VisibilityOS by PodPitch

Compare AI visibility with the same engine mix

An audit’s overall mention rate can move when you collect more answers from one assistant and fewer from another. Check each engine’s rate and use the same weights before you interpret the change.

A fictional two-round audit

Each fraction shows answers that mention the tracked brand out of all completed answers for that engine.

EngineRound 1Round 2
Engine A16 of 20 (80%)64 of 80 (80%)
Engine B16 of 80 (20%)4 of 20 (20%)
All answers32 of 10068 of 100
Pooled mention rate
32% to 68%
Rate with equal engine weights
50% both rounds

Both engine rates stayed flat. The pooled rate rose by 36 percentage points because Engine A made up a larger share of the second round.

Report a like-for-like comparison

  1. Pair the same questions across dates and keep the engine, model, market, and search settings consistent.
  2. For each engine, divide the number of answers that mention the brand by the number of usable answers. Keep failed and missing answers out of that denominator.
  3. Choose weights before reviewing the results and apply them to both rounds. In this example, giving each engine half the weight yields 50% in both rounds.
  4. Report the per-engine counts, pooled rate, fixed-weight rate, and missing coverage together. Show any engine available in only one round separately.