At Kolibri Games, our product decisions generally have three different origins.
We change our products based on:
- Community feedback
- Gut feeling
- Data
In the past, we have relied heavily on the first two: player feedback and qualitative decisions (essentially our own gut feeling). We only used data superficially and for around ten percent of our product decisions.
With this approach, we were able to grow Idle Miner Tycoon out of a student apartment into an incredibly successful game played across the globe. We made fast iterations and built a game together with the community, who rewarded us with loyalty and excellent app ratings.
Despite arguably being a cornerstone of our success, this approach alone wasn’t enough for us to unlock Idle Miner Tycoon’s full potential.
Almost four years after its release, the game is still one of the most successful idle games on the market and we believe that it has the potential to defend that spot for many more years.
In this episode, our CTO Oliver explains how we are unlocking IMT’s full potential by moving towards data-driven decision making with an initiative that we call “Data2020”.
The Project
We love to talk about our goal of becoming the most player-centric games company out there, about our signature player-centric approach, and how we manage to implement player feedback into our games in as fast as two weeks.
We talk about the efforts we take to read every single mail, comment and rating we receive for our games and why it pays off to do so and how we have made our games successful together with our players.
Yet there is one puzzle piece that we haven’t fully utilized yet: data.
We believe that adding a quantitative component to our decision-making processes will help us make the game even better and to unlock that unused potential.
Now, Data 2020 doesn’t mean that we are leaving player feedback behind or are becoming less player-centric. We are still reading every email, comment and rating for our games.
On the contrary, we are becoming even more player-centric by further improving our game with information on the behavior of the millions of players who don’t directly reach out to us with feedback.
The Goal
With this initiative, we are putting a strategic focus on making product decisions for Idle Miner Tycoon fully data-driven by the end of the year.
To achieve this goal, we are working hard on delivering the following key results:
90 percent of all our product decisions on IMT need to be backed by data.
That means that the question “why did we implement this feature?” should be answered with “because data told us to” 90 percent of the time.
Time to insight for all data-related questions needs to be less than one hour.
That means that whenever someone in the team needs data insights or has a related question, that person will receive an answer to that question in less than one hour. We need to ensure that we are extremely responsive here to not sacrifice our agility.
90 percent of all changes need to be validated with analytics.
That means that we need to test the vast majority of new features after implementation to ensure they are having the desired effect.
The Work
More concretely, to achieve our Data2020 goal and successfully leverage the respective key results, we are dividing our work into several initiatives.
We have drastically increased the rate at which we are gathering information and Insights about player behavior in Idle Miner Tycoon. Even though our data is fully anonymized, at mass it helps us to understand how, when and for how long players interact with certain features in the game. While we are getting answers to these questions through qualitative player feedback, we simply don’t have the resources or possibilities to reach the majority of our over ten million monthly active users in this way and come to quantitative conclusions.
We have started to conduct A/B Tests in an effort to find out which features or changes could have a positive impact on the game and how we could further improve the player experience.
Through CRM, we are improving our personalization and segmentation which, in turn, allows us to better interact with each individual player. We are making our game more dynamic in how it responds to player behavior, with the goal of eventually improving engagement and monetization.
With the help of Predictive Analysis, we are predicting future player behavior by analyzing the current state of our game and adapting it accordingly. How likely is it that a player churns tomorrow? If very likely, we will give the player a reward to keep them invested.
Successfully achieving all Data2020 key results and executing these initiatives won’t be an easy task, but we’re doing everything in our power, hiring talent to support our data and product teams, acquiring tools and building proper infrastructure to effectively hand and analyze billions of data events every month.
The Future
Data2020 will enable us to move even closer to our goal of becoming the most player-centric games company in the world.
Measures like effective player segmentation, highlighting new features and updates via push notifications, adjusting the game according to churn rates, or offer personalization will only make Idle Miner Tycoon a better game in the long run.
Through faster insights into data, we will improve the quality of our product decisions and eventually improve the quality of the individual player experience.
Data will play an important role for us from now on and will help us to turn a game for millions into a game for billons.
If you’re curious about our progress or how we are tackling some of the individual Data2020 challenges, make sure to follow our social media channels or check in to this blog from time to time.