By Chethan Ramachandran. This article originally appeared on VentureBeat, January 23, 2014.
This sponsored post is produced by Chethan Ramachandran, the CEO of Playnomics.
The world of mobile marketing moves fast, so why use outdated metrics that are no longer accurate representations of your application’s health? Here are five metrics you should be focusing on — and five you should lay to rest.
Five metrics that matter
1. User engagement: There are many different ways of measuring user engagement: return rate, session frequency, session length, time lapsed between sessions, intensity of play, average number of days played per week, etc. Regardless of app type or genre, when calculating user engagement, you should consider each of these submetrics and weigh them based upon what is most important to your business. For example, at Playnomics we developed the concept of an overall “Engagement Score,” which averages three subscores looking at attention, loyalty, and intensity. Marketers using our platform can quickly look at these scores for each application and understand how it’s performing.
2. Payback Days: Acquiring new users, organically or through acquisition buys, is an absolute must for any mobile developer. Even when spending thousands or millions of advertising budget on acquiring new users, knowing your cost per install (CPI) and optimizing channel performance, you can’t forecast what the exact return of investment (ROI) of each user will be. Enter the concept of “payback days.” When acquiring a new user, it’s important to know, within their first days of use, how soon that user will provide ROI. At a basic level this is calculated by looking at cost per user (or install) versus predicted user lifetime value (LTV). By knowing this piece of information, you will be able to make more intelligent decisions about which networks to continue to acquire new users from, and which networks you should avoid.
3. Predicted churn: User retention is a hot topic within the mobile industry because knowing when a group of users is at risk of churning not only gives you an idea of how your application is performing over all (is the content engaging), but it also shows how well your marketing efforts are working. After identifying a group of users with a high likelihood of churning, target them with content designed to increase their engagement and extend their lifetime.
4. Remaining lifetime: Going hand-in-hand with predicted churn, knowing a user’s predicted remaining lifetime, how many more days of use or play that user has left, provides you with an idea at how “sticky” your application is and offers you intelligent insights that can help inform decisions regarding additional content to provide to keep that person engaged. In certain instances, it can help with game design choices, identifying parts of the game where players are more likely to churn out.
5. Predicted lifetime value: Knowing how much value a user is going to add to the application over their lifetime is the holy grail of marketing metrics. It not only helps to inform acquisition buys but also game design choices, and marketing decisions. For example, if you identify a user as being low value, you can cross-promote them into another application where they may be more likely to monetize or serve them ads to monetize them. And it helps you provide VIP treatment and content that will increase engagement and retention to users that have been identified as high value — or have the potential to become high value.
Five metrics marketers should rethink or retire
1. App store ranking: With the constant changes in the way applications are ranked, and the increasing difficulty that lies in knocking a long-standing top 10 application off the charts, it no longer makes sense to focus heavily on how an app ranks. Increasingly, apps outside of the top 10, within any category, can develop a large user-base and perform strongly.
2. DAU and MAU: Daily and monthly active users (DAU, MAU) are the two metrics that have been around since the beginning of time. Though valuable in providing a snapshot at what your server load may be and a high level look at an application’s potential, they are not the only indicators of application longevity and success. DAU and MAU can fluctuate daily and monthly. Additionally, if an application has a high churn rate, but it isn’t receiving a consistent influx of new users, the app will quickly die out. On the flip side, an application can have a smaller DAU and MAU, but with a very low churn rate and high spend rate, it results more revenue than an application with twice the number of users.
3. Cost per install: A vital part of the user acquisition process, cost per install (CPI) is important, but only when looked at in reference to a user’s ROI and payback days. Which is better? Spending 25 cents on a user with a 10 percent ROI over 14 days, or purchasing a user for $1.50 with a 110 percent ROI over seven days? While the initial cost is going to be six times more, the payout is larger and more immediate.
4. Average revenue per user: Looking at ARPU or ARPPU (average revenue per paying user) only provides a snapshot of revenue at that moment, which is a very small part of the story. Because ARPU is based on a number of different factors, all of which can change day-to-day, it’s constantly in flux. With this being the case, calculating future revenue will be a shot in the dark. Instead of looking at ARPU, consider a user’s predicted lifetime value.
5. Retention rate: Retention rate is just one aspect of understanding user retention, and it should not be your only consideration when looking to improve retention through marketing campaigns and content. The story — and understanding it — becomes deeper, and the capability to affect a user’s retention becomes greater, when you know how users interact with the application. You should be asking in addition how often they return, how long they play for, and what the likelihood is of a user churning, and as noted before, folding this into your overall engagement score.
When guiding an application post launch to success, it’s important to look not only at the data of your users yesterday and today but also at how these users are predicted to perform in the future. This enables you to then make intelligent decisions about the type of users you should continue to acquire, and how you should communicate to encourage key behaviors for the longevity of your application.