Compare how many cta and layout records are available and how their outcomes were reported. Evidence quality is assessed per record; heterogeneous lifts are not averaged into a causal benchmark.
GrowthLayer contains 2 cta records and 2 layout records. They include 1 and 1 reported winners respectively, but those proportions can reflect library selection, context, and missing negative results rather than category effectiveness.
Compare individual hypotheses and treatments that match your situation. Do not use these library counts to prioritize investment without representative sampling and a publication-bias analysis.
Save experiments with consistent metrics and context so future comparisons can use reviewable evidence instead of reported winner counts.
Start free