Home / Blog / 07

How to check whether a site has real traffic

Traffic is the one quality signal that cannot be manufactured cheaply, which makes it the best filter available — and third-party estimates are still only estimates.

verification · 11 min read

THE ONE SIGNAL THAT CANNOT BE FAKED CHEAPLYDomain rating →Verified organic sessions →
Authority is acquired through links, which can be bought. Readership is acquired through being worth reading.

Authority scores can be raised by acquiring links. Design can be bought. Article volume can be generated. Author bios can be invented. Readership is the one attribute that requires either genuine value or sustained paid acquisition, and neither is compatible with selling links at a hundred dollars.

That makes verified traffic the most useful single filter in link prospecting. It is also the one people check least carefully.

What the tools give you

Ahrefs, Semrush, Similarweb and others all publish traffic estimates, derived differently. The backlink tools model organic traffic from ranking positions and keyword volumes; Similarweb models total traffic from panel and clickstream data.

Both approaches have characteristic errors worth knowing:

  • Organic-only models undercount sites with large direct, newsletter or social audiences. A trade publication whose readers arrive from a Monday email may show almost no organic traffic while having a substantial and highly relevant readership.
  • Panel-based models are unreliable at small scale. Below roughly 5,000 monthly visits the estimates become noisy.
  • Both are estimates. Treat a figure of 4,200 as "somewhere in the low thousands", not as a measurement.

Four cross-checks

1. Keyword footprint

Ask how many keywords the domain ranks for, and look at what they are. A real publication ranks for hundreds or thousands of long-tail terms in its field. Inventory ranks for its own brand name and little else.

This one is harder to fake than the traffic number itself, because ranking for a broad keyword footprint requires content people actually search for.

2. Page-level distribution

Pull the top pages by traffic. Two patterns to distinguish:

PatternReads as
Traffic spread across many articlesA working publication
All traffic on the homepage, none on articlesBrand searches only — no editorial pull
One viral post, nothing elseA single lucky hit, years ago
No page above 50 monthly visitsNobody reads any of it

3. The article you would be placed on

Domain traffic is not the same as page traffic. A site with 80,000 monthly visitors can place your link on a page receiving eleven. Ask which URL the placement will be on — or for a niche edit, check the host page directly — and check that page's own numbers.

4. Signs of life outside search

Does the publication have a newsletter with a visible subscriber count? Comments on recent articles? An active social account where actual practitioners reply? Conference appearances? Recent articles cited by other sites?

None is conclusive alone. Together they distinguish a publication from a content farm faster than any metric.

The high-authority, no-readers pattern

The single most useful thing to look for. DR 55+ combined with under 500 monthly organic sessions is the signature of a site built to sell links.

The logic is simple: authority is acquired through links, which can be bought or built. Readership is acquired through being worth reading. A site that has one and not the other has been optimised for a metric rather than an audience — and the metric it was optimised for is the one most buyers filter on.

FOUR CROSS-CHECKS01Keyword footprint02Page-level spread03The placement page04Life outside search
Domain traffic is the common metric; page traffic is the useful one. A site with 80,000 visitors can place your link on a page that gets eleven.

Traffic that exists but is worthless to you

The subtler failure, and one that a traffic filter alone will not catch. Three versions:

Wrong geography. 40,000 monthly visits, almost all from markets you do not sell into. Check the country breakdown before treating the total as meaningful.

Wrong audience. A site with genuine traffic from students, job-seekers or hobbyists rather than buyers. Real readers, wrong readers. Look at what the top pages are actually about.

Wrong intent. Traffic driven entirely by informational queries with no commercial adjacency. Interesting for brand exposure, close to useless for a placement meant to influence a purchase.

In B2B, 3,000 monthly readers who are all operations managers in your category beats 300,000 general business readers. The traffic filter is a floor, not a ranking.

A workable threshold

Ours is 1,000 verified monthly organic sessions, with an explicit exception for tight-niche publications where the audience is demonstrably the right one and the total is genuinely small.

That exception matters in B2B more than elsewhere. Some of the best placements available in a specialist category sit on publications with a few thousand monthly visits, because the total addressable readership is a few thousand people. Applying a consumer-scale traffic threshold to those excludes exactly the links worth having.

What the threshold is really for is catching the other failure — the DR 60 site with 200 visitors that no filter based on authority would ever reject.

Asking a supplier

One question: "What is your minimum verified organic traffic on a referring domain, and which index do you verify it in?"

Answers, ranked:

  • Good: a specific number, a named third-party index, and an acknowledgement that niche exceptions are handled case by case.
  • Acceptable: a specific number with no stated source — ask the follow-up.
  • Bad: a domain rating threshold instead. That filter selects for the exact inventory this article describes.
  • Worse: traffic figures supplied by the publisher. Self-reported numbers from a party with a commercial interest are not verification.

the short version

Traffic is the hardest quality signal to fake. Cross-check with keyword footprint, page-level distribution, the specific placement page, and signs of life outside search. Watch for DR 55+ with under 500 sessions. Allow niche exceptions deliberately.

see our thresholds