Skip to content
    Piiko
    ← All notes

    Note 06 · Paid acquisition

    Meta app ads: a first-campaign checklist

    Piiko·4 min read·Reviewed

    A small promise, put to the test.

    Two app creative cards and a little white megaphone connect to a phone displaying a confirmed outcome.
    Connect the promise in your ad to what people actually do in the app.
    The short answer

    Before spending, connect a real app, a tested conversion event, a clear creative promise, and a budget you can afford to learn with. Judge the users you acquire, not just the clicks.

    Decide what the first campaign should teach you.

    Before opening Ads Manager, write a short brief: the app and platform, the country and language, the user problem, the action you want, and the amount you can afford to spend learning. Start with a scope you can interpret rather than mixing every market and platform into one result.

    Choose a measurable primary outcome, such as installs, confirmed trials, or purchases. An install campaign can help test initial demand, but a cheap install alone says little about whether people will use or pay for the product.

    Use your unit economics to set a provisional spending limit. There is no universal daily budget or CPI target that makes a first campaign profitable. Set a total learning budget and a review point before launch.

    Connect the app and verify the event.

    Make sure the live store listing, platform identifiers, Meta app configuration, and measurement integration all refer to the same app. Follow the current setup guide for the integration you actually use.

    For example, AppsFlyer’s Meta integration guide explains its app-ID connection and in-app event mapping. It explicitly says a separate Meta SDK is not required for mobile attribution with that integration. Do not add overlapping SDKs simply because two tutorials mention them.

    Run a test journey. Confirm that the intended event arrives once, with the expected value and currency where relevant. A checkout tap should not be mapped as a confirmed purchase. Review your event definitions before asking an ad system to optimize for them.

    Attribution rules, privacy constraints, reporting delays, and modeled results can create differences between systems. Treat ad-platform reports as one view; reconcile business revenue with confirmed billing records.

    Give each creative a distinct reason to care.

    Use the app-promotion setup available in your account, select the correct app and platform, and choose the supported performance goal that matches your measurement. Meta Blueprint’s Advantage+ app campaign course is a starting point for understanding its automated campaign approach; available controls can vary by account.

    Test different promises, not just different background colors. For a hypothetical practice app, three concepts could show finishing a short session, getting feedback on a difficult task, and fitting practice into a busy day. Show the actual product experience and make the store page deliver the same promise.

    Preview every placement you enable. For Reels, Meta’s creative guidance recommends vertical 9:16 video and keeping important elements within the safe zone. Make the benefit understandable with captions as well as audio.

    Review the acquired users, not just the ad.

    1. Check the store destination, country, language, schedule, budget, and conversion event before publishing.
    2. After delivery starts, confirm spend and events are arriving as expected. Investigate broken tracking or a wrong destination immediately.
    3. Compare cohorts at the same age. Do not compare yesterday’s users with a group that has had a month to pay.
    4. Read cost per install alongside activation, trial or paid conversion, and contribution. Keep their denominators distinct.
    5. Record the decision: stop, investigate, change the promise, or run another bounded test.

    A good first campaign leaves you with a clearer next experiment. Increase spend when the evidence supports it, while checking how long recovery takes.

    Sources & further reading

    Examples are hypothetical. They illustrate the method and do not represent Piiko results or industry benchmarks.