Setting in-app events is a crucial step in saving money on useless traffic, it is the key to understand the user behavior.
Let me show you how we do it in SDM GROWTH.
In App Events 101
Use the App, Play the Game
At SDM GROWTH we always start with using the app, we download the client’s app and try it. This gives us the user experience on top of hints to the efficient in-app events to define.
We see what triggers us as a user to commit certain actions and what causes us to leave the app.
Let's skip the technical side of implementing events with the attribution SDK (Check out AppsFlyer\ Adjust \Kochava etc guids) and jump straight to the events we want to implement.
Main Event
First, we need to understand the main event the user should do, this depends on the app.
For example, in social casino we want deposits. In e-commerce app we are looking for purchases. In news app we want to see users read many articles, and so on. The main event is mostly based on how the app generates revenue. Once that is done, we have the first open and the main event, we will start tracking the user’s activity on the first time they open the app, then the users will perform some activity in the app and after a while we are hoping they will perform the main event. Between the first open and the main event we spend time and money on working time, reporting, buying more users from this placement and more. What if we could tell the future and know in advance which of the users, which we already acquired, will keep using the app and eventually hit the main event?
We are not magicians but we can setup a funnel of events to track the user’s activity so in between the first open and the main event we will track few more events and call them mid-events, they will help us to predict the user’s activity. Based on this prediction we will optimize.
Mid Funnel Events
After learning the application, we can define the mid funnel events, each app is different but same verticals have common ground. This is a tricky stage and we suggest hooking the client into this discussion, the client does not necessarily understand this process, but you can talk him into this by leading questions.
So, how we optimize based on the mid events?
Examples again, news app, the main event can be reading ten articles a day (assuming the app generates revenue from interstitials between pages). We learned the app and talked to the client. Decided to define a mid-event of section interest, we know from learning the stats that a user who read more than one section in the first day will eventually turn into ‘ten articles a day’ user. From now on we will check the numbers of sections in day 1 and we will optimize based on this, by that we will save the money of tracking and buying users which in low chances reach our main event. And we will focus on these which in higher chances reach the desired goal.
Another vertical we want to use for example is social casino, we learned and talked to the client. Defined a mid-event of 50 spins after day 3. Optimize based on this will once again save us money and time as we don’t need to keep buying traffic from the same placement for 30 days, we can tell after 3 days.
In e-commerce app we can use ‘add-to-cart’ or ‘insert payment details’ as a mid-event.
Very important is to be flexible in the mid event optimization, its not a hard rule. We suggest optimizing traffic based on mid event with flexible bounders.
Another tip is to have more than one mid event, the optimization will be more accurate, and the flexibility can be smaller.
setting up a wrong mid-event will make you chase after the wrong horse, which is not the winning horse. Make sure you perfectly understand the app and the users. So now we know the importance of In-App events, as they are the key to data driven decision making. Got any questions on how to get started? not sure your events are good enough indicators on your users? Contact us, we'd be happy to help! [email protected]