Data identifies churn segment in lifecycle marketing

A previous post focused on the how lifecycle marketing differs from live ops. Here we will dive deeper into lifecycle marketing. Specifically we will discuss how data analysis is used post launch to improve the native built-in onboarding flow, with the goal of improving retention.

Content drop announcements might be lowering your KPIs

Most games will announce new content when it is available. Users are greeted with splashy messages explaining the exciting new content. Developers feel a sense of accomplishment of pushing through a new release. But just how do these announcements affect your KPIs?

From our many customer implementations, we will examine the most common method of content announcements, a wide message sent to all active players, usually limited to once or twice per user.

Lessons from Candy Crush Saga Live Ops

At Gamesbeat Summit 2022, I attended a Live Ops panel with Jon Radoff (Beamable), Josh Yguado (Jamcity) and Jason Bailey (East side games). I asked the panel whether there were specific games genres where live ops would be less effective. Josh's answer surprised me: almost all genres could use live ops. Even a match 3 game such as Candy Crush leverages live ops to incredible effect.  So, how does Candy Crush do it?

Removing QA and internal users from analytics pipeline

When developers and publishers think of operating games as a service, what comes to mind are usually live ops, content drops and analytics. Product managers pour over reports to tease out player insights to make game and content changes; while live ops and lifecycle marketing teams look for nuggets to improve their events and campaigns. This post will address an overlooked issue - increasing numbers of QA and test users.

With more content drops, lifecycle campaigns and live ops events, the number of QA and testing users will increase dramatically.

How is lifecycle marketing different from live-ops

When Sonamine started to offer hands on services, we encountered confusion over what services meant. Customers asked if we were running user acquisition campaigns, or running live competitive service such as those offered by ESL Gaming. We explained that our services focused on enabling online interactions with players within the game. This excluded UA or in-person events. Even with that clarification, there was confusion over what these online interactions service meant. This post aims to clarify that question.

Hands on services for live ops and CRM

After a decade of offering player predictions to the game industry, Sonamine is expanding into services.  Specifically, any services involving live ops and CRM with game players.  Yes, our crack team of analysts and marketers will log into your mobile marketing and live ops tools, and set up messaging campaigns and AB tests!

What this means for Sonamine machine learning customers

Not only will you receive our predictive user scores for conversion, churn and revenues, but we will help you use these scores in live ops and lifecycle campaigns.  We will recommend the audience segments and create the messages, nudges and AB tests in your tool.  And we will create the reporting required to assess performance.  

In other words, we will get our hands dirty in the nitty-gritty day-to-day operations of turbo charging your existing live ops and lifecycle marketing with our machine learning scores.

What this means for new customers

Preventing churn in non-paying users with machine learning

Ever wonder how those special offers coming from mobile telecom providers are determined?  Almost certainly some aspect of these marketing offers involve a machine learning algorithm identifying which customers are likely to churn.   Game companies are beginning to take a cue from these industry leaders.  A previous study we discussed identified and effectively addressed churn in high value spenders in a game.  Today we get a chance to learn how a leading game company, IMVU, uses machine learning to address churn in non-paying users.  What follows is a condensed exchange between Nick Lim and Donnie Kajikawa, Senior CRM Manager at IMVU.  

Automated tools to produce and test AI models

The previous post on model drift highlighted the need to rebuild models as frequently as needed. How would this be done in practice? The most obvious path would be to have the data scientist rebuild the model, including generating the training dataset, model fitting, testing and comparison. Once the new model is completed, the next step would be to swap out the old model with the new model in a production environment.

Imagine now if the model needs to be refreshed on a weekly basis. Does that mean one data scientist is needed to maintain one model? After several rounds of model refresh and deploy, she would probably want to automate the process. Here's where automated tools come in handy.

Peek into real life model drift and retraining

We will start the new year with a series of articles about productionized machine learning. We will introduce a broad conceptual challenge and illustrate our solution. First up is model drift and retraining.

Model drift is the idea that a predictive model degrades over time. The usual culprit is that the real life environment is diverging from the modeling dataset. So the modeling dataset no longer matches the current environment. How quickly this happens really depends on the domain and the type of model. Sonamine focuses on consumer facing entertainment apps, and conditions there change rapidly.

