Rewarded Ads
Part of Measuring gaming advertising
Separating rewarded engagement from purchase intent
Understand what rewarded ad actions show, then assess stated purchase intent with a suitable survey and comparison group.
A player can choose and complete a rewarded ad for an in-game benefit without wanting the advertised product. Treat that choice as rewarded engagement. Measure purchase intent with a separate question asked of a suitable audience, and track later customer behaviour as a different outcome.
Know what the reward events show
Google's Android rewarded-ad documentation describes loading callbacks and full-screen callbacks for dismissal and failure to show. Those platform details do not establish that a particular publisher reports every event to an advertiser or grants the reward correctly.
Ask which stages the publisher can report: offer shown, opt-in, ad started, reward earned and advertiser action. State the denominator for each rate. Reward-earned events divided by ad starts measure progress through the incentive flow, while visits divided by impressions answer a different question. Neither is a purchase-intent score.
| Signal | What it supports | What it cannot establish alone |
|---|---|---|
| Opt-in | The player accepted the reward opportunity | Interest in the brand |
| Reward earned | The incentive condition was recorded as met | Desire to purchase |
| Advertiser tap | A destination action was recorded | A valuable visit in every case |
| Purchase-intent answer | A stated inclination at survey time | A future purchase |
| Purchase | A recorded customer action under a stated method | An incremental campaign effect without a suitable comparison |
Key Metrics from Rewarded Ad Campaigns
- Reward-earned rate
- Reward earned ÷ Ad starts
- Opt-in rate
- Opt-ins ÷ Offer shown
- Advertiser tap rate
- Destination actions ÷ Ad starts
- Purchase-intent lift (if applicable)
- Exposed group intent rate – Control group intent rate
Ask an intent question of the right people
Define the product, the eligible Australian audience and the decision period before writing the question. Keep reward wording out of the purchase question so respondents aren't led to think buying is required to earn the game benefit. Use identical wording and response choices for exposed and comparison groups, and record when each group was surveyed.
Group design matters most for voluntary rewarded inventory. People who opt in may already differ from those who decline, so a comparison between completers and all other players can mix audience differences with any effect of the ad.
IAB Australia recommends comparable test and control groups, adequate samples and random assignment where a controlled design permits it. Ask how the study handled eligibility, declines, failed ad loads and other campaign exposure.
If a suitable comparison cannot be formed, describe the result as purchase intent reported by surveyed players, with the sample and timing stated. Don't present it as intent lift caused by the campaign. An answer given straight after earning a reward may reflect that moment; it is not a later purchasing decision.
Keep behaviour as a separate check
A qualified visit, enquiry or purchase can show later behaviour within its tracking and attribution limits. It does not measure what a player said they intended to do, and an attributed action alone does not prove incrementality.
Use reward events to diagnose the placement, the survey to assess stated purchase intent, and customer events to assess later behaviour. Their agreement or disagreement can guide investigation, but none should be relabelled as another.
