When the concept of conversion value first appeared, it left a lot of app marketers confused.
To define, the conversion value is the only user activity postback the SKAdNetwork sends back to the app. This happens between 24 hours and up to 30 days after the install. It is a sum of six binary values, with 64 different combinations possible.
After one year in use, the concept of conversion value isn’t as confusing as it used to be.
Today, we can not only understand it but we can even talk about the best ways to get the most out of it.
Let’s dive right in!
1. Determining the main strategy
App advertisers are not a homogenous group. They have different goals, they advertise via different ad networks and measure different KPIs.
At the same time, they all want a conversion value strategy that corresponds with their KPIs and campaign optimization methods.
To achieve this, they use different strategies for measuring conversion value.
Their choice depends on the type of information they want to receive about the post-install activity of their users.
In most cases, these are engagement and revenue information.
For this reason, these are the two most dominant strategies app advertisers choose to build around conversion value. According to AppsFlyer, 84% of gaming apps use revenue-based strategies. On the other hand, the majority of non-gaming apps focus on measuring in-app events.
2. Starting simple
Conversion values may be limiting, but they bring different possibilities to the table.
These possibilities can guide advertisers in many different directions. For instance, they might want to start out with super granular strategies and experiments.
This is a trap that needs to be avoided.
When it comes to conversion value measurement, it is better to start with a simple setup.
For starters, advertisers should be using conversion value for reporting purposes, along with the data they already have on ATT opt-in users. This data is a valuable benchmark and a starting point for reporting based on conversion value.
Here’s what the starting setup can look like.
The conversion value setup should track important in-app events that occur frequently. This is especially important for apps that don’t have large user bases. According to Unity, these are the events that distribute the user base into groups of top 10% to 20%, the next 20% to 40%, and so on.
3. Mapping in-app events to bits
This is the most popular strategy app advertisers utilize for conversion value mapping.
As explained earlier, a conversion value is a sum of six values. These values are binary digits (bits) and they work as six on and off switches. Thanks to the combinations of these bits (0s and 1s), advertisers can know if certain events occurred in the app or not.
It is critical that advertisers track the right events in this process. The events they define should represent positive signals of user value.
These events can be observed either separately or in the form of a funnel.
By optimizing these six bits around in-app events, it is possible to measure early engagement, monetization, and retention signals.
Quick reminder: The window of tracking being between 24hours and 30 days but how does it work? The timing depends on the activity of the user. A 24-hour timer is reset every time the user fire a conversion value that is above the one the previous day and goes on like this. When the user reaches the highest conversion value number, the value is sent to the app.
This kind of strategy allows marketers to track the number of unique users who triggered an event, the number of times an event took place, or both. Based on this data, it is possible to build to predict the users’ LTVs.
4. Focusing on engagement events
Different in-app events carry different values for different apps.
For a lot of apps, the users’ engagement level in the first 24 hours is a great predictor of long-term LTV. For this purpose, app advertisers track engagement events to map conversion values (e.g., level completion, ad view).
For instance, marketers can choose a single event and observe the 64 combinations to measure how many times a user performed a certain event.
Let’s say the event is a one-minute session.
For instance, the number 000001 can suggest a user spent one minute within an app, while 111111 can mean the user spent 63 minutes in it.
However, most marketers go ahead and track multiple engagement events at once.
With six bits, it is possible to track up to six different events. Doing this gives fewer details about specific events, but it brings in a broader picture of post-install user engagement.
5. Focusing on conversion events
For some apps, the most valuable user LTV information comes from tracking conversion events.
In the same spirit as engagement events, the conversion value for these events tells marketers if a particular user triggered some of these events.
Let’s say we want to track six different conversion events for a single user:
- tutorial completion
- reaching level 1
- IAP (in-app purchase)
- friend invite
If a particular user only performs the first four events only, this would result in a recognizable binary digit (e.g. 111100). The marketer would then receive this information in the form of a conversion value from their setup and understand this user’s behavior.