Let's review a real life example, taken from a Sonamine customer for whom we are trying to predict a specific user behavior. It's a binary classification model with hundreds of thousands to millions of users. The data distribution is not heavily skewed with hundreds of input features. In short, it is a pretty generic classifier.

Audience Sync Service Live!

One common use of Sonamine predictions is to upload specific user lists, such as likely converters and high predicted revenue users, into user acquisition tools such as Facebook and Snapchat. Conceptually this allows mobile UA managers to find look-a-ilke audiences similar to these users with high predicted value.  As an extension of this capability provided to our Sonamine prediction customers, we are proud to announce the availability of the Sonamine Audience Sync service.   Now you can sync any audience, not just Sonamine predictions, to your Facebook custom audiences and Snapchat match segments.   Here's a link to the details.  Why would you want to do this?  

Integration with Facebook Custom Audiences

We have enabled integration with Facebook custom audiences. This feature allows you to synchronize Sonamine predictions with Facebook custom audiences.

How might this be useful? For example, you can now retarget Facebook users who are likely to churn out of your game. With lookalike campaigns, you can acquire FB users who are most likely to convert to paying users or have the highest predicted LTV. The list goes on.  

There are some intricacies to the synchronization.  First, we had to enable both the addition and 

Demystifying the AI hype for game developers

Even if you have been living under a rock, you would have heard of AI.  Why?  Because  developers use game AI all the time.  But the same term means different things to different people.  We will try to tease out the differences here.

Game AI refers to the programming that controls the behavior of computer generated characters, also known as non player characters (NPC).  As a general rule (sorry), game AI is a series of complex if-then statements.  These are usually pre-determined by the game designer.  The variables in the if-then statements can be dynamically generated or static.  Randomness can be injected within these if-then statements.  

Fixing biases and flaws in big data analyses of games and apps

Last Friday, I had the pleasure to attend an all day symposium titled "Societal Impact through Computing Research".  One keynote speaker was Ed Felten, deputy CTO of the United States.  To the audience, he posed three challenges regarding the use of big data and computing:

  1. how do we reduce the biases and flaws found in methodologies that are in widespread use?
  2. how do we incorporate notions of fairness and accountability into big data systems?
  3. how can we better secure our big data systems that will become increasingly mission critical?

Each of these challenges resulted in a lively and energetic discussion.  In this series of blog posts, I will discuss these 3 challenges from the perspective of analyzing the use of a game or app.  The reader should be fairly familiar with games, apps, mobile marketing and analytics.  Let's start with methodological biases and flaws.

RevenueSoon is now available

If you want to predict how much revenue each particular user will spend on subscriptions, micro-transactions and purchases, RevenueSoon™ is the product for you. As an integral part of the Predictive Player Segments Platform, RevenueSoon™ allows you to optimally allocate the appropriate user-acquisition budget and customer service resources.

What to watch out for in AB testing

A common cousin to game analytics is AB testing.  Many analytics tools such as Upsight are beginning to offer native AB testing capabilities within their suite. 

As a quick introduction, to run an AB test, you split players into two groups.  For one group (A) you do something with them; for the other group (B) you do something else. Something else could mean do nothing.  Then you compare the metrics of these two groups to see if there are differences. You can have multiple groups of course. For example, you give the A group some free gifts while leaving the B group alone; then you compare the spending and retention of the two groups.  If group A has 20% 14-day retention while B has 15% 14-day retention, we make the claim that free currency improves retention.

My 6 year old company Sonamine offers player predictions so our customers like to run AB tests to see if these predictions improve monetization and retention.  Here are some lessons learned from designing these A-B tests and interpreting the results.

Mistakes to avoid with Analytics

It's 2015.  Every game developer has heard about how important analytics and big data are.  So everyone's rushing into it; VCs fund new vendors such as Amplitude and Omniata; you are rushing to try out Amazon Redshift.  This blog post is about what to watch out for in analytics projects.  The points below are my personal opinions and are based on 6 years of Sonamine experiences deploying advanced predictive scoring, combined with 8 years at MicroStrategy, a BI tool used by the likes of EA and Activision.

Churn Prediction and Prevention in High Value Players

Many approaches to reduce churn are standard operating procedure in consumer oriented industries such as mobile telephony, credit cards and retailers.  These include  customer lifecycle management coupled with churn management techniques such as loyalty points and free gifts.  

