Retention is one of three components that you use to determine LTV (lifetime value of a customer) and in many ways most important to the success of a product (and the most difficult to improve significantly after launch). Three weeks ago, I wrote about the central importance of lifetime value (LTV) to the success of your game and your company. This week I want to discuss retention, its importance and how you can improve it.

How to define retention

Retention is how often players play your game and thus, also, how long they remain active players. As with all the LTV metrics, different companies use different measures of retention to determine lifetime value.

There are several components of retention for you to track and roll into your LTV formula. The first is the user lifetime. That is, if you acquire a user today can you predict they will still be playing the game in a week, a month or a year. Related to lifetime is how many days or sessions a user will play. For a large game, this metric can present a database issue, as you may end up collecting and storing gigabytes of information.

The second is how often they return to your game in a specified time period. This can be 1-day retention, 7-day retention, 30-day retention or whatever interval you deem valuable for gaining a more accurate estimate of LTV. You only measure day n retention for players who come back on day n. So if you get a player who comes back d2, d3, d4,d5, d6, d8, d9, d10-d100, they will not count for d1 retention or d7 retention.

The other retention related metric you should integrate into your LTV is your churn rate, when a player quits and never returns (great Gamasutra article on understanding and predicting churn). This can be measured by the days of play in a certain period (say past 7 or 30 days).

There is also a camp of analysts that believe you should not measure retention by days but instead by sessions. They argue that measuring sessions (and using that for LTV) is less arbitrary and better measures intensity. It also allows you to get a critical mass of data sooner (rather than waiting seven days you would wait seven sessions).

The importance of retention

In short, if people do not spend in your games, you do not have a company. As discussed in my first post on LTV two weeks ago, retention is one of the three categories (along with virality and monetization) that determine the value of a player and thus justify whether they are worth the cost of acquiring. If retention is low or non-existent, there is no profitable way to maintain your game.

The easiest way to understand the importance of retention is by looking at a simple example (in this example, I am excluding virality to focus on the effect of retention). Say you have a game that monetize well (it generates on average $0.10 for every user every day they play the game, a $0.10 ARPDAU). If a player plays a game once and never comes back, you generate only $0.10 from that player. If they come back 1,000 times, you generate $10. That is without any change to monetization. Let us also assume it cost $1 to acquire a new player.

In the first case, with the low retention, even if you bring monetization up to $0.15 (which is well above industry average if we don’t look at Japan), you are not even close to covering your user acquisition cost. Even if you get an incredible ARPDAU of $0.50, you can’t justify advertising. Conversely, if you can improve the game so an average player comes back 11 times, the game is now profitable at that $0.10 APRDAU. If you get the player to have 25 play seasons, you are more than doubling your advertising investment.

Improving retention

You need to design your game or product with retention in mind (not only in mind, but in focus) as it is very difficult to improve retention for a product after it launches (or even once it hits beta). While there are techniques that boost virality and monetization post-launch, rarely can you move the needle enough with retention if the game was not designed properly in the first place.

When designing or improving your game, there are several things to focus on to optimize retention (many of these ideas come from a great Quora thread on retention):

• Get in their heads. Dan Ariely, the author of Predictably Irrational, was on the Board of Advisors for my first game company (Merscom). He gave a piece of advice that has really stuck with me and is the cornerstone of how I look at game design. For a game to retain players, the person must be thinking about it when they are not playing. When they are in bed or at lunch or in a meeting they need to be thinking about how their crop or pet is doing or what move an opponent might be planning in a competitive environment. This is the number one key to building a game with strong retention.
• Make a great first time user experience (FTUE). Create a clean, fun, quick experience. You will lose a good portion of your players in the first few minutes so you should design the game to make that time great and leave the player wanting to play more. Avoid a tutorial, teach the player through the use of quests so they are enjoying the game from the first minute (do you really want to replicate a lecture hall).
• Create a fun game. You can use all the tricks in the playbook but if the game is not fun people won’t stay in it or return to it. Fun is not easy, so you need to take a step back regularly in the development process and make sure it is a fun game, not simply that you checked all the boxes in your GDD.
• Balance correctly. I mentioned balancing in my discussion last week about monetization and it is equally important for retention. Players should not feel like they completed everything in one session; they need to have to return to continue progressing. Balancing is also how you “get into their heads” as I discussed above: creating the appropriate tension makes them think about solutions when they are not playing.
• Incorporate multiple gameplay loops and mechanics. Although a few games stay fun with only one core gameplay loop and no real progression e.g., Bejeweled Blitz, most games retain players and extend their lifetime by adding mechanics. Beeline’s Smurfs Village does a great job of this, adding a mountain where the primary mechanic is treasure hunting and new mini-games that sometimes take more time than the core gameplay. By bringing in new mechanics, players do not get bored after days, weeks or months.
• Craft a compelling fantasy. The game should be aspirational, with the playing feeling they are working towards something very special.
• Make it truly social. Sending an eggplant to a friend does not create a social experience. If you add meaningful social engagement, players will come back because they feel an obligation to help their friends or are actually interacting (or competing) with friends.
• Update regularly with good content. Players will usually consume much more content than you expect (see my blog post on the challenge of creating sufficient content). If they run out of content, they obviously will stop playing. By providing a steady stream of interesting new content, you are giving players a reason to come back regularly.
• Use email lifecycle marketing. By sending out emails to re-engage players who have left the game and also to encourage active players to come back, you can increase both short-term and long-term retention.
• Make it global. The majority of gamers are not native English speakers. While they may try a game only in English, they are more likely to stay in (and enjoy) a game in their native language.
• Do not forget timers. Give the players content where returning is part of the game loop. This mechanism is tied to the need to balance well that I discussed above.
• Leverage guild play. Although primarily seen in MMORPGs, this mechanic was suggested on Quora and is a great idea. It is a form of cooperative game play, where players form guilds and complete quests cooperatively but also socialize with other guild members.
• Incorporate in-game events. Events, like contests or time-limited quests, give players new things to do and a sense of urgency because the event will not always be there (and the opportunity to earn a related reward).

With all of these areas, you should A/B test before implementing changes to all players. Some changes that logically feel like they will improve retention may have the opposite effect. Also, a change that improves retention may have a negative impact on virality or monetization (or both), that offsets the positive change and lowers LTV.

Retention does not exist in a vacuum

There are two important lessons about retention. First, I want to reiterate how hard it is to improve late in a game’s development cycle. You need to design for retention from day one. Second, as with the other LTV metrics, they all move in harmony. An increase in virality might improve retention (because all your friends are playing with you) and an improvement in game balance could benefit both retention and monetization. Conversely, if retention in your game is broken, it is impossible to generate a good LTV.

• Analytics
• General Social Games Business
• Lloyd's favorite posts
• LTV