How D7 ROAS is calculated
D7 ROAS follows a cohort from install through day 7:
D7 ROAS = revenue from the cohort in days 0 to 7 / ad spend on the cohort
A worked example: on 1 October a Meta campaign spends $4,000 and drives 2,000 installs. Those users generate $400 on install day, $900 by day 3 and $1,400 by day 7. D0 ROAS is 0.10x, D3 ROAS is 0.23x, and D7 ROAS is 1,400 / 4,000 = 0.35x.
D7 ROAS is only final once every user in the cohort has reached day 7, so the 1 October cohort is complete on 8 October. Reading it earlier undercounts.
Setting a D7 target
A D7 target is a proxy for a longer goal. Work it out from your own cohorts:
D7 target = payback goal x (D7 revenue / revenue at the payback day)
If your goal is 1x ROAS by day 90, and past cohorts made 30% of their D90 revenue by day 7, the D7 target is 1 x 0.30 = 0.30x. Campaigns above 0.30x D7 are on track; campaigns well below are unlikely to pay back by day 90.
The ratio differs by business model. Commerce and food delivery apps earn much of their early revenue in the first week, so D7 is a strong predictor. Subscription apps with a free trial earn almost nothing in the first 7 days, so trial starts or trial-to-paid conversion are better early signals.
Pitfalls
- Small cohorts. A few large purchases can swing D7 ROAS a lot. Use weekly cohorts for small campaigns.
- Mismatched windows. D7 by install date is not the same as revenue in the last 7 days.
- Changing mix. If the share of iOS, a new geography or a new offer changes, old D7-to-D90 ratios may no longer hold.
- Delayed iOS data. SKAN and AdAttributionKit postbacks arrive late and report coarse revenue ranges, so iOS D7 reads need care.
D7 ROAS in Linkrunner
Linkrunner reports revenue at D0, D3, D7 and D30 by install cohort, with spend synced from Meta, Google and Apple Ads, so D7 ROAS appears per campaign, ad set, creative and keyword without exports. LTV up to D30 shows how the curve continues. See ROAS measurement, or set targets with the ROAS calculator.