Don't miss
  • 2,006
  • 5,500
  • 5,765
  • 120

Improve your revenue with the GAMESbrief free-to-play game forecasting spreadsheet

By on October 6, 2011

Forecasting revenues for online games

I am a great believer in the “Keep it Simple” principle. I am often asked which metrics I consider most important. Clients and developers seem to expect me to rattle off a list of dozens, scores, maybe even hundreds of metrics that they should be tracking, analysing and tweaking.
The problem is that very few companies have the skills, resources or discipline to analyse this much data. They might capture it, but they are unlikely to use it. It might as well not exist.

Enter the GAMESbrief online games spreadsheet. It’s a simple, top-down approach that identifies the Key Performance Indicators of an online game that lead to revenue. It follows the funnel from users to revenue. It identifies the key steps along the way. It acts as a guide and as a development tool. It can become the heart of your business.

The metrics

I focus on six key metrics (Note that these links are to posts that will go live during October-November 2011 so you may get a 404 error at the moment):

  1. Monthly Active Users (MAUs)
  2. Daily Active Users (DAUs)
    • Engagement rate (DAUs/MAUs, derived from MAUs and DAUs)
  3. Retention rate
    • Churn (derived from Retention Rate)
    • Duration (derived from Churn)
  4. Conversion rate
  5. Split between whales, dolphins and minnows
  6. ARPPU for whales, dolphins and minnows

The objective is to keep the spreadsheet tracking the fewest number of KPIs that still enable you to get a good grasp of your business. I could add more (in particular, CPA and virality are notable omissions), but I’m not yet convinced this would help most people. The six metrics I’ve identified will help you identify where your game is performing above market benchmarks and where it is performing below them. The value of the spreadsheet lies in its simplicity.

How to use the GAMESbrief online game forecasting spreadsheet

When you download the spreadsheet, you will see that your online game is forecast to make $3 million in year one (because all online games can). You just need to go through each major metric and adjust the assumptions to reflect your own reality. Some of the metrics will vary by platform. (MMOs often convert better than social games and have better retention rates, but fewer MAUs, for example).

Improve the revenue of your game

Once your game is live, come back to the spreadsheet and compare live, actual data with the forecasts. Check which KPIs are above the market benchmark you would expect for your type of game and which are below market. The next step is key. Ignore the metrics that are above their relevant benchmarks. You need to focus your development team on fixing the areas where you can make the biggest difference fastest. So identify the problem areas and work on those. This is where the bad news starts. I can’t tell you exactly what is wrong at this point. You might, for example, see that your retention rate is very low. I have many ideas for how you can improve retention, but I don’t know what exactly is wrong with your game. Areas that might be an issue include:

  • High bounce rate: Lots of people land on your game, take one look at it and conclude “this isn’t for me”. (More common on, for example, a Facebook game where installation is easy than on a free-to-play MMO with a client download)
  • Early drop-off: Traditional game designers often use tutorials to set the scene. Casual/social gamers get bored and leave before they’ve even seen the game
  • Interest peak: The game can keep people interested for a week, but no longer

This is where you need to move from the GAMESbrief spreadsheet (which is a management/forecasting tool) into your live data. I’ll be writing further posts on how to dig deeper into your data, but the principles are:

  • Identify a problem (say, low retention)
  • Come up with a bunch of hypotheses (like those above)
  • Query your data to see if the hypotheses are true (e.g. if 80% of people take one look at your game and leave, you have a high bounce rate; if 50% of people don’t get to level 3, you have an early drop off)
  • Spend your next Agile development period fixing those hypotheses that are confirmed by your data
  • Repeat

Month on month, you should start seeing improvements to these metrics. All of these metrics point to revenue at the bottom line, so if your metrics are improving, so should your revenue.

What are the benchmark metrics are for social, mobile and online games?

That’s where GAMESbrief comes in. I get a lot of data about different games metrics. Whenever that information is public (revealed at a conference, embedded in a Slideshare deck, referenced in an interview), I’ll link to it. Each of the key metrics will have a reference page here on GAMESbrief. Every time a new datapoint is made public, I’ll add it to the page. Over time, I hope to build a comprehensive set of benchmarks that show, for example, what conversion rate to payers you might expect for a Facebook game, for a browser-based casual game, for a free-to-play MMO, an iPhone game and many more. You’ll be able to get a rapid sense of how you should be benchmarking your game by checking these pages regularly. (Oh, and I’m not arrogant enough to think I’ll capture all of the public references. If you see something that you think is relevant, add it in the comments or email me at [email protected]).

