Blog
What we learn running measurement at volume.
Methodology, benchmarks and the parts of ad measurement the industry tends not to say out loud.
Why a recall-based control group flatters your campaign
Splitting exposed from control by who remembers an ad conflates never-reached with simply forgotten. What that does to a lift number, and what to ask a vendor.
Reading a flat result without throwing it away
A metric that did not move is a finding, not a failure. How to tell a null result from an underpowered one, and what to do with each.
Mid-flight, verification, final report: three different documents
They answer different questions and carry different confidence. Treating them as one number is how measurement loses credibility.
What a benchmark is actually for
A norm tells you whether a result is good. It cannot tell you whether it is real. On using category benchmarks without hiding behind them.
What it means to be registered for measurement only
Under the IAB TCF a vendor declares its purposes. Ours are measurement and statistics — and the purposes we did not register matter just as much.
The hard part of cross-platform measurement is comparability
Every platform can report a number. Making those numbers mean the same thing is the work — and it is where consolidated measurement earns its keep.
These are placeholder briefs, not published articles. Write them in Strapi (Content Manager → Article) and they replace this list automatically.
Understand what happens within.
Bring us a question you can’t answer with the numbers you already have.