Data-driven development in mobile games

What is "data-driven development"? To put it in a nutshell, it's when you try to get away from thinking what works inside your team and let the data/metrics tell you the answer. 

How could this be used while making your indie mobile game?

When making a game, it's almost impossible to predict what players are interested in. You need to have some type of metrics and/or feedback to follow. It's not obvious that people who play your game, drop feedback or say anything why they stopped playing and even why they play the game for a lot of hours.

This is the moment when the analytics step in!


For mobile games, it's rather "straight forward" since you need the analytics to tell the retention of your players, time used within the game, how much each player generate revenue and so forth. When you release (or soft launch) your game, you can start gathering these metrics when the first players jump into the game. Of course, in most cases the players don't just "appear", you need to do user acquisition to do this. Since user acquisition is an art form of its own, I'm not going to go too deep into that within this article. (You can read some tips & tricks from our previous post)


For UA (User acquisition) you have your own metrics to follow. So basically what type of paid advertising works and what type of changes affect the metrics. The more you have the budget to try different methods, the more possible it is that you start to figure out what works and what doesn't. 

It's a long process to figure out the best adverts, which "sticks" into your core group. If you don't have a marketing budget for your mobile game, I highly suggest you to start contacting publishers and/or figure out ways to get the necessary budget. 

(Read more from our post: "How to get players for your mobile game?")

Here are the main things which you should focus on, when starting to iterate your game to get more players and how to know which parts you need to improve:


First, you need users to get into your game store page (Google Play, Appstore). You can easily see from the store analytics how many users visited your store page. Google Play for example enables A/B testing for your store page. What this means is that you can create multiple different store pages with different promotional materials and see which one of those works best. You should be having multiple different A/B tests on going all the time, even when it seems that your store page is getting a good amount of downloads. When the download numbers start to increase, you can start making minor tests for your materials, like changing the order of screenshots, changing a couple of words from your game description and so forth.

You need to reserve enough time (a month) for each A/B test to get enough data, then you're more able to see the actual results of your test.


When you start to get enough downloads for your game, you can start looking at the in-game analytics. The first metric to focus on is the first ten minutes of gameplay. Is your player playing the tutorial through, did he get stuck, did he leave and never come back? If you can see that the players quit the game before ending the tutorial, you have a clear plan to change the tutorial of the game.

Do not think that this is a bad thing. You already have proof that someone was interested enough in downloading your game, but the game just didn't feel good. Now you have all the proof to focus on making the game better, and not every part of it, just the tutorial part where most of the players left.


Once your players start to stick into the game and play it for multiple hours, you can start looking at the retention numbers. So how long exactly do the players stay and when do most of the players quit? Is it at level 50? Or maybe most of the players don't get the "Castle to level 6". When you're able to identify these type of issues, you now have another clear issue in your game, which needs work. Make update and see what happens!


If you use paid advertising (UA) to get players (which you should be doing) your plan should be the following:

  1. Use X amount of budget to get Y amount of players into the game
  2. Work with an update for 2-4 weeks
  3. See the analytics how long did the players stay within the game
  4. Identify core issues which need to be fixed
  5. Make an update to fix these issues and publish the new version
  6. Repeat the process from step 1

By following this "rotation" you will most likely start getting "loyal" players for your game, which stick around and even start using the money for in-app purchases and such. 

(NOTE: Please understand that if you can't get users to visit your store page and install the game and/or play it for more than 10 minutes even after several updates, this may be an indicator that you are working with a game which is just not interesting enough. This is harsh, but it's good to consider that it may be the time to "kill your darling" and move into a new project.)


Always aim to get analytics and/or other metrics to prove that you're working with something, that (even some) players find interesting. Also, aim to get this data rolling as soon as possible when starting to work with your game. This helps you to see a longer period of data and maybe even help you understand what type of "announcements" or promotional tricks got you more visibility, than others.

Also, the data which you gather is the best possible stuff to show for possible publishers and/or investors which you may be looking for your game. Without the data, you're just making predictions, assumptions and wild guesses.

I hope you enjoyed this article and if you did, consider sharing it with your friends! Also, subscribe to our newsletter to get notified when we publish new content!

Have a great rest of the week!


Consider supporting us by pre-ordering our upcoming game Soulrun right here. Pre-orders get the "Supporters Pack (12,99 EUR)" which contains unlock code for in-game content. PRE-ORDER SUPPORTERS PACK HERE 12,99€

Join our community of game developers!