The numbers will be wrong

The numbers that you choose to put in here will be wrong. Being wrong doesn’t matter. Give it your best shot. When you launch your game, some of the numbers in your forecast will be above benchmark; others will be below benchmark. Your job over the next 6-12 months is to refine and iterate until every metric is above the market benchmark. That’s what this spreadsheet is for. It is a tool to help you reach profitability. It is not a magic formula that will tell you exactly how much money you are going to make. Otherwise every single online game in the world would make $3 million in the first year. And we all know that’s not true, don’t we.

Where to go from here

You should download the spreadsheet, and read the rest of the posts about the metrics – linked below. It will be a valuable addition to your arsenal. I hope you find it useful.

 

CLICK TO DOWNLOAD THE SPREADSHEET

About Nicholas Lovell

Nicholas is the founder of Gamesbrief, a blog dedicated to the business of games. It aims to be informative, authoritative and above all helpful to developers grappling with business strategy. He is the author of a growing list of books about making money in the games industry and other digital media, including How to Publish a Game and Design Rules for Free-to-Play Games, and Penguin-published title The Curve: thecurveonline.com
  • David Barnes

    Very good idea.

  • http://www.gamesbrief.com Nicholas Lovell

    Thank you. I aim to keep improving it, while striving to keep it simple. And encouraging people to sign up to the blog when they get it too.

  • http://twitter.com/alexisbonte alexis bonte

    Good idea but you seem to have gone pretty high on the retention rate (75% Mom!) and can’t find the article you link to about retention. So in your experience what is a good retention rate after 30 days for: social game, browser MMO, download MMO?
    Old rule of 100 new users, you retain 10, you monetize 1 is now pretty off in my experience.

  • http://www.gamesbrief.com Nicholas Lovell

    I have gone high. My experience is that between 2-6 months is viable, so I plumped for squarely in the middle.

    This is – obviously – an iterative model so if you have any evidence that the retention rate at 75% is way high, I’d love to hear it.

  • http://twitter.com/VexingVision Björn Loesing

    I’ve modified the spreadsheet to include traditional seasonal impact on retention, revenue and paying user conversion. This makes it a lot more complex, but also a lot more viable.

    I have yet to see a retention rate of 75% outside the second month of a subscription based online game. Depending on the quality of game, 60% is considered a success and 40 to 50% is normal.

  • http://www.gamesbrief.com Nicholas Lovell

    Thank you. I’ll have a look a look. Of course, the spreadsheet works on “month 1, month 2″, not actual months, and all games slip, so I need to think about how to handle that.

    In practice, seasonality matters a lot; for initial benchmarking, I’m not so sure that it is key

  • http://www.gamesbrief.com Nicholas Lovell

    On the retention rate, thanks for that data. of course, in this case, the game is free-to-play, so someone doesn’t have to decide to fork out cash every month.

    if you are right, the average duration of every online game is <2 months. I wonder how you start to calculate the "average duration for people who played for more than 1 month"

  • http://twitter.com/VexingVision Björn Loesing

    I calculate “MAU” and “REAL MAU” – most retention rate is diluted by players who log in once, look around for half an hour and decide the game is not for them and never look back. These haven’t been the target audience anyway, so counting them in for retention is not necessarily what you’re looking for in the first place (although the number is very useful in determining how accessible/boring the initial gameplay/tutorial is).

    Instead, when looking at the REAL MAUs, I take people who have logged in at least twice on two different days. These are the ones who had an initial interest in getting invested into my games.

    Hope that helps!

  • http://www.gamesbrief.com Nicholas Lovell

    It does. But there are two approaches to forecasting:

    One is the “Keep it Simple” approach. How do you help a startup or a team considering making a fre-to-play game build a model to help them understand the key drivers that matter (leaving aside how accurate they are) and how they influence revenue.

    The second is to try to use the power of Excel to try to reflect reality.

    Your suggestion of MAU/Real MAU is excellent, but not simple. It fits the second case, not the first.

    i am considering a much more complex model (not until next year though) taking in all the feedback about MAUs, CPA, retention rates, seasonality and so on.

