Semrush, Ahrefs and Similarweb don’t measure people — they measure signals: search volume data, clickstream panels, ranking positions. Those signals can be inflated, and in some corners of the web they routinely are. Since every revenue estimate in our revenue models guide multiplies something by traffic, this checklist is the foundation the whole series stands on. Seven checks, in the order we run them.
1. Google Trends: the shape of demand
Search the brand name in Google Trends. Genuine branded demand grows the way reputations grow — gradually, with news-driven spikes that decay naturally. Artificial volume has a signature: a sudden vertical spike from zero to a high plateau, holding unnaturally flat, sometimes ending as abruptly as it began. Real interest breathes; manufactured interest is a square wave.
2. The footprint test
Reported search volume should be proportional to visible footprint. A brand allegedly searched 100,000 times a month has left tracks: press coverage, social profiles with real followers, app store listings, Reddit threads, employee LinkedIn pages. If tools show five-figure branded volume but Google’s own results for the brand are two thin blog posts and a marketplace listing, the volume and the reality don’t match — and reality wins.
3. Authority vs. traffic mismatch
DR/DA and traffic usually rise together, because both come from the same cause: people caring about the site. A DR 5–15 site showing tens of thousands of monthly visits is an anomaly that needs an explanation. Sometimes there’s an innocent one — a single viral page, a keyword nobody else targets. Often there isn’t.
4. Where does the traffic come from?
Open the site’s top keywords in any SEO tool. Healthy sites earn traffic across dozens of informational and commercial queries. A red-flag profile earns 80–95% of its reported traffic from a single keyword — frequently the site’s own domain name. Real businesses get branded searches in addition to topical traffic, not instead of it.
5. Does anyone link, cite, or discuss?
Traffic leaves cultural residue. Tens of thousands of monthly readers produce comments, backlinks from real sites, social shares, quotes. Zero engagement plus huge reported traffic is a contradiction. Check the backlink profile too: a handful of referring domains, mostly directories and other low-DR blogs, tells you what the “audience” actually is.
6. Check independent trust databases
Run the domain through ScamAdviser, WHOIS lookups and the Wayback Machine. Age matters — most artificial-volume plays involve domains registered within the past 12–24 months. Hidden ownership, shared hosting with dozens of similar low-trust sites, and a homepage that changed identity completely between Wayback snapshots are all part of one pattern.
7. The economic logic test
Finally, ask the analyst’s question: who benefits from this number existing? Inflated branded volume makes a domain and its ad inventory look more valuable to buyers who check tools but don’t verify. If the site’s business model is selling placements — and you can check its marketplace listings directly — the incentive to manufacture impressive third-party metrics is built into the model itself. That doesn’t prove the volume is fake; it tells you the burden of proof is on the number.
Scoring it
We treat any two failed checks as “unverified — discount heavily” and three or more as “assume artificial until proven otherwise.” In our breakdowns you’ll see this verdict in the At-a-Glance ledger at the top of each article, and revenue estimates for such sites deliberately exclude traffic-dependent income like display ads.
Why we publish this
SandBridge estimates website and company revenue in our Business Model Breakdowns. We can’t do that honestly without separating real traffic from manufactured traffic — so we published the checklist we use. It’s general education, not an accusation against any specific site; individual breakdowns cite their own evidence.