Research

Why there is no industry incrementality benchmark, what the spread between randomized experiments actually measures, and the arithmetic that tells you whether your own test can answer the question.

How much paid media now runs through Performance Max, Advantage+ and AI Max, what still decides results, and where people fit when the platform does the buying.

The ad is the largest factor in paid media return an advertiser controls. Why it belongs with a creative team rather than the media team or a folder of old assets, and what to change at every stage.

How big a creator to hire, when a brand is ready, what makes creator content work, and where AI influencers fit. Built on 345 controlled experiments, revenue data from real campaigns, and an interactive reach-per-dollar model.

What Google's AI Overviews take from organic search, how much optimisation wins back, and where the ceiling sits. 39 sourced figures, an interactive recovery model, and the measurement setup we recommend.

What advertisers can buy inside AI assistants today, what is in test, and what exists only in research. Covers ChatGPT, Google AI Mode and Copilot, why incremental return is still unmeasured, 89 sources, and a model for sizing a first test.

When a geo test or a media mix model is worth paying for, how to run a matched-market test that returns a usable answer, and how to carry the result into weekly reporting. With an interactive test-size calculator.

How a year of geo testing on two client accounts became a graded incrementality matrix that sets platform bid targets and moves budget daily. With a matrix-to-bid-target calculator.