high-quality opportunities. Better to land one post on a major blog than five posts on tiny ones." For the format comparison, see our page on guest posts vs niche edits.

Antonio Gabrić, outreach manager at Hunter, wrote on the Zapier blog in March 2024 about Hunter's own guest posting: "In 2022, we published 68 guest posts that resulted in 174 backlinks from DR70+ domains." One of the team's most successful campaigns: "I sent 102 emails and got 38% open rates, resulting in 26 guest post opportunities."

The guest post figures:

Digital PR and journalist platforms

BuzzStream's October 2026 pricing page puts digital PR link building pricing at $1,250-$1,500 per unique link, not including syndicated links. Typical campaigns cost $5-10k and net 6~7 unique linking root domains. Vince Nero's advice on the same page: "You should be willing to spend at least $1,500 to get a high-quality, non-syndicated, dofollow link in 2026."

BuzzStream's 2025 survey of about 70 digital PR agency leaders and freelancers, updated June 2026, puts the average monthly contract at approximately $5,458. The average cost per link for digital PR across all respondents is $597. Agencies charge about 50% more than freelancers ($6,357 vs. $4,200 monthly). Most US respondents placed their cost per link in the $0-$500 range, whereas most UK respondents placed theirs in the $500+ range; 54% of responses came from the UK. By service, costs range from $150 for guest posting to $800 for large hero campaigns; the page calls these category-weighted midpoints, not necessarily mathematical averages. BuzzStream's State of Digital PR 2026 report adds that the share of teams reporting cost per link of $750+ has surged from 3% to 10.2% in 2026. And 60.8% of digital PR professionals report that finding relevant journalists has become more challenging in the last 12 months. In uSERP's September 2026 survey, 53.4% say digital PR delivered their strongest results in the last 12 months.

Featured, the owner of HARO, released its own figures in September 2026. Human-written pitches are about 3x more likely to get featured than AI-written ones (no sample is given for the 3x). 32% of journalist requests are rejected, and 35% of journalists block AI-generated pitches. More on HARO backlinks on our site. Linkifi's June 2026 analysis of 2,366 winning expert quote pitches found HARO, Connectively, and SOS accounted for ~73% of link placements. The sample is winning pitches only, so these are shares of placed links, not of pitches.

The digital PR figures:

How many links are lost, and how fast

Links decay, and the two best measurements sit years apart. In Linkody's June 2026 pull of more than 865,000 real backlinks that its users actively track, 89% are still live after 3 months and around 63% after a year. A later account split, added August 2026, covered 139 accounts and 234,138 links from accounts that added at least 30 links in 2025. In that split, the median account lost 1.8%. Ahrefs's web-wide study, updated in February 2024, found 66.5% of the links pointing to 2,062,173 sampled websites have rotted since January 2013.

The Ahrefs study covers nine years: 74.5% of the links in the study are considered lost, with the causes listed as dropped (47.7%) and link removed (34.2%). Linkody's 2026 figures carry a caveat: the clock starts when a link enters Linkody, not when it was first published. Loss is uneven: in the same 139-account split, 73% of accounts lost less than 10% of their links. Across all tracked links, half of all dying links do not survive twelve months, and the average lifespan to death is about 20 months. Host quality shows: sites under DA 10 kept only 33% of links live, against about 70% for DA 30 to 49. Linkody's older 2022 case study drew a separate survival curve: 82.36% still live after 1 year, 56.61% after 7 years.

Our own term sits here: every delivered link carries 12-month replacement cover, replaced like for like if the link is removed or its page is taken down. When links go, reclaiming lost links is the tidy-up job.

What we see in QA: we open every placement after it goes live, and again later. We check that the page is indexed, that the link is still in the body and still followed, and that the host site still has readers. Since 2017, we have built over 120,000+ links annually for 7,000+ clients, and that re-check is part of every order.

The loss and retention figures:

Anchors, nofollow and links per page

The nofollow attribute sat on 32.7% of pages in 2024, up from 29.5% of pages in 2022, per the HTTP Archive Web Almanac's SEO chapter. Sponsored and ugc markup is rare: in 2024, it was just 0.4% for sponsored and 0.3% for ugc. Those are usage shares across the web, nothing more.

