Measuring gaming advertising

Measure a gaming campaign against the result it was bought to achieve. Before launch, agree what counts as delivery, which audience and geography the report covers, and what evidence would support the goal. Keep exposure, interaction and outcomes separate: each answers a different question.

Start with a measurement brief

Name the games or apps, devices, placements, creative versions, campaign dates and Australian eligibility rule. List the measures each supplier can provide for each format. A package containing game-world signs and rewarded ads needs separate format reporting; a combined engagement rate would hide how players encountered the message.

Key Measurement Metrics for Australian Gaming Ad Campaigns

Keep the event definitions separate

For served intrinsic placements, ask suppliers to report ad requests, impressions and viewability separately, and document their definitions.

For rewarded ads, confirm which events the supplier reports and how each is defined. Reporting may vary by implementation, so do not assume an event available in the app or SDK is included in campaign reporting.

Name the denominator for every rate. A completion rate based on ad starts cannot be compared directly with a visit rate based on impressions. If two reports differ, first align their dates, time zones, format definitions, invalid-traffic treatment and measurement coverage.

Match evidence to the goal

For recognition, establish credible exposure and consider a brand study that compares responses from exposed people with a suitable eligible control group. Check which inventory the study includes and what question it asks. A measured difference in recall or awareness is distinct from a sale.

For an action campaign, define the useful destination event before launch. A visit, install, registration and meaningful product use are different outcomes. Label results as attributed under the stated method unless a sound comparison supports an incremental claim.

Design a comparison around the question

Turn the campaign objective into a hypothesis before launch, then choose metrics and a method that can test it. IAB Australia's guidance on digital ad impact studies emphasises that measures should align with the business objective. The same campaign can have different valid measures depending on what the business is trying to achieve.

Where the question is whether advertising caused an outcome, a controlled experiment can distinguish incrementality from correlation. Plan the comparison before activity begins. It should show what the campaign changed, not just report events that happened after exposure.

Make the test and control groups as comparable as possible. The guidance recommends random assignment, matching groups on relevant characteristics such as demographics and prior exposure, and ensuring each subject is in one group rather than both. Allow for a sample size suited to the precision required: detecting smaller differences requires larger samples.

Use results to guide the next decision

Decide in advance what campaign choices the result could inform. Controlled experiments can assess the effects of targeting, ad context, format and frequency. They provide evidence to optimise current or future campaigns, not a retrospective scorecard.

Keep the question broad enough to reflect how gaming activity sits within the wider campaign. Controlled experiments can examine cross-media effects and the interplay of digital with other media; geographic tests can assess the effect of online activity on in-store sales.

Present findings with the hypothesis, chosen measures, method and result together. This lets decision-makers judge whether the evidence answers the original question. It also avoids treating a result from one format or campaign setting as proof that another will perform the same way.

Read the Australian report on its own terms

State the rule used to classify an opportunity as Australian and whether planning and reporting use the same rule. Show unknown geography separately where available. Label reach as people, households, accounts or devices, and ask whether overlap across games is removed.

Request breakdowns by format and placement, the share that could be measured, and material device or game-version exclusions. Keep an unknown viewability classification unknown. At review, ask what demonstrably ran, what opportunity the intended audience had to encounter it, and what evidence supports the claimed outcome.