How Many Backlinks Do You Actually Need?

Ask five agencies how many backlinks you need and you'll get five different numbers, most of them matching whatever package that agency happens to sell. There's a real figure for your site, and you can work it out yourself in about twenty minutes using the same competitor data we run before quoting a campaign. Below is that method, step by step, along with how quickly to build once you've got your number, why copying a competitor's Domain Rating gives you the wrong answer, and the situations where the honest figure is zero.

How many backlinks do I need to rank?

There's no fixed number.

What you need depends on the search results page you're targeting, so anyone quoting a figure before they've looked at your SERP is guessing.

Four things decide it.

The first is the page-level link profile of the results currently ranking for your query, meaning the links pointing at those individual URLs rather than at their domains. The second is how relevant those links are, because ten links from sites that genuinely cover your subject will usually do more than a hundred from sites that cover everything. The third is how your page compares with the ones it's up against, since links work hardest on a page that already answers the query well. The fourth is what you've got already, because the gap is what you're buying, not the total.

Google has been consistent on this for years.

John Mueller has said the total number of backlinks is "completely irrelevant" to how its systems assess a site, and has separately advised against focusing on the absolute count, partly because no tool can measure it objectively in the first place. Two links can carry very different weight, so counting them tells you little.

Links do still matter.

Backlinko's analysis of 11.8 million search results found that the number of distinct domains linking to a page was the strongest correlating factor of everything it examined, and that the page in position one had on average 3.8 times more backlinks than positions two to ten. So the useful question isn't whether you need links. It's how many more you need than you've got, for the specific pages you want to rank.

That's a piece of analysis rather than a guess, and it's the first thing we do before quoting anyone for our link building services. The method is below.

How do I work out my own number?

Run a gap analysis against the pages actually ranking for your target query, which frequently aren't the businesses you think of as your competitors. Here's the process.

Start with one query and one page. Pick the commercial query that matters and the single URL on your site you want ranking for it. Not the topic, not the homepage. Everything that follows is page-level.

1. Capture the real SERP. Search the query in a private window, using the location and Google domain your customers use. Google.co.uk returns a materially different top ten to google.com. List the ten organic URLs, ignoring the ads and the AI Overview, because you're measuring what it takes to rank organically.

2. Pull page-level data for each of the ten. In Ahrefs' Site Explorer, paste each URL and switch the mode from "Domain" to "Exact URL". Record four numbers per result, being referring domains to that URL, URL Rating, estimated organic traffic to that URL, and referring domains to the whole domain. Semrush's Backlink Analytics does the same job with the URL scope, as does Majestic. Ahrefs Webmaster Tools is free for domains you've verified in Search Console, so you can get your own page's figures at no cost before deciding whether a subscription is worth it.

3. Read the spread, not the average. Sort the ten by referring domains to the URL. The distribution is almost always heavily skewed, which makes an average close to meaningless. As an illustration, take a SERP where the ten pages have 3, 5, 8, 11, 14, 17, 24, 46, 88 and 240 referring domains. The average is 45.6, which describes none of them. What you want is the floor and the middle. The floor, 3 in this case, is the entry ticket, the least any page got away with. The middle, around 14 to 17, is what a page needs to sit there comfortably rather than clinging on.

4. Strip out the results you can't copy. Wikipedia, government sites, major news brands and category-dominant marketplaces rank on domain strength and can hold a position with almost no links to the page itself. Exclude them. If four of the ten are untouchable and the remaining six sit between 8 and 20 referring domains, that range is your target.

5. Subtract what you already have. Pull the same page-level figure for your own URL. If the ranking pages you could realistically displace sit at 8 to 20 and your page has 2, your number for this query is roughly 10 to 20 additional referring domains, plus a margin, because the count is only half of it and relevance is the other half.

6. Repeat for three to five queries and take the highest. Pages rank for clusters, not single terms. Running the analysis across the handful of queries you actually care about stops you optimising for the easiest one and being disappointed by the rest.

To speed up the prospecting, Ahrefs' Link Intersect and Semrush's Backlink Gap tool both take several competitor URLs and return the domains linking to them but not to you. That list is your outreach shortlist, and it's far more useful than a raw count, because it tells you which specific publishers already consider your subject worth covering.

For the wider framework this analysis sits inside, our guide on how to plan a link building campaign that supports long-term SEO growth covers turning these numbers into a programme with objectives attached.

Why is matching a competitor's DR a lazy way to do this?

Domain Rating describes a whole site, and the thing you're trying to rank is one page on it. Aiming at a competitor's DR sets you a target that doesn't measure the job.

Three problems come with using it. It isn't a Google metric, since DR belongs to Ahrefs, Domain Authority to Moz and Trust Flow to Majestic, and they routinely disagree about the same site because they're built on different crawls and different maths. It's logarithmic, so the distance from DR 20 to DR 30 is nothing like the distance from DR 70 to DR 80, which leaves "we need to hit DR 50" as a target with no consistent meaning. And it's an average across an entire domain, so a DR 80 site can and does publish pages with no links that rank for nothing.

There's evidence pointing the other way, and it's worth knowing about. Backlinko's study found domain-level authority correlated more strongly with first-page positions than page-level URL Rating did. That finding is real, though correlation is doing a lot of work in it. Strong domains have more of everything, including more content, more brand searches, better internal linking and better technical foundations, so they rank well for reasons a domain score only picks up second-hand.

