How Many Backlinks Do You Actually Need to Rank?

When it comes to ranking your website on search engines, one of the most common questions is, “How many backlinks do I need to rank on page one?”

The short answer? It depends — but not in the hand-waving way that phrase usually means. The number is calculable, the inputs are public, and the method we use to work it out takes about ten minutes.

One correction before we start, because the two are not the same count: the thing to count is referring domains — how many separate websites link to a page — not backlinks. The question gets asked in backlinks. It gets answered in domains.

Here’s how to work out yours, with real numbers from our own campaigns as calibration.

The Short Answer

Your number is about one page, not your whole site — and for most keywords it’s smaller than people expect.

Ahrefs publishes the closest thing to a public answer that exists, because its keyword difficulty score is built from how many websites link to the pages already sitting in the top 10. On an easy keyword that’s a handful of sites. At the top of the market it’s thousands. For a keyword at difficulty 6, Ahrefs’ Keywords Explorer¹ estimates roughly 7 referring domains to reach the top 10.

So the honest answer isn’t a single figure. It’s whatever closes the distance between the pages already ranking for your keyword and where your page stands today — and that distance is measurable.

Our own case studies show how wide the spread runs. The smallest campaign we’ve published a result for was 10 placements. The largest reached 80+, spread across three fashion brands for one client. Both ended in 4x traffic or better. In our experience the difference is rarely budget — it’s how large a gap the client started with.

Just ask our lifestyle affiliate client who saw a 37,000% organic traffic increase from just 10 placements. Not a typo. Strategic placement beats mass quantity.

Here’s how to work out yours.

Referring Domains, Not Backlinks — and Why It Changes the Number

One website that links to you from 200 pages — a sidebar, a footer, a template — is 200 backlinks and one referring domain. It’s one vote, cast once. Count backlinks and you’ll conclude you need hundreds where you need a dozen, and buy accordingly.

The vendors already agree with the correction, inside their own products. Ahrefs describes how it scores keyword difficulty in one sentence: “We pull the top 10 ranking pages for your keyword and look up how many websites link to each of them.” Websites, not links. Semrush² weighs the same unit — the median referring domains of the ranking URLs — alongside link quality, the authority of the ranking domains and the make-up of the results page. So both build difficulty from the referring domains of the pages already ranking: Ahrefs from that count alone, Semrush as one of four inputs. (Ahrefs then publishes its difficulty calibration in backlinks rather than domains — a unit trap we come back to below.)

The same correction applies one level up, to the domain scores everyone quotes. In Ahrefs’ analysis of a million U.S. SERPs, published January 2025, referring domains correlated with ranking position at 0.255.³ Domain Rating managed 0.131 — roughly half as much.

Ahrefs⁴ is blunter about its own metric than we would be: “DR is not a Google ranking factor. Google doesn’t use Ahrefs’ Domain Rating.” It’s a third-party estimate, and Ahrefs positions it as a tool for choosing which sites to build links from.⁵ That’s exactly how we use it — a sourcing filter, never a ranking promise.

How to Work Out Your Number

This is the method we use before we quote a campaign. It borrows the vendors’ unit — the referring domains of the pages already ranking — but the rest of it is ours: the vendors publish neither the position range below nor the exclusion of the top three.

Do this per keyword, not per site. Your gap on a long-tail term may be nearly closed while your gap on the head term is a year of work, and you only find that out by measuring both.

1. Take one keyword and pull the top 10 organic results. Ignore the ads. Ignore any result that isn’t the kind of page you’d publish.

2. Read the referring domains for positions 4 through 10. Referring domains, not backlinks. We skip the top three on purpose: in our experience those slots are usually held by incumbents and household names that would rank on recognition alone, and including them drags the bar above what the keyword actually costs to enter.

3. Take the median of those seven numbers. That’s the entry bar — roughly what it takes to be in the conversation on that keyword.

4. Subtract the referring domains already pointing at your page. At your page, not your whole site. The difference is your number.

One honest caveat about the method above: it treats every referring domain as one vote, which is the right simplification for a calculation you can run yourself in ten minutes. When we quote a campaign we weight each domain by its authority instead, and benchmark against the median competitor rather than the strongest — ten DR 70 domains and ten DR 20 domains are not the same gap, and one giant in the results shouldn’t set the bar. The count gets you the estimate. The weighting is what changes the plan.

5. Add headroom, and re-measure. The bar moves while you build, because the sites you’re catching are still building too. Treat the first figure as a reading, not a finish line. We re-run the measurement quarterly on the campaigns we manage — it’s the only way to know whether you’re closing the gap or holding station.

The ten-second version

If you’d rather not open a backlink tool at all, Ahrefs’ keyword difficulty score compresses the same idea into one number, built from the referring domains of the top 10. Watch the unit when you read its published calibration, though: Ahrefs states these bands in backlinks, not in domains, even though the score itself is computed from referring domains. Its calibration, updated December 2025:

For a keyword at difficulty 6 — Ahrefs’ worked example is “best womens hiking boots” — Keywords Explorer estimates roughly 7 referring domains to reach the top 10.

