SpendWhat Meta charged for the ad in the window (USD).
ROASPurchase value ÷ spend, as attributed by Meta. 2.0 means $2 of revenue per $1 spent. It runs higher than Shopify's own numbers because Meta claims credit for view-through and multi-touch sales, so use it to rank ads against each other, not to reconcile revenue.
Purchases / CPAMeta-attributed orders, and spend ÷ purchases (cost per acquisition). Lower CPA is better.
Hook rate3-second video plays ÷ impressions. Did the first three seconds stop the scroll? Around 25–30% is healthy for feed video; under 20% usually means the opening needs rework. Static-image ads have no plays, so they show n/a.
Hold rateThruPlays (15 s or to the end) ÷ 3-second plays. Of the people who stopped, how many kept watching? Low hold with a high hook means the opening over-promises.
CTRLink clicks ÷ impressions. Whether the ad moves people to the site.
FrequencyImpressions ÷ reach: how many times the average person saw the ad. Above about 3 with a falling CTR is creative fatigue.
RankingsMeta's own quality / engagement / conversion ranking versus competitors for the same audience (above average, average, below average). Only shown for ads with 500+ impressions.
VerdictScale: ROAS at or above the scale threshold with a healthy hook. Keep: above breakeven. Fix the hook: converts but the opening is weak. Refresh creative: above breakeven but frequency is high and CTR is dropping. Landing/offer problem: people stop and click but don't buy. Kill: below breakeven, or no purchases. Not enough data: spend is under the minimum. Thresholds are editable above.
Trailing windowAll figures cover the selected window (Last 7d etc.) ending on the latest day with data; "vs prior" compares to the window before it. The feed re-pulls the last seven days every morning, so recent days settle as late conversions arrive.