6. Tracking revenue points
Conversion values don’t have to be observed as events – they can also be seen as revenue points.
In fact, this is the second most popular strategy for conversion value mapping.
In this strategy, every conversion value stands for a certain amount of revenue spent by the user.
Here’s the gist of it.
At the install, the CV (Conversion Value) is set to 0 by default. If a user buys a $0.99 pack, this can return a CV of 1. For a $2 purchase, the CV can be set to 2, etc. In the same manner, this can be done all the way to the conversion value of 63. This kind of strategy would allow advertisers to measure revenue from zero to $63.
This strategy should be determined based on the existing benchmarks. For instance, if users typically spend between $0.99 and $9.99 in the first two days of using the app, it’s only reasonable to set the CV 1 at $0.99.
Is this strategy a good fit for all apps?
This approach can be applied to all apps that make money from IAPs. However, it is the best fit for apps that monetize a lot of users within the initial 24 hours. Want to learn more about tracking and probabilistic attribution? Check out our article of the subject.
7. Tracking revenue groups
There is another, different way to measure conversion values based on revenue. In this setup, conversion values stand for different revenue ranges.
If a user spent anywhere from $0.99 to $3.99 in the app on in-app purchases, this can return a CV of 1. If they spent between $3.99 and $6.99, this can mark a CV of 2.
What’s the point of this approach?
With these postbacks, it is possible to figure out the revenue average from each group. As a result, the advertiser will gain access to different user value ranges. This information can then be used to fine-tune user acquisition strategies to target groups of users with higher CVs.
This approach is not reserved for IAP-based apps only. In the same spirit, it can be used to track ad revenue.
For example, if a user brings in anywhere from $0 to $0.05 by watching ads, this can return as CV 1. If they bring $0.5 to $0.10 in ad revenue, this returns as CV 2, etc.
8. Bit optimization for value measurement
All of the above-mentioned conversion value tips and strategies may seem to be working separately from one another.
This can but doesn’t have to be true.
The thing is, all of these conversion value strategies exist on the same base – the six bits. When deciding how to use these bits, marketers have a great deal of flexibility to achieve their goals.
There are three most common ways they optimize their bits:
- Flat method – using all six bits to measure a specific KPI (e.g. revenue)
- Split method – splitting bits to measure different aspects of user behavior (e.g. three bits for revenue points and three for conversion events)
- Combo split method – the split method combined with an additional yes/no signal (e.g. if the user was logged in)
What makes these methods appealing?
Obviously, the flat method brings granular insights about a specific goal. On the other hand, the split methods bring in diverse information about users.
For instance, they make it possible to find out if a user made a $10 purchase, signed up, completed a tutorial, and reached level 10. Yup, all of this can be deciphered from a single CV number.
9. Testing conversion value formulas
Once determined, the conversion value strategy isn’t set in stone. All of these strategies can and should be tested and fine-tuned.
For example, it would be good to test which six events that occur on day 1 have the biggest impact on long-term LTVs. On the other hand, by revenue range tracking, it’s desirable to test different revenue ranges.
However, making a complete strategy switch is a tricky move.
Changing the conversion value setup will affect measurement and optimization across UA channels.
10. Key takeaways for insightful conversion values
To end this list, here’s a short overview of the main things that make a conversion value strategy insightful:
putting the main focus on engagement and/or revenue
- for measuring revenue impact – identifying revenue ranges that matter for existing users
- for measuring engagement impact – detecting the most common and most significant early engagement signals
- launching the strategy and analyzing how the triggered conversion values are distributed
Take full control of conversion values with predictive modeling
Conversion value alone doesn’t tell you enough about the value of your users.
However, by combining it with predictive modeling, you can get the full picture you’re craving. That’s why we created Tempr., predictive and automation technology ideal for the privacy-first era. Interested? Schedule a demo or sign-up for our newsletter.