How to Vet a Website Before Buying a Link (12-Point Checklist)

Most people vet a site in about eleven seconds. They glance at a Domain Rating, see a number above 50, and say yes. That number is a third-party estimate of link strength. It says nothing about whether anyone reads the site, whether Google trusts it, or whether the specific page your link will live on has ever ranked for a single keyword.

The checklist below is the one we run before any placement, and it’s the same one you can run yourself with a free Google search and whatever SEO tool you already pay for. It takes about fifteen minutes per site once you’re used to it. Fifteen minutes is cheap compared to $150 spent on a link that was dead on arrival.

What actually makes a backlink worth paying for?

A backlink is worth paying for when the page hosting it is one a real person could plausibly find, read, and click through from. That’s the whole test, stated simply. Everything in the checklist below is just a way of measuring how close a candidate site gets to that description.

The rest is proxy metrics. Domain Rating, Domain Authority, Trust Flow — these are useful for sorting a list of 200 prospects down to 40, and useless for the final decision. They’re calculated from link graphs, which means a site can inflate them by acquiring links itself, and plenty of site owners do exactly that before they start selling placements. Traffic is much harder to fake convincingly, and the shape of that traffic over time is harder still.

One more framing point before the list. You’re not evaluating a domain. You’re evaluating a page, on a domain, in a topical context. A link insertion (also called a niche edit) on a DR 65 site’s abandoned 2016 tag archive is worth less than a link on a DR 38 site’s article that ranks fourth for a keyword your customers actually search.

What are the 12 things to check before buying a link?

Run these in order. The first four kill most bad prospects fast, which saves you from spending time on the more subjective checks. If a site fails points 1, 3, or 5, stop immediately — those three are close to unrecoverable.

• 1. Organic traffic, and the shape of the curve. Pull the site into Ahrefs, Semrush, or Similarweb and look at the twelve-month organic traffic graph. Don’t just read the headline number — read the direction. A site holding steady at 3,000 monthly visits is a better host than one that peaked at 40,000 eighteen months ago and now sits at 2,800. Red flag: a cliff-edge drop that never recovered, which usually means an algorithmic hit or a manual action, and a vertical spike with no gradual build, which usually means bought traffic or a viral fluke. We won’t place on anything under 1,000 monthly organic visits regardless of how good the curve looks.
• 2. Where the traffic comes from. Check the country breakdown in your tool of choice. A UK fintech client gets very little from a site whose traffic is 78% Indonesia and 9% Bangladesh, even if the raw number looks healthy. Red flag: heavy traffic concentration in a country with no relationship to the site’s language or subject, or a top-country list that reads like a list of the cheapest available traffic sources.
• 3. Is the specific page indexed? Copy the exact URL you’ve been offered and search site: plus that URL in Google. If it returns nothing, the page is not in the index and your link cannot pass anything. Red flag: zero results, or a result that returns the homepage instead of the page. This check takes eight seconds and it’s the single most skipped step in the entire industry.
• 4. Does the page rank for anything? Run the URL through your tool’s organic keywords report. A page that ranks in the top 20 for even a handful of low-volume terms has demonstrated that Google is willing to show it. Red flag: indexed but ranking for nothing at all, or ranking only for its own title and the site’s brand name. Those pages are technically alive and functionally inert.
• 5. The outbound link profile. Open the site’s linked-domains report and sort by frequency. You’re looking for clusters. If a home-improvement blog is sending links to eleven online casinos, four CBD vendors, and a payday lender, its editorial standards are for sale and Google has almost certainly noticed. Red flag: any gambling, adult, pharmaceutical, or short-term-loan cluster, and outbound links in languages the site doesn’t publish in.
• 6. Sponsored versus editorial ratio. Scroll the last 20 posts and count how many are labelled sponsored, partnered, guest contributed, or carry an obvious commercial slant. Red flag: more than roughly a third of recent output being paid content. At that point the site isn’t a publication with occasional advertising, it’s an advertising channel with occasional publishing, and it’ll be read that way.
• 7. Is there a “write for us” page advertising link sales? Search the site for pages named write-for-us, contribute, guest-post, or advertise. A submission page asking for original expert contributions is fine and normal. Red flag: a page listing prices, turnaround times, DR figures, or “we accept dofollow links in any niche.” That page is a public confession, indexed and crawlable, and it tells Google exactly what the site does.
• 8. Do the authors exist? Click three author bylines. Do they have bios, photos, LinkedIn profiles, other bylines elsewhere? Red flag: every post attributed to “Admin” or “Editorial Team”, author photos that reverse-image-search to a stock library, or twelve different author names that all registered the same week and never wrote anywhere else.
• 9. Publishing cadence. Look at the dates on the last fifteen posts. A living site publishes at some recognisable rhythm — weekly, fortnightly, whatever suits it. Red flag: nothing for nine months and then eleven posts in four days, which is the classic signature of a dormant domain being reactivated to sell placements. Also watch for sites where every post is backdated to look older than it is.
• 10. Is the DR inflated relative to traffic? Compare the two directly. A DR 70 site with 900 monthly organic visits is telling you that its link profile was built deliberately and its content never earned an audience. Red flag: high authority score with negligible traffic, and a referring-domains graph that jumps from 40 to 900 in one quarter. Check where those referring domains came from — if they’re all footer links, directories, or expired-domain redirects, the score is decoration.
• 11. Topical cluster depth. Search the site for your subject. If you’re placing a link about commercial espresso machines, the site should have several other articles about coffee equipment, not one lonely post surrounded by crypto, weight loss, and travel insurance. Our own standard is at least three other genuine articles in the same cluster. Red flag: a site that covers everything, which means it specialises in nothing and exists to host whatever it’s paid to host.
• 12. Design and spam signals. Load the site on your phone. Is it readable? Are there four display ads above the fold, a pop-up asking you to subscribe before you’ve read a sentence, and an interstitial? Red flag: aggressive ad density, broken layouts, mixed fonts from three different themes, expired SSL, or a footer copyright date two years stale. None of these individually is fatal. Together they describe a site nobody is looking after.

