Website traffic glossary

Every metric on a hap.io report, defined in plain language.

Traffic reports compress a lot of measurement into a handful of numbers. These are the terms hap.io uses, what each one actually counts, and the misunderstanding that most often attaches to it.

Visits and visitors

Monthly visits. The number of visits (sessions) a website receives in a month. A visit is one continuous period of activity by one person; the same person visiting morning and evening is two visits. Visits are not people — that difference is why habit products like search engines and feeds post visit counts far above their user counts.

Daily unique visitors. The estimated number of distinct people who visit on a typical day. "Unique" is always an estimate on the open web: people switch devices, clear cookies and share computers, so treat this as a scale indicator, not a headcount.

Global rank. A site's position when all tracked sites are ordered by traffic — #1 is the most visited site on the web. Rank is a comparison, not a measurement: a site can gain visits and still lose rank if the sites around it grow faster.

Engagement

Pages per visit. Average pages viewed in one visit. High values suit content and commerce sites where people browse; low values can be perfect health for a tool that answers in one page.

Average visit duration. How long a typical visit lasts. Duration is only meaningful within a category — a video site and a calculator site succeed at opposite ends of it.

Bounce rate. The share of visits that end after one page. A bounce is not automatically a failure: a visitor who gets their answer on the first page and leaves satisfied still counts as one.

Where visits come from

Direct. Visits with no referring source — typed addresses, bookmarks, app opens, and anything a browser cannot attribute. For famous sites, direct is the brand itself showing up in the data.

Referral. Visits arriving from links on other websites. The referring domains count is how many distinct sites link traffic in — a rough footprint of a site's presence across the web.

Search. Visits from search-engine results, split informally into unpaid ("organic") listings and paid placements, which some reports list separately. A site whose search share is high lives on being found; a site whose direct share is high lives on being remembered.

Social. Visits referred by social platforms. Social shares understate social importance — much of it arrives later as direct visits — but they do show which sites are built to be shared.

Top countries. The geographic distribution of visits. Read it as audience location, not company location: a US company can have a mostly international audience, and the shape matters more than any single share.

Desktop vs mobile. The device split of web visits. Apps are measured separately, so a low mobile-web share can mean a weak mobile presence — or a product whose mobile life happens entirely inside its app.

Search presence

Organic keywords. The number of search terms for which a site appears in unpaid results. Breadth here signals a content footprint; it says nothing by itself about how high the site ranks for the terms that matter.

Backlinks. Links from other sites pointing in. Counts vary widely by how they are measured, so backlinks are most useful compared against a site's direct competitors, not as an absolute score.

Apps

Downloads / installs. Store-reported install ranges for Android apps. Stores publish ranges, not exact counts, so app pages show them as published. Apple does not publish iOS download counts at all — iOS pages use ratings and rankings instead, labelled as such.

Ratings and rating distribution. An app's average star rating, its total rating count, and the spread from one to five stars. The distribution tells the truer story: two apps can average 4.2 with one loved-and-hated and the other merely liked.

Estimates and measured reports

Basic report. hap.io's starting report for every tracked site: estimated monthly visits with a published range, global rank and referring networks. Estimates are for scale and comparison.

Full report. The measured report: traffic sources, top countries, engagement and trend, published only when the measurement supports them. Where a figure cannot be measured responsibly — audience demographics, for instance — hap.io leaves it out rather than modelling it. How website traffic is measured explains why the two kinds of figures exist at all.