Marketing Effectiveness now designs the holdout region for you

Holding a region flat is the most useful single thing a brand team can do to make next year's read defensible. It is also the step most often left out, because choosing a genuinely comparable region takes time and judgement.

What shipped

• Candidate control regions proposed automatically, ranked by similarity across volume, mix, access and seasonality.
• A similarity score per candidate, with the dimensions that drove it shown rather than summarised away.
• Minimum detectable effect calculated up front, so you know before starting whether the test can answer your question.
• A read-out schedule with the twelve-week point marked, since earlier reads tend to mislead.

Available now across the platform. If you are self-hosting it is in the current release, and the similarity model runs entirely inside your environment.