After the announcement of our partnership in August, we received several questions around how existing customers would leverage the partnership. In this blog post we will walk through how a Teak.io end user would set up and experience the benefits.
To demonstrate how this works, this blog post will feature EXAMPLES throughout the walk through.
From a Teak end user perspective, nothing changes from a day-to-day workflow. She stays within the intuitive and responsive Teak.io web interface. No additional log-in or training is required. All the capabilities such as audience creation, messaging and rewards remain available.
Since Sonamine is a machine learning platform, Teak end users will need to list out some use cases that they are interested in. These use cases will determine what type of machine learning models Sonamine will be using. The Sonamine customer success team will meet with Teak end users and brainstorm these use cases.
In the case of churn reduction, the end users might want to segment out users based on their likelihood to stick around if they are given some rewards. Alternatively, end users might want to know who will definitely not churn, so they can target rewards accordingly.
Another common use case is to optimize first time conversions. In this case, the end user might want to segment users based on their conversion elasticity based on certain discounts.
EXAMPLE : For this blog, let us assume that we want to group users by their likelihood to lapse in the next seven days. This will help us target our retention rewards.
EXAMPLE : Let us also assume we have several actions that we can take to nudge non-spenders into their first purchase. And Sonamine will machine learn which is the optimal action for each user.
Once the use cases are identified, the Teak End User will need to set up the integration within the Teak Web Interface. This integration is completed for each game within the custom player properties section. For the use cases identified, one or more player properties will be defined here.
Clicking the “New Player Property” allows the Teak end user or administrator to name and describe a new Sonamine user property. Make sure to specify the “server API” will be updating the user property. Here you can name the user property in a descriptive manner. Either the number or string setting will work well.
EXAMPLE : For the retention use case, we could create a Sonamine_Churn user property. It will have integers. There will also be a Sonamine_Conversion user property that contains strings.
Once the integration is completed in the Teak Web Interface, Sonamine’s platform will start to perform machine learning tasks in the background. The data used for this machine learning is what has been stored in Teak. As a result, there is no additional work required on the part of the end user, or the customer’s data engineering team. Optionally, additional data that the end user wants to provide can be uploaded as needed.
For most integrations, Sonamine will update the Teak user properties within a matter of weeks. Games with more data may require one or two more weeks before the user properties start to be updated.
The Sonamine user properties within Teak will be continuously updated on a daily basis.
Sonamine User Properties within Teak behave just like any other Teak user properties. An end user can use them individually or combine them with other user properties to create audiences. The power of staying within the Teak web interface shows itself. The end user does not need to learn a new tool or perform any copy-paste or file uploads. The intuitive Teak web interface allows the end user to slice and dice the Teak audience using advanced Sonamine machine learning output, all through simple drop down filters.
EXAMPLE : For the retention use case, to create an audience that will likely lapse, we would use the Sonamine_Churn user property as a rule for the audience. The screen shot below shows in greater detail how that would be defined using easy drop down filters.
Once audiences have been created using Sonamine user properties, they can be used just like any other Teak audiences. Again the power of staying within the Teak web interface enables seamless application of machine learning within day-to-day messaging campaigns.
EXAMPLE : Knowing that an audience is like to lapse, we can offer them extra rewards to retain them. We can set up a Tiered Reward in Teak, offering these users a larger incentive with their next message.
All marketers remember tedious manual steps such as asking the data science team for a list of users, waiting for the list, and then importing CSV file into the marketing tool. The partnership between Teak and Sonamine makes life easier for the marketer. Now you can get the same advanced machine learning insights that your larger competitors with bigger teams have, all within the same web interface.
For a limited time until September 2025, Sonamine is offering a 60 day money back guarantee to OneSignal customers. Come experience the ease and simplicity of the First Time Spender Nudge package and watch your conversions soar.
For a limited time until September 2025, Sonamine is offering a 60 day money back guarantee to OneSignal customers. Come experience the ease and simplicity of the First Time Spender Nudge package and watch your conversions soar.