Retention is critical in many free to play games.  Another way to say this is that developers need to reduce player churn.  So it may seem odd that  not many churn reduction approaches have been documented within the games industry.  This blog post will review a paper presented by the team at Wooga regarding one of their attempts at churn reduction.  The authors attempted to reduce churn by selectively offering certain players free in-game currency as a means to entice them to continue playing.   This work was done in collaboration with the Artificial Intelligence Lab at École Polytechnique Fédérale de Lausanne.

Improving user acquisition effectiveness with big data Part 2 - Using excel solver to allocate UA budget

This is part 2 of the user acquisition optimization blog series.  Part 1 covered how you can compare different acquisition channels more accurately by incorporating predictions into the data.  Channels really meant any source or campaign set up.  View a video on this website if you want to learn more about making user predictions.

Part 2 will build on part 1, and show how you will use this channel comparison table to efficiently allocate your UA budget.  This post is a hands-on tutorial on using the Excel solver, so be ready to fire up your own spreadsheet.  

Improving user acquisition effectiveness with big data - Predicting number of paying users by channel

Users are the lifeblood of games.  To get a constant stream of players, developers and publishers are spending huge amounts of money, driving up the cost per user.  The complexity is daunting as there are easily over 50 different places where you can spend money to acquire players.  It is a never ending game, no pun intended, to find the channel that provides the lowest acquisition costs balanced with high quality players.  

The question we will address in this article is how would we use big data to optimize the large amount of money that is used to fund player acquisition.  We will assume that you are able to attribute players to the different acquisition channels, either completely or partially.

Big data and Games Part 3 - Option value of data

In the Big data and games - part 1 - why ask why we described the characteristics of big data, how it will challenge the culture of asking and knowing why before acting. Big data and games - part 2 - algorithms and experts, and how management must utilize both these algorithmic experts and human experts  in the appropriate contexts.  In this final part 3, we will discuss the utility value of big data and games.  As before we will rely heavily on Mayer-Schonberger and Cukier, as well as the colossal "Games Analytics" edited by El-Nasr, Drachen and Canossa. 

First let us talk through how data is different from other types of assets.  Many resources such as coal or oil can only be used once.  

Big data and games - part 2 - algorithms and experts

In Big data and games - part 1 - why ask why, we described the characteristics of big data, and explained how big data would challenge the need for causation.  In this second installment, we will discuss how big data will affect the role of subject matter experts.  As before, we will draw on the experiences of Sonamine and the excellent book by Viktor Mayer-Schonberger and Kanneth Cukier.

Human subject matter experts?

Doctors are experts: they spend many years studying to gain a large amount of knowledge.  In their daily practice, they apply their knowledge to decisions in diagnoses and treatment, the outcome of which continue to hone their knowledge.  They have to keep up with new research and studies in order to refine their understanding of human body and diseases.

Big data and games - part 1 - why ask why

There has been quite a lot of hype around big data (BG), with every vendor adding it to their marketing slogan and tag lines.  Having been in data and consumer marketing for over 20 years, I was quite dismayed to watch the promise of big data following the Gartner Hype cycle...  This series of blog posts hopefully cuts through and clarifies the issues surrounding big data, and more importantly applies them to the digital games industry. 

Why most game analytics companies become ad networks

Have you noticed that there are a lot of analytics companies in the game space recently? They offer pretty looking charts and require you to integrate an SDK.  This article tries to answer the question of why many of them seem to become advertising companies.

For example, the following companies started as game analytics providers and then very quickly added advertising to their offerings:

Sony Online Entertainment leverages Sonamine Predictive Player Scores

Happy new year everyone!

There is a lot of industry interest in using data analytics in games.  Looking at the gamasutra or linkedin job board today, I see that Ubisoft, 2K games, MachineZone, Ngmoco, Microsoft, Z2live, Activision etc are all looking for data analysts.  Especially after it was revealed that the Democrats used a predictive scoring algorithm to allocate scare volunteer resources to get-out-the-vote, interest in using predictive analytics has never been higher, or more hyped.

Freemium games are not normal

Freemium gameplay behavior fits a power law

After analyzing several dozen freemium games, Sonamine discovered that the player behavior is better approximated by a power law rather than the more commonly known normal distribution.  For example, if you plot of the number of days since their first play against the probability of a user converting to paying status, you will get a graph that looks like this:


If this behavior followed a normal distribution, the distribution would look more like a usual bell curve. 

© 2009-2022 Sonamine LLC