As a mobile app grows, the focus of the marketer shifts from user acquisition to user engagement. Enabling users on the top, middle or bottom funnel is important to have a sustainable mobile app with revenues and growth in user transactions. As the valuation game has shifted from looking at user counts to user transactions, this area of marketing has become the major component for marketeers spending plans. And wherever the spending increases, fraud follows. With more and more focus and information focussed on user-acquisition Fraud aspects, this article is intended for marketers who may be oblivious to re-engagement campaigns fraud.
Key Points to remember for Reengagement Campaigns-:
Attribution Window: While user acquisition attribution windows have crashed from the era of 28 days down to 24 hours (and in some case 3 hours!) the AW for reengagement campaigns are mostly still at the default 28 days level. Many advertisers are unaware on how to determine and configure this value based on their industry and user behaviour. There is no set value which you should set (it varies for industries and your customer’s behaviour patterns), but understanding what is the right value for your app is important (rather than simply accepting the default value). Generally values around 7 – 14 days are more than enough, since logically user behaviour is not influenced beyond this period in this era of continuous digital access and repeated advertising across channels. Look at the click to order patterns for your internal activities (e.g. push messages) and see how most people are responding. Something similar should be enough here also
View Through or Click Through: Again, many advertisers have shifted to Click Through for acquisition campaigns, but still continue with View through for re-engagement campaigns.
Inactivity Window: This is the single most important parameter which will decide whether you are passing engagement from Organic to Inorganic or not. A user needs to be inactive before it makes sense for you to spend money to make him active. An aggressive Inactivity Window only benefits shifting users from Organic to Inorganic without contributing to the bottom line. Also remember that Inactivity Timer is relevant not only for new users but also existing users who may already be active on your app
Lets look at different fraud aspects applicable in re-engagement campaigns :
Capturing Organic Activity: Typically re-engagement campaigns require the advertiser to enable passing all user activities post backs to the network. This implies that every time a user opens the app organically also, the re-engagement network gets to know about it. Now, for the network to simply fire clicks towards attribution platform for those device-ids (with the assumption of the user opening the app sometime in the default 28 days attribution window) to get converted is an easy way to capture organic traffic. The network may mix pure fraud traffic of firing clicks in the background with no user engagement with some ad campaigns shown to users. In effect the advertiser will see absolutely no benefit from the campaign and the total sales would simply move from organic to inorganic.
Key parameters to consider for detecting this :
- Click to Reengagement Time Spread: This is the time distribution of the difference between the time when the customer clicked on the reengagement Ad and when the user finally opened the app, resulting in the re-engagement of that user to be tracked on the attribution platform. When a network fraudulently fires clicks to hope for capturing the organic app opens, the spread of the click times and reengagement time will be spread all over. Also look at the concentration of click to reengagement time spread in the first few hours.
- Count of Clicks for same DeviceID: For a network trying to capture organic users, the clicks need to be repeatedly fired to capture users who may not have opened the app organically during the attribution window. Finding repeated clicks just outside attribution window being fired with the same deviceId is a clear-cut indication of organic theft.
As is the case in general with Performance Marketing, fraud is difficult to fathom since performance is clearly visible. Unfortunately, tweaking attribution platforms tracking capabilities is a easy enough way for fraudsters to show “performance” even when no ads are actually shown to users. Unfortunately, hardly any attribution platform tracks fraud in this space and is only recently started looking at user acquisition fraud. Which means that the advertisers need to be prepared to deep dive into statistical analysis to ensure they are getting what they are spending for. Or use a anti-fraud tool which protects you across all channels – mfilterit (mfilterit.com). “Performance” is not a benchmark for “fraud”. Protecting yourself is.