Mobile app attribution is vital to measure the effectiveness of a mobile app marketing campaign. The key role of a mobile app attribution technology is to accurately attribute key events like installs, opens, registrations, purchase etc. to the channel where the user came from. This allows advertisers to optimise their marketing dollars towards the most effective channels. Some of the key mobile attribution players in the market at the moment are TUNE, Appsflyer, Adjust, Kochava, Apsalar and it’s getting more exciting with the likes of Google Firebase & Facebook Analytics joining the fray, though it’s still early days for them.
From the very start, it’s very important for an advertiser to choose the right measurement partner which aligns with their business goals and their marketing strategy. Another reason why advertisers should be careful in choosing the right partner is because unlike tracking for online advertising which is cookie based and fairly simple to integrate on a website, integrating an attribution partner is a bit advanced and requires an sdk (Software Development Kit) to be integrated on the mobile app. Hence, once a partner is integrated it is an uphill task to switch over from one to another.
So how does one choose which mobile attribution technology they should be using? This is a very common question that we get from our clients since all of them on the surface seem to be providing a very similar set of features.
I have listed down four key important features to check before choosing a mobile attribution partner for your app:
1. Attribution Setting
Key factors to check on the attribution setting are:
• What is the default attribution window for both deterministic & non-deterministic methods?
• Does the attribution partner allow advertisers to customise the attribution window according to their needs?
• Does the attribution partner allow advertisers to customise the type of attribution they warrant ie. Identifier, google referrer & fingerprint?
• Does the attribution partner allow advertisers to enable impression attribution if needed?
Fingerprint attribution is the fall back non-deterministic attribution method in cases where the identifier or the referrer are not passed. Hence, for ads appearing on most mobile browsers it becomes the default attribution method. Also, since the introduction of ‘Limit Ad Tracking’ on IOS 10, fingerprinting is the default attribution method for such devices. This makes it very important for advertisers to have clarity on the exact parameters that form a device fingerprint and what combination of those parameters are used for attribution.
3. Fraud detection
Mobile ad fraud is by far the number one problem and challenge that faces the mobile marketing industry. That makes it even more important for an attribution platform to be the gatekeeper for any such activity and attribute accurately to legitimate sources. Every attribution platform has unique methods of fraud detection and prevention which vary from ineffective to very effective. Key things to check in here are:
• How does the attribution platform detect bot installs?
• How do they counter click spamming on a campaign level?
• What is the window for de-duplication of installs?
• Are they able to provide with uninstall data on a campaign level?
4. Availability of data
One of the main key features of any attribution platform is the data that it provides and presents to advertisers.
• Does the attribution partner provide an API access to pull raw data?
• Does the attribution partner provide raw click, install & event logs?
• Do they provide a wholesome report with all aggregate data i.e. listing clicks, installs, events, revenue on a campaign and sub-publisher level?
• Do they provide retention & cohort reports on a campaign and sub-publisher level?
• How long do they store the advertiser’s raw data?
Besides the ones mentioned above, there are other secondary factors like reattribution window, asynchronous clicks, pricing, deep linking methodology and managed access for marketing partners. Also, in our experience of working with all the above mentioned attribution partners, there are also subjective factors like quality of product support and troubleshooting which vary from partner to partner.