On anchors, SEO Arcade's September 2026 study scored 303 ranking pages from 46 complete search results against their own backlink anchors. Jeremy Rivera writes: "Exact-phrase anchors: nothing. The anchor that reproduces your keyword word-for-word showed no relationship with position." And: "How many different sites link to you still matters more than any of it." It is a cross-section, not an experiment.

Ahrefs's 2020 study covered the top 110k sites by Ahrefs Rank. It found 10.6% of all backlinks to those sites are nofollow, with rel="ugc" on only 0.44% of referring domains and rel="sponsored" on 0.01%. Digital Gratified's March 2026 SaaS study found roughly 17 to 18% of the referring domains pointing to successful SaaS sites are nofollow. Our page on dofollow and nofollow links explains the attributes. And how many links does a page carry at all? The Web Almanac 2025, published January 2026, puts the median page at just six outbound links across both desktop and mobile. Internal links dominate: a median of 43 on desktop and 39 on mobile.

The anchor and attribute figures:

Links and AI answers

Better-linked pages get cited more often in AI answers. In SE Ranking's November 2025 study of 129,000 unique domains and 216,524 pages across 20 niches, websites with up to 2,500 referring domains receive an average of 1.6-1.8 ChatGPT citations. Those with over 350,000 referring domains receive 8.4 citations on average. In Link Builders' October 2026 study of 1,329 AI Overview citations across 208 search queries, the median cited page had 123 referring domains. The median page-one result that was not cited had 28. "Links seem to help twice: once by lifting the page up the results, and again by making it more likely to be cited once it is there," writes founder Justin Davis. The study covers one niche and one country, and in the page's words it shows a link, not a cause.

Ahrefs measured the overlap in March 2026 across 863K keyword SERPs and 4M AI Overview URLs. 37.9% of URLs cited in AI Overviews also appeared within the first 10 blocks, down from about 76% in July 2025. ZipTie's study covered 25,000 real user queries, published May 2025 and updated September 2026. It found that if your website ranks #1 in Google's traditional results (blue links), you have a 25% chance of being used as a source in AI Overviews. And Pew Research Center, July 2025, analysing 68,879 unique Google searches from March 2025: the vast majority of these summaries (88%) cited three or more sources.

Practitioners are acting on it: in uSERP's September 2026 survey, 81.9% believe backlinks influence whether AI engines cite their brand (a belief, not a measurement). The rest of the survey's AI rows, and Reporter Outreach's 2026 figure, sit in the table below.

Amanda Natividad captured the mood in her September 2026 newsletter: "The irony is not lost on me. In this increasingly clickless world, we still desperately want those links." And Wil Reynolds, as printed in Amanda Natividad's newsletter: "A backlink to your site, your research, your content, whatever it is, means somebody thought it was worth linking to."

The AI citation figures:

Content that earns links

Most content earns nothing. Backlinko's 2019 study of 912 million blog posts measured how little earns links and which formats earn more, and the table below carries the four figures.

Statistics pages themselves earn links. Sunny Patel, an SEO Consultant & AI Strategist, counted on his own site in September 2026: "On sunnypatel.co.uk, five stats and data pages hold 31 unique referring domains between them, checked on 24 September 2026. … Page age and promotion were not controlled for." His other 21 articles, service pages, landing pages and portfolio pages hold 24 referring domains combined.

That is why this page lists its sources the way it does.

The content figures:

Which metrics link builders use

Aira's 2022 survey asked link builders which metrics they use to measure the authority and/or quality of a link. The most popular answer with 67% of the votes was Domain Rating by Ahrefs, followed by Domain Authority from Moz at 42%. Asked what they measure, the clear winner with 53% was rankings, followed by search visibility at 36%.