One caveat, and it’s Ahrefs’ own: keyword difficulty “is an absolute metric — not a relative one”, scored from the backlink profiles of the top-ranking pages, so “it doesn’t account for your specific website’s authority.” It tells you where the bar is. Only step 4 tells you how far you have to jump.

Links to the Page vs Links to the Site

“How many backlinks do I need” is almost always “how many does this page need to rank for this keyword”. Those are different questions with different answers, and mixing them is the quickest way to overspend: a site’s total is everything it has accumulated across every page over its whole life, while your page needs only enough to clear one keyword’s bar.

Links to the page you want to rank do the ranking work for that page. Links to your domain raise the floor under every page on it, which is why — in our experience — a new site has further to travel than an established one to reach the same position.

Internal linking is the bridge between the two. Linking from your strongest existing pages to the page you’re trying to rank moves authority you’ve already paid for onto the page that needs it. It’s the one kind of link building with no placement cost at all, and one reason the honest answer to “how many” is usually smaller than people fear.

Measure at page level. Ahrefs says plainly that its domain-level score doesn’t predict rankings as well as its page-level one.

What Moves Your Number Up or Down

Links are still a live ranking signal, and Google says so. Its ranking-systems guide⁶ lists PageRank among the active core systems — Penguin, the standalone link-spam system, was folded into core ranking in 2016 and now sits in the retired list — and Cherry Prommawin told Google Search Central Live APAC 2025 that “links are still an important part of the internet … and we use them for ranking”, as reported by Search Engine Journal in July 2025.⁷

But Google’s own language has cooled. When it revised its link-spam policy on 25 September 2024⁸, the sentence that had said Google “uses links as a factor in determining the relevancy of web pages” did not survive⁹; today the policy defines only the abuse. The documentation that still describes the signal¹⁰ calls links a signal, not a factor.

One live input among many, then. Three things move how much of it you need.

Your existing link profile

If your site is brand new, your threshold is higher. You’re building authority from scratch, and you’ll need more referring domains than an established competitor to reach the same position.

Closing that gap is what we’re hired for, and our floors are published: DR 30+ on Listicle Publishing, DR 20+ on Outreach Link Building. The measured average sits well above both. As of August 2026, the 1,068 live placements we could match to a current domain rating average DR 58 — 71% at DR 50+, 34% at DR 70+, 99% at DR 20+. One honest caveat: those are the ratings those domains carry today, not the ratings they carried the day each link went live.

How competitive the keyword is

Head terms and long-tail terms are different jobs. Ahrefs is blunt that there’s “no set number of backlinks” and no sweet spot, and in guidance last updated January 2025¹¹ its own example for the head term “tea” estimates roughly 756 unique backlinks needed to reach the top 10 — an estimate of what it would take, not a measurement of what the incumbents have, and note the unit again: this is one of the places a vendor quotes backlinks rather than domains. A long-tail term in the same niche can sit at single-digit referring domains. That’s why every fully managed campaign we run starts with a competitor gap analysis.

One caution about the benchmarks quoted on this question: they’re often older than the pages quoting them. Backlinko’s widely cited “3.8x more backlinks” page¹² carries a 2025 update stamp but never states when its 11.8-million-result dataset was collected, and never claims it was re-analysed. Read the data’s age, not the stamp.

The quality of what you build

Not all links are equal, and the difference isn’t a famous logo. We vet for niche relevance and real organic traffic, and given the choice we’ll take a placement on a site your buyers actually read over a bigger name with no topical connection to your subject. That’s our sourcing judgement, not a measured law — but it’s the one we build on.

When More Links Won’t Help

Sometimes the answer to “how many more backlinks do I need” is none, and buying them — from us or anyone — is the wrong move. Four cases where the number isn’t your problem:

The page isn’t eligible. If it isn’t indexed, no quantity of links will rank it. Check coverage before you check competitors.

The intent doesn’t match. If the top 10 are all comparison articles and yours is a product page, you’re not short of links — you’re the wrong shape. Keyword difficulty won’t warn you about this: Ahrefs¹³ states that KD takes no account of on-page factors at all.

The page is too young. Ahrefs’ May 2025 study¹⁴ found 72.9% of pages in Google’s top 10 are more than three years old, and only 1.74% of newly published pages get there within a year. Links accelerate that clock. They don’t skip it.

There’s no topical connection. Google describes links as a signal it uses “when determining the relevancy of pages” — so a site with nothing to do with your subject is a poor fit for that job. Relevance is the first filter we apply when we source, for the same reason.

Fix those first. Then count domains.

Real Campaign Results: What Does It Actually Take?

We count placements — one editorial link on one publisher — rather than backlinks. A note on that last row, because the number on our case studies page is bigger: 30+ is what we tracked to the client’s main ecommerce store, and that store is what 4x’d. The wider engagement covered three of their fashion brands and 80+ placements in total. Ten placements at one end of that range, then, and 80+ at the other. Both were wins. The difference was the size of the gap each one started with. Explore our full case studies to see how.

How Many Per Month, and Is There a Safe Number?

