Most of the web measures readers by following them: cookies, pixels, cross-site profiles assembled over months. We do none of that. Everything we know about engagement comes from what happens on our own pages, counted in aggregate, with no individual file on anyone.
What we count
- Page views per publication and per article.
- Time on page, measured while the page is open, not inferred afterward.
- Scroll depth: how far down an article readers actually get.
- Return visits, counted with a first-party, anonymised marker.
- Newsletter sign-ups and ad clicks.
What we never collect
- No cross-site tracking. What you read elsewhere is your business.
- No data bought from brokers to "enrich" audience profiles.
- No individual reading histories, sold or shared. They do not exist to sell.
- No device fingerprinting, ever.
Why context beats cookies
Someone reading a guide to carry-on sizing is telling us about an upcoming trip, directly, with no surveillance required. The article itself carries the intent, which is the same argument that makes niche audiences outperform broad reach in the first place. SIG 09 makes that case in full. Advertisers get relevance, readers keep their privacy, and nobody is followed home.
We measure the room, not the people in it.
What partners actually see
Aggregate numbers, weekly: how many readers, how long they stayed, what they clicked. Nothing that identifies a person, because there is nothing to identify. The shape of that report, and the questions worth asking about any report, is covered in what a weekly report should tell you. The commercial side of the arrangement lives on the advertisers page.