Free Player Retention Guide Get Now
IMVU (by TogetherLabs) is the world’s largest avatar based social networking app. Sonamine has been partnering with Together Lab since 2019.
In an interview with Senior CRM Manager, Donnie Kajikawa, she discussed how a complete retention program solved key challenges by combining different capabilities with data science. It is summarized here.
Like most free-to-play games and apps, IMVU had a very large churn rate. These are non spenders who installed, registered but abandoned the app pretty quickly. Morever the required capabilities to run retention campaigns were distributed over separate departments.
As game marketers, we spend a lot of time and energy on optimizing user acquisition. But we tend to ignore the tail end of the funnel, that is, churn. In fact if you can decrease churn by 10%, it's like improving your UA by 10%
I’ve been running programs for a long time, and one of the hardest things is to coordinate among different departments and tools. Many ML projects fail because it’s hard to get the ML output to the different places that it can be used.
Sonamine's backend was used to identify likely churners who were active, and could be engaged with a retention reward.
Sonamine set up several different AB tests in the CRM tool, Leanplum. These AB tests were used to identify the level of retention rewards most likely to appeal to different user segments, such as different platforms.
Specialized reporting not available in the CRM tools allowed analysis on several KPIs over a 90+ day horizon. This was critical to identifying the payback period and addressing concerns with unexpected impacts.
Based on the AB tests results, Sonamine helped to design an automated reward system to enable one-to-one personalized rewards, without needing precious game development resources. This reward system is now used in live ops events and streaks.
We decided to leverage Sonamine services, as an extension of the internal team, but also to get to market faster. If you are just starting out, we recommend working with a partner first. You can always bring it in house later.
Within 4 weeks we had our first test in-app campaign running! It’s a testament to the team effort. And we have been sprinting ever since. We are now even running conversion campaigns using the same automation paradigm.
Currently in the aggregate of all our campaigns, the rewards group has a 23% higher ARPU compared to the control group!
We find that their DAU level is 10% higher than a control group that does not get them. And interestingly, this elevated DAU level is sustained even 60 days after the initial reward!
High School Story, Hollywood University and Choices are successful narrative mobile games built over the last decade. Sonamine has been partnering with Pixelberry since 2014 to improve their game performances.
Like other successful free-to-play games, Choices had great retention and relatively low payer conversion rates. There were precious few resources to devote to increasing the payer conversion rate.
A breakout content oriented game such as Choices required a constant stream of new content to satisfy players. There was no bandwidth to integrate new SDKs for live ops or CRM tools.
Growth stage games focus their management attention and resources to driving new players. The payback logic is straightforward, more UA leads to more players and revenue. However, increasing the payer conversion rate by 10% is similar to increasing ROAS by 10%.
To preserve internal resources for game improvements, Sonamine provided an end to end package to drive first time conversions.
Sonamine backend connected to the analytics tool used and retrieved the required data for analysis. There was NO data engineering resources or work required on the part of Pixelberry.
Customized data science was used to target users likely to respond to a nudge message.
Custom integration was set up with the Leanplum system so that users could be flagged and instantly eligible for the nudge messages and alerts. User flags were also removed once they were no longer picked up by the machine learning model. This was critical to ensure that users were not spammed with irrelevant messages.
New content genres, business models such as subscription were introduced over the years. These new game features were integrated into the First Time Spender Nudge package, thus enabling new data to be used for predictions, and new nudge options.
Sonamine services team created and set up the AB tests and messaging within the CRM tool, Leanplum. Such end-to-end service allowed Pixelberry product and live-ops teams to focus on other features and events.
By creating a long term holdout control group to compare performance over several years, the incremental revenue of the first time spender nudge was calculated. It was an extremely high ROI number.
Sonamine moves the needle with actionable user predictions.
- Oliver Miao, CEO Pixelberry
Instead of spending time and money on new development within the game client, the New Spender Nudge used the existing Leanplum implementation combined with offline data science. The first AB tests were running within weeks of game launch.
The First Time Spender Nudge maintained its higher performance compared with a holdout user group over several years. There was no indication of "revenue acceleration", the phenomenon of hastening purchases but not increasing the overall LTV. This is a direct result of proactive tweaks and automated machine learning.
IMVU (by TogetherLabs) is the world’s largest avatar based social networking app.
Like most free-to-play games and apps, IMVU had a very large churn rate. These are non spenders who installed, registerer but abandoned the app pretty quickly. Morever the required capabilities to run retention campaigns were distributed over separate departments.
As game marketers, we spend a lot of time and energy on optimizing user acquisition. But we tend to ignore the tail end of the funnel, that is, churn. In fact if you can decrease churn by 10%, it's like improving your UA by 10%
I’ve been running programs for a long time, and one of the hardest things is to coordinate among different departments and tools. Many ML projects fail because it’s hard to get the ML output to the different places that it can be used.
We decided to leverage Sonamine services, as an extension of the internal team, but also to get to market faster. If you are just starting out, we recommend working with a partner first. You can always bring it in house later.
Within 4 weeks we had our first test in-app campaign running! It’s a testament to the team effort. And we have been sprinting ever since. We are now even running conversion campaigns using the same automation paradigm.
We find that their DAU level is 10% higher than a control group that does not get them. And interestingly, this elevated DAU level is sustained even 60 days after the initial reward!
Currently in the aggregate of all our campaigns, the rewards group has a 23% higher ARPU compared to the control group!
Ut sunt minima inventore perferendis odit. Quis vero odio qui velit quasi quia cupiditate commodi. Consequatur enim consectetur sed itaque quisquam et. Nobis suscipit amet.
Enim quam temporibus cupiditate recusandae omnis et ut rerum laborum. Deserunt dolorum et. Quisquam ipsa sed veritatis cumque facere qui mollitia. Laudantium earum ullam voluptate voluptatem. Illum provident molestiae dolores assumenda omnis pariatur. Et rerum sint.
Debitis aut voluptas totam cum rerum. Sit voluptatem molestiae iure eos. Sit et repudiandae quo hic tenetur qui qui.