On safety. Google’s spam policies¹⁵ define link spam by intent — links created primarily to manipulate rankings — not by rate. There is no published per-month limit and no published rate threshold of any kind. Our own read of that: what draws scrutiny is the pattern, not the pace — a hundred links a month from sites that exist to sell links is a problem at any speed, and ten editorial placements a month on real publications is not.

On paid placements. Google’s spam policies address this directly: buying and selling links is a normal part of the web’s advertising economy, and such links are not a policy violation when they’re qualified with a nofollow or sponsored link attribute. We mark placements that way for any client who asks us to.

On pace. Fast enough to close your gap inside a horizon you can live with. These are the cadences we actually run:

Most of our clients see measurable ranking improvements within 3–4 months, with early signals — better keyword positions, rising impressions — often in the first 6–8 weeks. We go deeper on that in how long link building takes to drive results.

And no, we don’t do PBNs or guest post farms. Placements go on vetted, niche-relevant sites with real organic traffic. That’s a vetting standard we apply by hand rather than a hard gate in software, which is one reason the measured figure above is 99% at DR 20+ rather than 100% — the other being that those are today’s ratings, not the ratings at placement.

What This Costs

Once you have a number, the budget is arithmetic. Editorial placements start at $250 each, so a twelve-domain gap is a few thousand dollars and a two-hundred-domain gap is a five-figure programme. That’s a useful thing to learn before you commit rather than after.

It’s also the argument for measuring first. The campaign that costs least is the one aimed at the keyword where your gap is a dozen domains instead of two hundred. See what a placement costs.

Does This Still Matter Now That AI Answers the Question?

Worth knowing before you set a link budget: the ranking is no longer the end of the chain.

In March 2025, Pew Research Center¹⁶ tracked 68,879 Google searches across the browsing data of 900 U.S. adults. On visits to a results page that carried an AI summary, people clicked a traditional search result 8% of the time. On visits without a summary, they clicked 15% of the time — nearly twice as often. And the summaries aren’t evenly spread: 60% of queries beginning with a question word (such as who, what, when or why) returned one.

So a first-page ranking is still worth having, but visits to a results page carrying a summary produce fewer clicks on traditional results. That changes the budget maths. It doesn’t change the method.

The second surface behaves differently again. Ahrefs’ study of 75,000 brands (May 2025)¹⁷ found branded web mentions correlated with AI Overview visibility at 0.664 — well ahead of Domain Rating at 0.326, referring domains at 0.295 and backlink counts at 0.218. Note that the ordering of the link metrics flips on that dataset: DR edges out referring domains there, the reverse of the classic-SERP study earlier in this post. Correlations, not causes — Ahrefs says so, and so do we.

There’s a split worth understanding here, because the assistants are not one thing. Google is explicit that its AI answers aren’t a separate machine: its own guidance¹⁸ describes grounding as “relying on our core Search ranking systems to retrieve relevant, up-to-date web pages from our Search index” — the same systems in which links from other prominent websites are a documented quality factor¹⁹. The others are a black box by comparison. OpenAI²⁰ documents a search crawler and states that blocking it keeps a site out of ChatGPT’s search answers, but publishes nothing about how it ranks what it finds. Perplexity²¹ documents a first-party index of “over 200 billion unique URLs” with its own ranking pipeline, without naming its signals. Anthropic²² documents the mechanics of Claude’s web search but not the index behind it. So the honest position isn’t that links stopped counting in AI assistants — it’s that only Google has said.

The practical read is that being cited by an assistant looks closer to being mentioned than to being linked. A placement that carries your brand name on a page an assistant reads is doing two jobs, not one.

That’s a pointer, not the map. AI visibility deserves its own post. This one is about the number.

Summary

Your number is your gap. Measure it per keyword, at page level, in referring domains: take the median referring domains of positions 4 through 10, subtract what your page already has, and build the difference.

On low-competition terms that’s often a dozen domains. On head terms it’s hundreds, and at the top of the market it’s more than that. The figure changes as you build, so treat the first reading as a reading rather than a target — we re-measure quarterly.

And before you buy anything, check that links are the constraint at all.

How PressHERO Can Help

Two services do most of the work of closing a gap.

• Listicle Publishing — from $250 per placement, live within 14 days, DR 30+ guaranteed. Your brand featured in the “best of” lists your buyers and AI assistants read.
• Outreach Link Building — from $250 per placement, live within 28 days, minimum DR 20 and averaging DR 50+. Precision targeting on vetted, niche-relevant publishers.

Every placement carries a 12-month guarantee: if a link comes down inside that window, we replace it free. No PBNs, no guest post farms, no long-term contracts.

TL;DR — It’s Not a Numbers Game, It’s a Strategy Game

You don’t need a universal number. You need enough to close your gap — and that number is knowable before you spend anything. Measure it per keyword, at page level, in referring domains.

You don’t need us to run the measurement — the method is above, and it’s yours to use.

Just ask the fashion ecommerce brand that reversed their traffic decline with 30+ DR 70+ placements to their main store — part of an 80+ placement programme across three of their brands — and 4x’d their traffic in 12 months.

What we sell is closing the gap it finds: listicle placements from $250, live within 14 days, or outreach link building on vetted publishers averaging DR 50+. Both carry the 12-month replacement guarantee.