How do you read a traffic curve rather than a traffic number?

The number tells you where a site is now. The curve tells you what Google has been deciding about it. Those are different questions, and the second one matters more when you’re buying an asset you want to still be working in two years.

Four shapes come up repeatedly. A gradual climb over 18-24 months is the healthiest thing you can find — it means content is compounding and nothing has gone wrong. A flat plateau is fine, if unexciting; the site has found its ceiling and stayed there. A sharp cliff followed by a flat line at a much lower level almost always means a core update or a manual action, and buying into that is buying into someone else’s problem. The fourth shape is the one people misread most: a near-vertical spike that appears in one month and decays over the next three. That’s usually a single post catching a news cycle, or bought traffic, and it doesn’t indicate anything durable about the site.

Widen the date range to two or three years before you judge. Seasonal businesses look catastrophic on a twelve-month view if you happen to catch them in the trough.

What separates a pass from a hard stop?

Use the table below as the quick reference. The middle column is what you want to see; the right column is where we walk away rather than negotiate, because these signals rarely improve and the cost of being wrong is a link that never counts.

How many points does a site need to pass?

There’s no scoring formula that survives contact with reality, so don’t build one. Treat points 1, 3, 4, 5 and 10 as gates — a failure on any of those ends the conversation. The remaining seven are weighted judgement, and a site can fail two or three of them and still be a reasonable buy.

An example of acceptable failure: a well-read regional food blog with genuine traffic, real authors, and a tight topical focus, but a slightly dated design and no author photos. That’s a site run by someone who cares about food and not about web design. Fine. An example of unacceptable failure: strong DR, decent-looking traffic number, excellent design, and a published price list for dofollow links. That site has told you what it is.

What should you check when a vendor is doing the vetting?

Ask for the live URL before you pay, not after. Any vendor unwilling to show you the exact page in advance is either hiding the site’s quality or hasn’t secured the placement yet. Then run points 1, 3, 4 and 5 yourself — four checks, five minutes, and they catch the majority of bad placements.

Also ask what happens if the link is removed. Sites get sold, redesigned, and pruned; links vanish. Our own policy is free replacement for twelve months, and if a vendor won’t put any replacement window in writing, price that risk into what you’re willing to pay. If you want to see how a vetting standard translates into an actual delivery process, our approach to placing links inside articles that already have readers lays out each stage, including what we reject and why.

Which checks matter most for a small budget?

If you can only run three, run the traffic curve, the page indexation check, and the outbound link scan. Those three catch dead domains, deindexed pages, and link farms respectively, which between them account for most of the money wasted on backlinks.

The subtler checks — author reality, cadence, cluster depth — matter more as your budget grows and your site gets more competitive. At ten links a month on a site in a soft niche, you can absorb a couple of mediocre placements. At a hundred links a month in a niche full of experienced competitors, patterns in your profile start to matter, which is where how you distribute anchor text across those links becomes as important as where they sit.

It’s also worth knowing what you should be paying before you start, since price and quality correlate loosely at best — see what backlinks actually cost across the market for the ranges. And if you’re weighing whether to buy at all, we’ve written candidly about where paid links sit relative to Google’s policies.