Documentation

Support

Unity Ads User Acquisition

Campaigns

Attribution

Open User Acquisition dashboard

Unity Ads User Acquisition

ROAS campaign best practices

Review the recommended best practices for ROAS campaigns with Unity Ads User Acquisition.
Read time 3 minutesLast updated 2 days ago
Before you create a Return on Ad Spend (ROAS) campaign, review the following recommended best practices.

Optimization-specific best practices

The following best practices are for specific ROAS optimization types.

In-app Purchase

Refer to the following In-app Purchase (IAP) campaign best practices:

Set adequate budgets

When you launch a new IAP ROAS campaign, set budgets large enough to acquire at least 10 purchasers per day during the first week. This ensures your campaign can collect data quickly and exit the Learning phase sooner. To determine appropriate budgets, use your average CPI from previous campaigns as a reference. Multiply this by 10 users per day to calculate a minimum daily budget. Adequate budgets ensure your campaign can bid competitively and meet the Learning phase data thresholds more quickly. If you have previously run D7 IAP ROAS campaigns with a mature install cohort, the dashboard recommends a D7 ROAS goal for each geo. Unity calculates this recommendation based on your observed ROAS from the D7 IAP ROAS campaigns. To set custom ROAS goals for each geo, refer to historical D7 purchase data.

Learning phase best practices

The following best practices are specific to the Learning phase of your ROAS campaign.

Collect unattributed postbacks from the Unity SDK

To accelerate your campaign's learning phase, use the Unity SDK to collect additional SDK data and ensure that you send Unity the corresponding post-install event’s unattributed postbacks for at least two weeks before you create your new ROAS campaign. This data won’t be counted as part of the learning threshold, but helps to train the machine learning models that enable optimization. When you supply unattributed postbacks and SDK data before campaign launch, it can improve initial optimization predictions and help shorten the Learning phase.

Avoid setup changes during the Learning phase

While your campaign is in the Learning phase, avoid making changes to your campaigns settings, such as budgets or targeting parameters. Performance fluctuations are normal during this phase. Making changes during the Learning phase can disrupt the machine learning process and extend the Learning phase. After your campaign reaches Live status, you can optimize your settings for better performance.

Adjust ROAS goal if install volume is too low

While you should generally avoid setup changes during the Learning phase, you might need to adjust your ROAS goal if your campaign can't reach the required install threshold. If your campaign has been in the Learning phase for an extended period without meeting the install requirement, lower your ROAS goal. This allows the system to place higher CPI bids and acquire more users, helping your campaign exit the Learning phase.

Focus on United States and Tier 1 countries during the Learning phase

During the learning phase, the recommended best practice is to launch with US or multiple top performing Tier 1 or Tier 2 countries. Tier 3 countries typically have lower user density and monetization rates, which can make it difficult to collect sufficient data during the Learning phase. This is especially true for IAP and Hybrid campaigns, which require payer data.

MMP integration best practices

The following best practices are specific to your preferred Mobile Measurement Partner (MMP) integration setup.

Send install data from all sources

When you share install data with Unity, ensure that you include data from your entire user base, including attributed, unattributed, and organic installs. Refer to the Introduction to Attribution for definitions of each attribution type.

Share both values and revenue for Purchase and Ad Revenue events

When you share Purchase or Ad revenue event data with Unity, share both values and revenue for each event. Sharing this complete data improves Unity's User Acquisition models and can improve campaign performance.

Set postback windows to lifetime if applicable

If your MMP supports lifetime postback windows, always set the window to lifetime. Otherwise, set your postback window to the maximum length. Setting your postback window to lifetime allows your campaign to learn from your app's entire install history and can improve campaign performance.