skip to Main Content

How to work with SKAD traffic in mobile games: lessons and tips

Alexey Gavrilov, marketing producer at mobile game publisher, AppQuantum, offers insights into working with SKAD traffic for 12 months in the ATT era.

The release of iOS 14.5 last year shook up the mobile world. In addition to this, Google has also announced its own version of ATT. Thereby, we continue to adapt to our new reality: continue to exchange knowledge, test hypotheses and learn to work without user data in hand. Now is the time to take stock and share our tips. Over the past year, mobile publisher AppQuantum has developed a system for working with iOS traffic for mobile games. In this article, we want to share our methods of buying, attributing, and scaling SKAD traffic effectively.

main points
First of all to make you aware of the conditions that we are going to discuss. The three main ones, circulating close to each other are ATT, IDFA and SKAdNEtwork.

Application Tracking Transparency (ATT) is a framework, with which app developers ask users to allow them to use an advertising ID. Now, users can prohibit applications from sharing their personal data: gender, age, device on which the application was installed, time and date of installation.

Users who choose in are a small portion of the overall population, with the last information from Flurry showing that the global percentage is only 25%.

Identifier for advertisers (IDFA) is a device marketing identifier, which advertisers use for ad attribution, retargeting, working with lookalike audiences, analytics, and other tasks.

SKAd network (SKAN/SKAD) is Apple’s attribution technology, on iOS 14.5+ devices. SK stands for StoreKit Apple’s framework, through which developers interact with the App Store and in-app purchases. A d advertising and Network are explicit.

Prior to ATT and SKAD, mobile advertisers could track user data and pass it all to the MMP (Mobile Measurement Partner) throughout its life within the app. Today, the amount of data available is catastrophically insufficient to make marketing decisions. Because of this problem, others stretch out like a web. We will explain in more detail exactly what these problems are and how you can recognize and fix them.

The challenges of working with SKAD
First, there were issues with testing creatives, building purchase predictions, and attributing traffic. Each area required a workaround. Let’s take a look at the specific challenges and the solutions AppQuantum has found for them.

Traffic distribution
Almost every marketer working with iOS traffic has encountered attribution issues. Now, to attribute traffic accurately, you first need to determine buying patterns. We will mention the two main ones:

  • When buying traffic for a GEO is split by sources For example, the United States on Facebook, Germany on Google, France on Snapchat, etc. When you buy a country’s traffic through a separate source, it’s easier to track its dynamics. The downside of this option is a huge amount of implicit scans.
  • When purchase is tested on all sources and GEO By turning off and on each source, you can measure the campaignefficiency. It’s not a cheap method and should only be used when you’re willing to spend more than $50,000 (but the amount greatly depends on the amount of organic traffic and CPI). You need to be prepared for the prospect of first buys that won’t bring tangible results, but instead bring a lot of knowledge to AU and analytics teams.

These options are not perfect and may not be right for you. But they work for us and you can use these methods, customize and modify them, finding the best templates for your specific projects.

Limited number of advertising campaigns
Per Apple’s rules, you can now only track 100 active campaigns for each ad network. Most Self-Reporting Networks (SRNs) have limited the number of campaigns for advertisers to new10 or 11. Campaigns above that use the source to teach their machine learning algorithms.

If you have already targeted for wWorldwide, you now need to group GEO segments into tiers – 1, 2, 3. Traffic, in turn, will need to be assessed at the campaign level. The GEO level is not relayed in SKAD postbacks but you can see the spread by country in the traffic source itself.

Plan the next purchase
The biggest changes in user acquisition after SKAD occurred in advertising campaign forecasts. For example, we previously planned the purchase with user and cohort level data, but this is no longer available. There is now only data at the campaign and source level, and therefore the purchase must be planned at these levels.

SKAD attribution allows us to track the first 24 hours after installation, so we can predict traffic, using day zero (d0) data.

Say you’re tracking revenue by conversion value (CV). The metric itself is a number between 0 and 63. It indicates the value or quality of the conversion. In the CV you can follow the events of the game, in-app purchase revenue and, in some cases, advertising revenue. The latter can be used to forecast future purchases using this formula:

LTV = SKADRevenu_d0 * predict coef_d0

The downside of this option is that the revenue amount in d0 might be insufficient for quality forecasting, but overall it works well.

In addition, you should not delete the data of users who have opted in to tracking (ATT). Despite the fact that these users are few, they still exist, so this data will be useful for forecasting.

The share of opt-in users may vary depending on the source: for example, according to our data, the percentage is high in TikTok. According search, TikTok users are more responsive to brand messages, calls to action, and advertisements in general. It is possible that more users will understand what data tracking is for. So you can assess TikTok traffic and extrapolate the results to your other sources.

What the future holds
By introducing such a severe personal data protection policy, Apple has started a trend that will only get stronger. Granular targeting, aimed at each particular user, along with user-level analysis and probability attribution, will be a thing of the past. Developers will need to further explain why they need users to opt in to data tracking. For now, we’re working with what we have and looking for new ways to work “in a vacuum”, without user data, keeping our finger on the pulse and sharing our last discoveries with you.

Source link

Back To Top