Wide and unconfigurable lookback windows raise concerns
SKAdNetwork attribution windows and the 60-day first launch window
For the SKAdNetwork to award attribution, a certain sequence of events must occur within specific time windows before the postback can be sent to the winning ad network. This includes the time window between the impression or click and the install, plus an additional first launch window.
For instances where a user sees an ad but doesn’t click it (eg, impression), it could take 61+ days to award attribution. The advertiser will only receive a postback if the user downloads and installs the app within 24 hours after seeing the ad, and then launches the app inside a 60-day window. The SKAdNetwork will get a report of the install only after the user launches the app.
For instances where a user not only sees the ad but also clicks it, a postback for an attributed install could take up to 91+ days. The SKAdNetwork will only report the install if the user downloads and installs the app within the first 30 days after the click and launches the app for the first time within the same 60-day window.
Read more about receiving ad attribution through SKAdNetwork here.
Understanding lookback windows
Lookback windows play an important role in marketing attribution, as they determine the allowable time frame between an ad view or click and the resulting conversion (install or in-app event) to which it can be attributed. While 30 days usually is the default lookback window to deterministically match a click based on a device identifier, such a long lookback window is not recommended
Lookback windows: MMP vs. SKAdNetwork
Per the table above, the SKAdNetwork has a wide click lookback window of 30 days, which is unconfigurable, meaning advertisers cannot tighten it to suit the specific goals of their campaign or nuances of the mean-time-to-install (MTTI) for their app and vertical.
The importance of attribution window configurability
Marketing runs on a cause-and-effect model. The goal of a campaign ad creative is to drive the user to complete a specific action. If the user engages with your ad by clicking it but then takes a break for 30 days until finally coming back to perform the originally intended conversion goal, was the campaign effective? Reason would say no.
Analyses by Kochava data scientists and analysts have shown that tightening lookback windows greatly improves the correlation between a marketer’s ad spend and the conversions it drives. Excessively long lookback windows often lead marketers to overpay media partners for conversions they didn’t actually cause.
This is why Kochava often recommends advertisers set much shorter lookback windows, in some cases as tight as 1-3 days, or even a matter of hours or minutes in instances of an incentivized campaign. Marketers will not have this flexibility with their SKAdNetwork attribution.
Prior to the release of Apple’s SKAdNetwork, marketers had grown accustomed to scaling their lookback windows to meet their marketing campaign goals and to account for differences between app verticals and industries. With the SKAdNetwork, lookback windows are wide and cannot be configured or tightened. As a result, there is potential for over-attribution and increased potential for fraudulent activity.
Questions still remain about SKAdNetwork lookback windows
The SKAdNetwork’s approach to lookback windows stir up more questions.
Will Apple eventually support configurable lookback windows?
There is currently no mention of the ability to configure or customize these windows; however, we hope the need for configurability is recognized and implemented within future iterations of the SKAdNetwork.
How will the long lookback windows impact attribution accuracy?
With the limited data provided within SKAdNetwork postbacks, it will be difficult to determine the impact on attribution accuracy caused by the 30-day click lookback window in particular.
There’s potential that real-time attribution processed by an MMP on the cohort of devices that opt-in to the AppTrackingTransparency (ATT) framework can provide insights when compared to the SKAdNetwork attribution outcomes of the same campaign overall. Marketers may be able to use such insights to refactor their ad spend accordingly.
What about ad fraud?
A key reason that Kochava advises customers to tighten lookback windows is to mitigate the impacts of common fraud tactics like click flooding/click spamming and also prevent organic sniping of installs. It’s not yet clear how Apple will holistically detect and prevent such tactics, but offering configurable lookback windows would be a step in the right direction.
Stay up to date
With the continuous uncertainty of SKAdNetwork and ATT framework, Kochava strives to provide accurate and digestible information. If you have questions or want to stay up-to-date on the latest information check out our iOS 14 resource center, contact us, or fill out the form below to subscribe to our newsletter.