Look at referring domains to the specific URL instead, along with the URL Rating or Page Authority of that page, and how relevant the linking sites are to your subject. A DR 35 site that writes about your topic every week, has real organic traffic and links out sparingly is a better prospect than a DR 75 general-interest site running forty sponsored posts a month. Check the linking page's own traffic before you decide what the link is worth.

How fast should I be building them?

There's no published limit, so pace the campaign to what your outreach can genuinely produce rather than to a number someone has quoted you. Any agency offering you a safe links-per-month ceiling has invented it.

It's worth being precise about the risk.

Google has never documented a velocity threshold, and its link spam policies don't describe one. What those policies describe is the character of the links, meaning links exchanged at scale, paid for without disclosure, or placed on sites with no editorial standards. Since the December 2022 link spam update, Google's stated approach has been to neutralise links it doesn't trust rather than penalise the site receiving them. The usual outcome of a bad campaign isn't a manual action. It's that you've paid for links that pass nothing, and you find out months later.

A spike matters mostly because of what causes it.

Forty referring domains in a fortnight, all from sites with similar templates, similar outbound linking patterns and no organic traffic, is a footprint. Forty referring domains in a fortnight because a piece of original research got picked up is completely normal, and sites earn that pattern legitimately every day.

Natural is also relative to your own history rather than an absolute figure.

A site that's earned two or three referring domains a month for three years can absorb a large campaign without looking odd. A three-month-old site with eleven referring domains that acquires ninety in a month has changed its own profile ninefold, and there's no history for the new links to sit within.

For most sites we work with, a sustainable rhythm through blogger outreach is a handful of quality placements a month, sustained, with a larger asset-led push once or twice a year. That's partly a safety consideration and mostly a question of what's achievable, since genuine editorial placements take time to pitch, write and land.

When is the honest answer "none"?

The answer is none when links aren't what's holding the page back. Build them anyway and you've paid to accelerate a page that wasn't going to rank for reasons links don't touch, so it's worth ruling these five out first.

The page isn't indexed. 

Run the URL through the URL Inspection tool in Google Search Console. If it's excluded by a noindex tag, canonicalised to a different URL, blocked in robots.txt or simply not indexed, links to it do nothing at all. The check takes thirty seconds and gets skipped constantly.

The page doesn't match intent. 

If the top ten for your query are all comparison guides and you're pointing a product page at it, the mismatch is the problem. No volume of links will make a product page the right answer to a query where every ranking result is a guide.

The content is thinner than the SERP. 

Where every ranking page covers eight subtopics, includes original data or answers a set of related questions and yours does none of that, links are amplifying something that won't hold the position. Improve the page first and the links will work harder when they arrive.

You already out-link the SERP and still don't rank. If your page has more referring domains than everything on page one and it's nowhere, links clearly aren't the constraint. Something else is, whether that's intent, quality, technical setup or a site-wide trust problem.

The site has no topical footprint yet. 

Pointing links at a three-page site with no supporting content is expensive and slow. Building the content around the subject first makes every later link count for more, which is the subject of our guide to building topical authority with content and links.

None of these is a reason never to build links.

Each one is a reason to fix something first, so the links you do build are pointing at a page that can hold its position once it gets there. We'd rather have that conversation at the quoting stage than six months into a campaign, which is why the gap analysis comes before the proposal.

What does this mean for my budget?

Once you know how many referring domains your target pages need, you can turn that into a cost and a timeline. Two figures give you the plan, being the size of your gap and what a placement realistically costs in your niche.

Most of the published cost data is American and quoted in dollars, so the figures here are converted at roughly $1.33 to the pound. An analysis by PressWhizz of nearly 23,000 real placements found a median price of about £84 and an average of about £121. Those are publisher fees only, before anyone has done any work. Ahrefs has put the average nearer £270, and a 2026 survey of SEO professionals published by Editorial.Link found the average price they considered acceptable for a quality link was around £380.

Those figures aren't contradicting each other, and understanding the difference is worth more than any of them individually. The lower one is what a publisher will take to run a post. The higher one is what a placement costs once someone has built the prospect list, checked each site for real organic traffic and sane outbound linking, negotiated, briefed and written the piece, placed the link in context rather than in a footer, and verified it's still live months later. Pay the publisher fee alone and you're buying access to a site. The vetting, which is what decides whether the link passes anything, is the part you're choosing to do yourself. That same survey put the average minimum monthly budget for competing in genuinely difficult niches at around £6,300, with gambling named most often as the most expensive.

Add the tooling if you're running the analysis in-house.

Ahrefs starts at about £22 a month on Starter, £97 on Lite and £187 on Standard, Semrush Pro is around £105 and Majestic starts around £38, all billed in dollars so your figure moves with the rate. Ahrefs Webmaster Tools is free for domains you've verified, which covers your own site's data if not your competitors'.

Then do the arithmetic.

If your gap analysis says you need 15 additional referring domains across two target pages, and a sustainable pace is four or five quality placements a month, that's a three to four month programme rather than a one-off purchase. Budget for the pace rather than the total, and hold it for longer than feels comfortable, because links need to be discovered, assessed and supported by the page they point at before they show up in rankings.

If you want the number for your own pages rather than a general answer, send us the URLs and the queries you care about. We'll run the gap analysis and tell you what the SERP actually requires. Get in touch with the team.