How to read a backlink profile before you buy a site

Why the backlink profile matters more than the traffic graph

Every site listing comes with a traffic screenshot. Nobody shows you the backlink export unless you ask for it, and that’s exactly why you should ask first. Traffic is a lagging indicator. It tells you what happened over the last twelve months. The backlink profile tells you what’s going to happen over the next twelve, because links are the asset that keeps paying rent long after the seller stops updating content.

I’ve bought sites where the traffic chart looked clean and the link profile was a mess underneath it. The traffic held for a few months after purchase, then slid, because the links propping the rankings up were the kind that stop counting once someone at Google’s spam team gets around to a fresh crawl of the linking domains. Backlink profile analysis is how you catch that before you wire the money, not after.

Pull the raw export, not the summary score

Every backlink tool gives you a single number: domain rating, authority score, whatever the vendor calls it. That number is a weighted guess built from a model you can’t see. It’s fine as a first filter, useless as a verdict. Export the actual list of referring domains and go through it by hand, or at least by spreadsheet formula. You’re looking for patterns a summary score can’t show you: a cluster of links from sites that were clearly built for one purpose, a handful of domains linking through five or six different URLs on the target site, dates that bunch up in a way normal link acquisition never does.

Ahrefs, Semrush, and Majestic will all give you slightly different domain and link counts for the same site because they crawl different slices of the web and weight recency differently. Pull at least two and compare the referring domain lists, not just the headline numbers. If one tool shows 400 referring domains and another shows 90, that gap is information. It usually means a batch of links came from a source one crawler indexes and the other doesn’t, which is often a sign of a link source built for volume rather than for actual traffic or citation.

Anchor text tells you what happened, not what should have happened

A natural link profile has messy anchor text. People link to your homepage with your brand name, with the bare URL, with “this site,” with the article title, with nothing descriptive at all. When you sort a backlink export by anchor text and see the exact target keyword repeated across dozens of unrelated domains, that’s not a natural pattern. Nobody writing a genuine recommendation types your five-word commercial keyword as the anchor.

Build an anchor text pivot table. Group by exact match, partial match, branded, and naked URL. A site that’s been actively built with commercial anchors will show exact match sitting at 15 to 30 percent or higher. A site that grew organically usually shows branded and naked URL dominating, with exact match anchors making up a small slice. Neither pattern guarantees anything about the future, but a heavy exact-match skew tells you the previous owner was pushing rankings with links rather than earning them, and that kind of profile is more exposed when algorithms reassess how much weight over-optimized anchors deserve.

Referring domains matter more than total backlinks

A site with 8,000 backlinks from 40 referring domains is a different asset than a site with 800 backlinks from 400 referring domains. The first pattern usually means sitewide links: footer credits, widget links, directory listings that stamp the same URL across every page of a linking domain. Those links get counted once by the linking domain’s authority but they don’t represent 8,000 separate votes of confidence. The second pattern, more domains with fewer links each, tends to reflect genuine, distributed interest.

When you’re evaluating a listing, always look at the referring-domain-to-backlink ratio before you look at the raw link count. Sellers who quote “50,000 backlinks” in a listing description are, more often than not, quoting a number inflated by a small set of domains linking many times over.

The velocity chart is where the bodies are buried

Every backlink tool has a “new and lost links over time” chart. This is the single most useful view for a buyer because it shows acquisition behavior, not just outcome. A site built through years of content and outreach shows a bumpy but generally upward line, new links arriving in small batches tied to specific pieces of content or PR mentions. A site that bought its way up shows spikes: flat for months, then a burst of forty or fifty new referring domains in a two-week window, then flat again.

Spikes aren’t automatically disqualifying. A site that got picked up by a roundup or went semi-viral on a forum can show a spike too. The difference is what’s inside the spike. Click into the domains that appeared during it. If they’re a mix of recognizable publications, forums, and niche blogs with actual traffic, that’s a real event. If they’re a wall of domains you’ve never heard of, with generic names, thin content, and no organic traffic of their own, you’re looking at a link campaign, and you need to price the deal as if those links could stop counting.

Check whether the linking domains are alive

A domain showing up in a backlink export doesn’t mean it’s still a working, indexed, traffic-getting site today. Take a sample of thirty or so referring domains, more if the profile is large, and manually check each one: does it resolve, does it have an estimated traffic number above zero in the tool you’re using, is the content still there or has it been repurposed into something unrelated. I do this by hand because it’s the fastest way to catch a profile built on a network of domains that have since been abandoned or repurposed. If a third of your sample is dead or deindexed, extrapolate that ratio across the full referring domain count and mentally discount the profile’s real strength accordingly.

What disqualifies a deal

A few patterns move a listing from “worth negotiating” to “pass.” Heavy concentration of links from domains that share hosting infrastructure or ownership signals, visible in shared WHOIS patterns or identical site templates across dozens of linking domains, is one. A backlink profile where the newest links are months older than the listing date, meaning the seller stopped acquiring links the moment they decided to sell, is another, because it suggests they know the link source is about to dry up or become a liability. And a profile where a large share of referring domains sit in unrelated niches with no topical connection to the site’s subject matter is a signal the links were bought in bulk rather than earned through relevant outreach or content.

None of this means the site is worthless. It means you price it as a content and domain asset, not as a site with durable link equity, and you budget for replacing that link foundation yourself if you want the rankings to hold.

What a clean profile actually looks like

A profile worth paying a premium for shows referring domains that are topically relevant to the site’s niche, anchor text dominated by branded and natural phrases, a velocity chart with irregular but sustained growth rather than spikes, and linking domains that are still live, indexed, and getting their own organic traffic when you check them. It won’t be perfect. Every real site picks up a few spammy or irrelevant links over time just by existing on the internet. The question isn’t whether bad links exist, it’s whether they’re the exception or the foundation.

The buy decision

Backlink profile analysis isn’t a pass or fail test with a clean cutoff, and anyone who tells you a specific score guarantees the rankings survive a change in ownership is guessing, because nobody outside Google actually knows how a given link is weighted at query time. What you’re doing when you read a profile before buying is pricing risk. A clean, relevant, slow-grown profile is worth negotiating up toward the asking price. A spiky, exact-match-heavy, dead-domain-riddled profile means you negotiate down, or you plan to rebuild the link foundation yourself after the purchase, and you price that rebuild into your offer.

Do this work before you send an offer, not after you’ve paid and started watching the rankings for signs of trouble. The export is free. The lesson is a lot cheaper to learn on someone else’s spreadsheet than on a site you already own.

If you want help running this kind of audit before your next acquisition, or building out the link side of a site you already run, The SEO Desk covers backlink profile analysis, keyword research, link building, and technical SEO from the operator’s side of the table.