In Data analysis, Games Analytics

Dashboards, slice&dice and behavioural segmentation are all terms used within the general catch-all topic of analytics.  No wonder analytics is a confused term within the games industry.

And it remains an intense debate how publishers and developers can get value from analytics; what level of investment in data and the environment is appropriate and how to deliver insights that improve the player experience and moves the metrics significantly in the right direction.

It is clear that analytics is much more than dashboards tracking MAU, retention rates and revenues.

But this seems to be what most people would currently define as analytics.  However the real value is producing actionable analytics – taking decisions from exhibited player behaviours and changing the game.

If you are still making important design decisions or running A/B tests from gut feel and intuition, then you really have to ask yourself if your analytics environment is delivering

So what does the data really mean?  And is it actually contributing to the success of your game?

Consider this scenario: “Lots of players left the game at level 4 yesterday.  Let’s make some design tweaks and hope that we don’t lose any more players today’.   Sound familiar? This is a typical scenario where primary dashboard data is used as the sole barometer of game performance.  Yes, it flags there is a problem and tells you where the players left the game; but not why they left.

Thankfully the ambition of the Games Industry is much greater than this, and behavioural data insight is fast becoming one of the central pillars of successful game production.  Key to this shift is the focus on actionable analytics, and the ability to use all that rich player data to make accurate game design decisions, tailored to individual playing styles and based on real-time player feedback.

Making changes in this way takes the guesswork out of game design decisions.  But to reach this point publishers and developers must look beyond simple statistics and understand how their analytics environment can be enhanced to help make game improvements with confidence.

Here are some examples of how the analytics journey can go from simple stats to proactive player Relationship Management based on predicting behaviours and giving players the experience they require to maximise retention and revenues:

  • Average sessions per player is 1.2

This statistic is meaningless as it is not clear if this is good or bad; or how we increase the average sessions per player to 1.3.  Unfortunately many dashboard reporting tools contain a great deal of this noise.  Take some time to look across your dashboard reports and decide which ones allow you to make decisions.

  • 70% of session one players will not return

This is better as it leads to further questions:

  • How long is this first session?
  • What level mission did the player reach?
  • How did they respond to the tutorial?

Such questions deliver clear cohorts of users to enable you to understand actual playing experiences and the potential barriers to engagement.  The ability to segment users and slice & dice data is key for this level of insight.

  • 50% of Session One Players left because they didn’t understand the tutorial.  35% left because they ran out of resources. 15% left due to high mission failure.

Now we are really getting somewhere.

Through slicing & dicing the player data in a meaningful way you can then start to test design improvements with confidence based on exhibited behaviours.  During Session One, players can be now funnelled into different treatment groups based on their likelihood to leave the game and not return – easy missions for novices, fast-track experience for experts, gifting to players who run out of resources.

The true value in analytics is providing clarity in how players are responding to your game and actioning this information. Ultimately you need to understand the different segments of players in your game and make the user experience as responsive as possible.  Either by targeted in-game messaging, or by giving players options based on their competency and playing styles.

Development time in Beta is precious so having an analytics environment that supports decision making based on actual player insight is fundamental. It is all about making the right calls; working analytics hard to de-risk the development process, optimise your game and make players happy.

Recent Posts

Start typing and press Enter to search