What Makes a Backlink Worth Paying For? 7 Checks Before You Say Yes

The metric almost everyone checks first is the one a seller can fake for $15. In a 2024 research series by Xamsor, the team paid black hat freelancers between $15 and $85 per domain and watched five fresh test sites climb to Ahrefs DR scores of 44 to 59. None of those domains had a reader, a ranking, or a reason to exist. Every one of them would sail straight through a "DR 40+" filter.

That is the core problem with buying links: you are usually shopping with someone else's scorecard. I have vetted thousands of prospect domains while building link campaigns, and the sites that looked best on a metrics screenshot were often the worst deals on closer reading. These are the seven checks I actually run before agreeing to pay for a placement, in the order that saves the most time.

1. Check the traffic before anything else

Organic traffic is the hardest thing for a seller to fake at scale. To manufacture it, a site has to actually rank for something, which is exactly the property you are paying for. So I start there and ignore authority scores completely at this stage. Drop the domain into Ahrefs' free traffic checker or the free Semrush overview. I want a site doing at least a thousand organic visits a month, spread across a decent number of pages instead of one lucky post, with a 12-month line that holds steady or climbs.

The mismatch is the finding. A DR 68 site pulling 300 visits a month is not a bargain hiding in plain sight. It is a site whose authority score and reality have come apart, and you should assume the score is the fabricated half.

2. Treat DR as a filter, never a verdict

Domain Rating measures one thing: the linking patterns pointing at a site. Ahrefs is honest about the limits. Their own explainer on the metric notes that DR ignores search traffic, domain age, and brand popularity, and warns that "large amounts of low-quality backlinks may actually increase your DR, not decrease it."

Used correctly, DR is a sorting tool. It helps you cut a 500-domain list down to 80 worth reading. It cannot tell you that any single domain deserves your money, because single domains are exactly what the $15 manipulation trick targets. This is also why the scoring criteria we use at LinkIntel weight live traffic and relevance above any authority number: the score is one input, never the answer.

3. Ask who actually reads the site

Before paying, read the five most recent post titles and try to describe the site's reader in one sentence. If you can ("marketing managers at small ecommerce brands"), relevance is workable. If the last five posts cover crypto exchanges, dog food, wedding rings, and a casino guide, the audience is nobody, and Google's systems are better at spotting that pattern than most buyers are.

Topical fit beats raw authority in almost every trade. A modest home-improvement blog linking to your power tools page passes a plausibility test that a stronger general-news domain never will. It is also the only kind of page that sends referral clicks that convert.

4. Read the outbound neighborhood

Open three recent posts and count the external links. Where do they go? A healthy editorial site links out a handful of times per article, mostly to sources. A link farm reads differently: eight to forty outbound links per post, pointing at commercial pages in unrelated niches, with the classic casino, CBD, and essay-writing cluster showing up sooner or later.

Check for a "write for us" page with a public rate card while you are at it. A site that sells placements to anyone is a neighborhood, and you are buying a house in it next to every competitor with a credit card. Your link inherits the company it keeps, and it gets diluted by every other paid link on the page.

5. Confirm Google still cares about the site

A site can look alive and be algorithmically dead. Two minutes settles it. Search site:theirdomain.com and see whether the site is indexed at a size that matches its archive. Then take the exact titles of two of their newest posts and search those: fresh content from a healthy site gets indexed within days. Finally, look at the traffic trend line for a cliff, the signature of a site caught in a spam or core update.

A devalued site will still happily sell you a link. The placement goes live, the report looks fine, and the link passes nothing, because Google stopped counting that site's votes months ago.

6. Look at where your link would live

Placement on the page matters as much as the domain. An editorial mention inside a paragraph of a relevant article is the format every study of link value assumes. A slot in a "useful resources" list at the bottom, a footer, or an author bio is worth a fraction of that, whatever the domain's metrics say.

So ask for specifics before money moves. For a link insertion, get the exact URL and the paragraph where your link would sit, then check whether that page itself ranks for anything or gets any traffic. For a guest post, ask which section of the site it will publish under and whether posts there get internal links from the rest of the site. A page that can rank compounds: it passes equity, sends readers, and keeps earning after you paid once.

7. Price, terms, and the disclosure question

Only now does price mean anything. Current going rates for real placements are covered in our link building cost breakdown, but the short version is that quality editorial placements are rarely cheap, and a price far below market is usually the checks above failing quietly. If a DR 60 "editorial" link costs $40, you now know why.

You should also know what the rulebook says before you spend. Google's spam policies treat buying or selling links for ranking purposes as link spam unless the link is qualified with a rel="nofollow" or rel="sponsored" attribute. Plenty of the industry operates in the gap between that policy and its enforcement, but that is a risk decision you should make with open eyes, not one a seller should make for you by omission.

Finally, terms separate professionals from farms. Ask how long the link is guaranteed to stay live, what happens if the post is deleted or the link switched to nofollow, and whether there is a replacement or refund policy. A seller who verifies their own placements and offers a defined issue window has skin in the game. A seller who goes quiet at these questions has answered them.

The five-minute version

Traffic first, DR as a filter only, a nameable audience, a clean outbound neighborhood, proof the site is still indexed and counted, an editorial placement on a page that can rank, and a price with terms attached. Two failed checks is my walk-away line, at any price. The full pass takes about five minutes per domain once it becomes routine, which is nothing against the cost of a year of bad placements.

The last habit worth building: judge links after you buy them, not just before. Rankings and referral data will tell you which sellers were worth it, and our guide to measuring link building ROI covers exactly what to watch. And if you are not yet sure which of your pages even need links, run the free link audit first: it maps your gap against competitors page by page, so you spend on the pages where a link changes something.