How SiteLens works
We want every number on this site to be honest. Here is exactly where it comes from.
Technology detection (high accuracy)
We fetch the homepage like a browser would. We look for fingerprints in HTTP headers, cookies, HTML, script URLs, meta tags, DNS records (MX, NS, TXT) and the SSL certificate. The fingerprints come from the community-maintained, open Wappalyzer-format database (7,628 technologies). Each detection on a report lists the evidence we found. We respect robots.txt.
Traffic estimates (ranges, not exact numbers)
Only a site's owner knows its exact traffic. Everyone else, including large paid tools, estimates it. Independent tests have found average errors around 50% even for the biggest tools, and much worse for small sites.
We use public popularity lists that are licensed for commercial use:
Each list's rank is converted to visits with a model of the form log(visits) = a + b · log(rank). When several lists cover a site we combine them; when they disagree, the range gets wider. We show a range covering roughly two out of three cases. If a site is not in any list, we say there is not enough data instead of inventing a number. log(visits) = a + b · log(rank)
Calibration status
The model was fitted on real traffic data on 2026-10-01. Fitted on 2195 sites with known traffic.
Pageviews, engagement and visitors (modeled)
We don't see a site's analytics, so pages per visit, visit duration and bounce rate are industry benchmarks for the kind of site we detect, clearly labelled as typical rather than measured:
| Site type | Pages / visit | Duration | Bounce |
|---|---|---|---|
| Online store | 3.5–6 | 3m–6m | 35–50% |
| Blog / content site | 1.4–2.2 | 1m–2.5m | 55–75% |
| News / media | 1.8–3 | 1.5m–3.3m | 50–70% |
| Web app / SaaS | 2.5–5 | 3m–8m | 30–50% |
| Documentation | 2–4 | 2m–5m | 40–60% |
| Forum / community | 3–6 | 4m–10m | 30–50% |
| Website | 1.8–3 | 1m–2.5m | 45–65% |
Earnings and worth (modeled)
Display-ad revenue = pageviews × revenue per 1,000 pageviews (RPM) × share not ad-blocked. RPM depends on where visitors are: $3–8 for the US, UK, Canada, Australia and Northern Europe; $1.2–3.5 for the rest of Europe, Japan, Korea and the Gulf; $0.3–1.2 for India and most other markets; $0.8–4 when we don't know. We assume 8–25% of pageviews are ad-blocked.
If we detect an ad network (such as Google AdSense) we call it estimated ad revenue; if not, ad potential, because the site probably earns in other ways. Worth is 24–40 months of that revenue, a common range for content sites. For stores and software products it is only a floor.
Every step has a range. We combine them the statistically correct way (in log space), so ranges stay useful instead of multiplying worst cases together.
Verified traffic
Site owners will be able to connect Google Analytics or Search Console to show exact numbers with a Verified badge. Verified sites are also used to calibrate estimates for everyone else.