Short answer: optimize for the event closest to revenue that gives each campaign enough conversions to judge, then test value optimization when eligible. Meta’s in-app purchase value requirement is at least 15 purchases with values in the last 7 days, plus the required event setup. Compare Ads Manager with your own cohort revenue and contribution after costs. Set a ROAS goal from your campaign’s history and the return your business needs.
Page Contents
What can you optimize for in a Meta app campaign?
In the app promotion objective there are three levels:
- App installs. Meta finds people likely to install. Install volume alone does not tell you which users will pay.
- App events. App event optimization finds people likely to complete a standard or custom event after installing, such as a purchase, a subscription or a level reached. Custom events work too, but Meta notes you cannot optimize custom events for value, because they carry no value parameter.
- Value. Maximize value of conversions finds people likely to spend more, not just people likely to buy once.
You do not have to move through these in order. The rule is the same at every level: pick the deepest event that still gives the campaign enough conversions a day for Meta to find more of the right people and for you to judge your ads. Ideally that is 30 a day per campaign, and 10 a day is good. On a small budget, 1 or 2 a day is fine; it just takes longer to read. Budget follows from it: cost per conversion times the conversions a day you want. For a fuller explanation of event optimization, see what is AEO on Meta.
What do you need before value optimization works?
Meta’s requirements for app conversions are:
- an active integration: the Meta SDK, an MMP or the Conversions API;
- a purchase event that sends a value and a currency;
- at least 15 in-app purchases with values over the last 7 days (or, for ad-monetized apps, 15 in-app ad impressions with at least 2 distinct values over 28 days);
- app events configured for value optimization in Events Manager. For SKAdNetwork campaigns, enable it in the SKAdNetwork event configuration; Meta says that switch is not needed for events sent through Aggregated Event Measurement.
Meta adds that access can take a few weeks after you meet the thresholds. Before switching, check that purchase events arrive with accurate amounts and currency. Missing or incorrect values undermine purchase-value optimization. Keep actual transaction values separate from lifetime-value projections.
Highest value or ROAS goal?
Once value optimization is available, you choose a bid strategy:
| Bid strategy | What Meta does | Use it when |
|---|---|---|
| Highest value | Prioritizes spending the budget while maximizing purchase value | You want more conversion value without a specific ROAS goal |
| ROAS goal | Keeps the weekly average ROAS near a target you set; may underspend when the market is competitive | Protecting a return matters more than spending the budget |
The current Ads Manager name is ROAS goal. It is an average target, not a guaranteed minimum return. Highest value benefits from variation in purchase values; if your purchases all have similar values, compare it with purchase-event optimization rather than assume it will improve results.
How do you set a ROAS goal, including for a new app?
Meta’s bid-strategy guidance gives a starting example of a ROAS goal 10% to 20% below recent average ROAS, using the campaign’s attribution window. With no history, it suggests highest value to establish a baseline. Check that any starting goal also meets your economics; do not lower it below an acceptable return just to increase delivery. A new game has no universal target, so first establish what its acquired users produce within the period you can afford to wait.
Your acquisition break-even point depends on what a user contributes within that period after store fees, refunds, taxes and other variable costs. The business also needs enough margin to cover fixed costs and profit. How to calculate LTV without a lifetime of data covers the projection.
How long until you can see ROAS results?
Allow for the time between installation and purchase, plus reporting delays. Meta’s ROAS goal guidance says to evaluate weekly averages and wait at least 7 days after an adjustment before re-evaluating. A subscription with a longer trial may need more time before the acquired cohort’s paid revenue can be judged.
How do you measure real profit, not the ROAS Meta reports?
Meta’s ROAS is Meta’s attribution of revenue it can see. Profit needs your own numbers:
- Cohort revenue. For the users acquired on each day or week, how much did they pay by day 7, day 30 and later? Pull it from your MMP or your own revenue data, by campaign.
- Contribution. Take off store fees, refunds, taxes and variable delivery costs. Revenue divided by spend is ROAS; it is not profit.
- Blended view. Compare total spend with total new-user revenue across channels, because any platform’s attribution can credit it with users who would have come anyway.
Account for the reporting delays and attribution limits of your iOS setup before changing a goal. How to fix mobile attribution on iOS covers the setup, and what ROAS means for mobile apps covers the definitions.
How should you switch without wrecking what works?
Run the new value or purchase campaign alongside the one that works, with its own budget sized from the conversions a day it needs. Do not move all your spend at once. Use your proven ads, so the test is about the optimization, not the creative. If the new campaign gets more value for the money after a few weeks, scale it about 20% a day, or 50% to 100% a week if you want to move faster, as long as the return holds (budget scaling for app campaigns). If purchases fall to too few a day to read, loosen the ROAS goal or go back to the purchase event.
Does creative change when you optimize for value?
Creative remains the main lever. Different promises can attract users with different conversion and revenue profiles, even under the same optimization setting. Read cost per conversion and revenue per user by creative concept, not only by campaign. For subscription apps, the trial-or-paid choice is covered in should you optimize for trial starts or paid conversions on iOS.
Frequently asked questions
How do I optimize Meta campaigns for in-app purchases, not just installs? Choose app event optimization for the purchase event and budget from cost per conversion. Our guide is 30 conversions a day ideally and 10 as good; smaller campaigns can run on 1 or 2 and take longer to judge. Test value optimization once Meta’s separate purchase-value and setup requirements are met.
How do I set a minimum ROAS in Meta? Choose maximize value of conversions as the performance goal and ROAS goal as the bid strategy, then enter the target. Meta’s starting example is 10% to 20% below recent average ROAS, but the target must still fit your economics and is not a guaranteed floor.
How do I measure ROAS from in-app purchases? Send each purchase to Meta with its value and currency through the SDK, an MMP or the Conversions API, then measure cohort revenue against spend in your own data, net of store fees and refunds.
How long until I can see ROAS results on Meta? Judge weekly averages, and wait at least 7 days after a change, as Meta advises. Allow more time where the purchase or trial cycle and reporting delay require it.
Do I need 50 conversions a week? No. That is really old guidance we would avoid. Aim for about 30 conversions a day per campaign, with 10 a day being good.
