Identify where players are leaving your game and fix it. Retention is all about quantifying player metrics, finding problems and forming hypothesis to understand and ultimately fix these problems. If User Acquisition is about funneling water into the bucket, retention is patching the holes in the bucket.
Vocabulary
Day 2 Retention – Day 2 retention measures what percentage of people return to you app. This is generally the fastest drop in retention rate. An extremely high-performing game like Clash of Clans has a day 2 retention rates of 60% (meaning 40% never open the game again). Other successful games have been known to have retention as low as 10% - 20%.
Day # Retention – Retention is often measured days 2 – 7 (the first week), then Day 14, then Day 30.
In the previous chart, you can see the biggest drop in retention is between Level 1 & Level 2. From this we can surmise that there must be a problem with this step, and can start forming hypotheses about why players are dropping out here.
Like what you're reading? This blog is updated every week! Subscribe on the right hand bar for updates via Twitter or Email.
Day Based Retention should look like this chart. The curvature should be smooth and flatten out over time. If you see in sudden dips in your retention—that’s where you have room to optimize. Most of the optimization opportunity occurs early in the lifecycle, because that’s where the most players are lost.
Secondary Levels – break down your retention numbers by hour, by play session, by achievements in the game. The more data you can gather about your players activity in your game the more data you will have for identifying and fixing sources of player abandonment.
Compulsion Loops – are in game cycles that keep players engaged. Each stage of a compulsion loop should enhance the next stage to keep players interested.
Compulsion loops are strongest when they are small and simple. Players want to know how to make progress and how to compare it against others. Too many loops will create confusion and dilute players’ engagement with the game. Loose cycles with a lack of player agency will create player apathy. The player needs to feel they are progressing through the loop because of their own skill & decisions.
If your game does not offer clear indicators of progress, consider adding an abstract indicator of progress—like points or XP.
Data Driven Approach to Optimizing Retention
Fixing a game costs time and money so it’s important to prioritize these tasks by their ROI. Fixes that will result in the greatest ROI should be prioritized first.
Step 1: Identify a Problem
Analyze saved reports to identify issues. Look at relative retention over time and over player achievements. Find the biggest drops and address those first. Define a success metric or KPI to summarize what results a fix for this problem should deliver.
Example: Day 2 retention is 20%, after analyzing data we find that the biggest drop in retention happens between the loading screen and completing registration (90% - 40% absolute or 44% relative retention). A fix for this issue should increase relative retention at this step to 80%)
Step 2: Form Hypothesis
Create a hypotheses how to fix the issue. Forecast the lift or improvement that would come from implementing your hypothesis and weigh that against the estimated cost & effort to implement the issue. If possible, seek additional data to verify your hypothesis, such as Player Surveys & Focus Groups. Once you feel confident that your hypothesis will solve the problem according to your established success metric, it’s time to move on to the next step.
Example (Continued): Players find the registration process off-putting. Having never played the game before they fail to see the value of taking a few minutes to register in order to play our game. Many players abandon at this step without ever giving the game a chance. Solution: Allow player to play the first level before asking them to register. Effort Level: 2, Estimated Lift: %60.
Step 3: Implement and Analyze Results
The development team should be able to give an estimate for the amount of time & effort that will be needed to implement the fix. Once they complete all steps necessary and it’s been tested for basic functionality, update the game and watch the retention numbers closely. Analyze the new data and restart this optimization process.
Early Game Retention
Early game retention is the most important. Players here are most likely to abandon your game as many are still evaluating it for the first time.
- A/B Testing should be done for everything
- Endless design iterations until everything is optimized to its maximum
- Game must be intuitive and easy to play
- Technical Stability is especially important to new players. No bugs or crashes will be tolerated.
- Difficulty must not drive away new players. Game should start easy but also offer challenges for advanced players to achieve.
- Other players are often the best content. Include social features to let players develop a community within your game.
Late Game Retention
Optimizations to late game retention will affect the lowest number of players, but these players are likely to be the most dedicated and easiest to monetize.
- Low Cost Content: Late game players have already consumed most of your game’s content. When they feel the game has nothing left to offer them they may move on to other games. Try to add as much low-cost late-game content as possible such as higher building and character levels. These changes require very little effort to implement.
- Events can drive engagement and provide novel experiences for advanced players who have already finished most of what your game has to offer. Good events should be repeated at intervals.
- New features and gameplay modes will keep late-game players interested, but are costly to implement.