Our own data says the two scores sit closer than people assume. In the July 2026 sample from our free DA/DR checker, published with our Domain Authority vs Domain Rating write-up, the tool had logged 29,632 checks since it launched in February 2026. That is 15,357 unique domains, latest check per domain. Mean DA sat at 27.9 and mean DR at 28.6, with a median DR-minus-DA gap of -1. DR was above DA on 42.1% of domains, at least 20 points above on 12.5%, and at 50 or more with DA under 25 on 2.1%.

Two practitioners put the reader-first view. Dasara Kushi, May 2026: "A backlink from a respected sneaker enthusiast newsletter beats a backlink from NYT for a sneaker valuation site — even though NYT has 10× the domain authority." And Frey Chu of the Ship Your Directory newsletter lets one question guide his spend: "Will a real person see this link, click it, and land on my directory because it's genuinely relevant to them?" … "That's the one question that I let guide my link building spend."

How we use the pair: DA sets the tier and price for our Guest Posts, and DR and live traffic tell us whether the number is earned. Our Curated Links are chosen on relevance, referring domains, DA, organic traffic and overall site quality.

Figures we left out

The ranking pages repeat a set of figures we could not trace. Each is either not on the page it is credited to, or credited to a primary we did not open. It is listed here and kept out of the tables above. The pages repeating the first three credit them to Authority Hacker's survey; the archived survey text does not contain them.

The figures as repeated, and why each is not here:

The mix-up runs the other way too. The Ahrefs figure in our first table is about traffic, not links: 96.55% of all pages in Ahrefs's index get zero traffic from Google. As the SEOmator blog put it in August 2026: "You will see "96.55% of all web pages have zero external backlinks" in dozens of 2026 roundups."

FAQ

How many backlinks do I need to rank?

The median top-10 page in Semrush's Ranking Factors Study 2024 had 6 referring domains, against an average of 112. A few heavy pages pull the average up, so the median is the honest benchmark. Whole sites carry more: WebFX measured medians for page-one sites from 76 in Apparel to 3,027 in Finance & Insurance.

How long do backlinks take to work?

Most movement lands between 2 weeks and 3 months, in Andy Zhao's 50 tracked campaigns (April 2026), with the peak window at 4-8 weeks. The 2022 Authority Hacker survey of 755 link builders put the average at 3.1 months, and Moz's 2016 measurement was 10 weeks per rank jump.

How much does a backlink cost in 2026?

A guest post costs $295 on average bought directly from a site and $461 through a vendor, in BuzzStream's June 2026 data. SaaSLinks put the listed median at about $140 across 105,258 websites that sell links, and BuzzStream's October 2026 data puts a link insertion at about $179.

What reply rate should I expect from link outreach?

Expect 4.5% across cold email generally and 13% in Hunter's digital PR / link building category, from 31 million emails sent in 2025. The 2019 Backlinko and Pitchbox study of 12 million outreach emails found 8.5% got a response. Small, relevant, personal campaigns report far higher rates.

How many backlinks are lost each year?

Around 63% of the 865,000+ backlinks Linkody tracked are still live after a year (June 2026). In a later split of 139 accounts that added links in 2025, the median account lost 1.8%, and 73% of accounts lost less than 10%. Ahrefs's study, updated February 2024, found 66.5% of links rotted since 2013.

Do backlinks still matter for AI search?

The data says better-linked pages get cited more. In SE Ranking's November 2025 study, domains with over 350,000 referring domains received 8.4 ChatGPT citations on average, against 1.6-1.8 up to 2,500. Link Builders' October 2026 study found cited AI Overview pages had a median of 123 referring domains, against 28.

Do most SEOs pay for links?

Most do: 74.3% of link builders said they pay for links in Authority Hacker's 2022 survey of 755. What we manage on every order is publisher quality: real readers and real traffic, DR and DA cross-checked, the link in the body of the article, and a check after it goes live.

Which metric should I use to judge a link, DA or DR?

Both measure link profile strength. Aira's 2022 survey found 67% of link builders use Ahrefs DR and 42% use Moz DA. In our own checker's July 2026 sample of 15,357 domains, the two sat close: mean DA 27.9, mean DR 28.6. DA sets our Guest Post tier; DR and live traffic are cross-checked.