<b>The traffic spike that paid no bills</b>

<b>The traffic spike that paid no bills</b>

A content site celebrated a 30% month-over-month traffic jump and started planning around the new 'audience'. Ad revenue, oddly, stayed flat.

The clue was in the network and engagement split. The new sessions clustered in a single hosting-provider network, had a 0.0s engagement time, and a 99% single-page rate. Real humans don't behave that cleanly. It was a scraper farm hammering one URL.

They added a filter segment to exclude that network and rebuilt their baseline so future reports compared real to real.

The '30% growth' became an honest 4%, and ad RPM per genuine session was restored from a deflated $6.10 to its true $11.80.

The lesson: a traffic spike with flat money is a question, not a celebration. Check engagement and network before you build a strategy on ghosts.