    I still think that, for many purposes, the really simple model I started with is a better tool for the job.

  • http://twitter.com/VexingVision Björn Loesing

    Agreed on the MAU/Real MAU issue being too complex. For me, when analyzing biz potential of a game, I would not be able to get an even remotely accurate calculation without at least changing the retention rate and revenue rate from month to month. It just doesn’t work that way – Christmas Sales Booms are occassionally 600% of the July-Holiday-Everyone-Is-Gone-Depressions due to significantly higher PU and much much higher ARPPU.

    It doesn’t really even out, and any (even basic) game revenue forecast should take the major fluctuations into account.

    Also content updates are a major drive in retention and reactivation and should be accounted for. Retention is lower in months without major content updates, which in turn affects the later months. It’s hard to recover from a sudden drop in retention mid-year to get back on track when projecting continuous growth.

    Maybe it’s just me being German and detail-obsessed, though. There’s always that.

  • Pingback: Games Brief to help game devs forecast freemium revenues | The Appside

  • Heather Stark

    Sensible advice about keeping it simple,  hypothesis-testing, and reality-checking as real data comes in.   Gotta argue, though, about the value of using DAU/MAU as an engagement metric.  Because of the way it is constructed,  DAU/MAU is not and will never be a direct measure of engagement.  And as a proxy measure it can be misleading, even catastrophically so.   We seem to have got into this weird state where people are trying to replicate the data Facebook burps out, for its own sake – when what they really need to know is how engaged their players are.   Which DAU/MAU simply can’t tell you.   For your own data – you deserve better.   Measure engagement, by all means.  But do it directly.   Then you’ll really know who’s doing what.  And you’ll be better able to guess why.
    For the nitty gritty detail, check out:
    http://insightanalysis.wordpress.com/2010/07/21/facebook-dau-and-mau-what-they-tell-you-and-what-they-dont/

  • Pingback: Improve your revenue with the GAMESbrief free-to-play game forecasting spreadsheet – Games Brief | Logicamp's Blog

  • Pingback: What share does the platform take? | Nerd Gamers

  • Pingback: CPA, CAC and Customer Acquisition costs | Nerd Gamers

  • Pingback: How many downloads? | Nerd Gamers

  • Pingback: 为什么开发商必然会开发Android游戏?又存在哪些挑战? - 雷锋网 - 专注于移动互联网创业&

  • Pingback: 来做爱 - 开发商必然会开发Android游戏的七大原因

  • Pingback: 开发商必然会开发 Android 游戏的七大原因 » 全信息

  • Pingback: 获利比IOS更丰厚?开发商必然会开发Android游戏的七大原因 - AKEPC-便携数码志 AKEPC-便携数码志

  • Pingback: 开发商必然会开发Android游戏的七大原因 | 绿卓网

  • JoeJoe K

    Does the spreasheet assume that New Users = MAU and that a new user will only use it once, never returning. Where else can I dig up figures for New Users, MAU and DAU if the info is not public?

  • rupazero

    We collect MAU and DAU figures on their respective pages: http://www.gamesbrief.com/2011/10/monthly-active-users/ and http://www.gamesbrief.com/2011/10/daily-active-users-daus/

    As for new users and returning users, this aspect is covered by retention, churn and/or duration: http://www.gamesbrief.com/2011/11/retention-rate-churn-and-duration/

  • JoeJoe

    thanks!

    another question: based on this spreadsheet, in one of his decks he states that whales are .5% of the users … how did he mathematically derive that number

  • Cari

    Nicholas
    Just learned of your website, downloaded the spreadsheet and read a few chapters of your book.
    What a wonderful service you are providing.

    Have you done any research in the educational games arena?
    Thank you for your wisdom.
    Cari

  • Aaron Cammarata

    Hi Nicolas –

    Fantastic sheet – thank you! You’ve saved me a lot of wheel-reinvention. Great site!

    I’m curious if you have seen any data points to support organic growth. Specifically, what percentage boost are we likely to see with organic discovery, and what is it a function of? For example, if we do a boost campaign, but are not feature by Apple, is it likely that, say, 1% of new users will be “heard from a friend” referrals? 5? Is there an ongoing additive effect that is a percentage of active audience, for example? Would love to hear your thoughts on organic numbers, as I couldn’t see it anywhere on this site or elsewhere (fiksu, etc.)

    Thanks!
    Aaron