Total sessions vs Meta-driven sessions

Bounce rate: overall vs Meta

Engagement rate: overall vs Meta

Meta CPA vs Mystery Picnic search CPA

Google search account matched to market by name (Americas → USA, Australia → AU, CA/NZ/UK matched directly). One line per market — toggle to declutter.

Meta CPA vs CPS (Meta only)

CPS = Meta cost ÷ GA4 (Meta) sessions, at account/market level — GA4 doesn't break out by campaign, so this can't be shown per-campaign. A falling CPA alongside a falling CPS can mean cheaper, lower-intent traffic is diluting results rather than creative genuinely improving. One line per market — toggle to declutter.

Meta CPA by day, selected range (since 30 Apr by default)

Campaign summary

Ads in the selected campaign(s), selected date range

Hook Rate and Hold Rate for video ads are not shown here — calculating them from this sheet's own video fields produced inaccurate results (Hold Rate exceeding 100% in some cases). These are instead computed directly via Windsor.ai inside a Claude chat: paste the prompt generated below into Claude and it will pull Hook Rate and Hold Rate for these exact ad names live from Meta.

GA4 Meta traffic quality — per market

GA4 (Meta) isn't broken down by campaign in this sheet, only by market and date. Shown per selected market below — this reflects all Meta traffic in that market, not the selected campaign(s) specifically.

Mystery Picnic search performance — per market

Google Search isn't broken down by campaign either, only by market and date — same limitation as GA4 above. Shown here for context on whether Meta creative changes are coinciding with search CPA movement in the same market.

Generate analysis