Timo Riikonen

Perhaps the most important task for a management board or CEO in gaming company is to decide development and marketing budgets between games, both old and new. They can do those decisions based on their gut feelings and/or based on metrics. If they use metrics in the decision making, these are the most important numbers they should get for all of the games.

Before going to the list of actual metrics, one place where some game developers have a pitfall still is that they have not understood the importance of data selection: All politics analysts know that if they make a web query where they ask users on web questions on politics, they will get significantly incorrect results. This is because people who browse in internet answering to these types of queries do not represent very well ordinary people who go to vote. For some reason there are many game analysts that analysis basic on a weekday represents weekends as well. I think this is dangerous assumption to do and the high speed in game industry can affect negatively in doing analysis overly fast. So consider carefully when you do analysis on one day data and when you do it for a full week. Calculating a week complicates the mathematics, but I hope you are up to the challenge J.

1.     New users: active on days 14 – 41 (metrics)

It is easy to calculate and understand growth: you can calculate how many users logged in yesterday and compare that to four week earlier and the difference is the growth, positive or negative. But companies that focus on growth will change their products too fast and thus they will suffer in profitability. This is because growth numbers are harsh on older games with large player basis.

So instead of growth, you need to measure feed metrics: how many new players you get and contrary to most of the advice you may read from internet, you should forget how many players leave the game! Growth numbers are the metric #1 for capital investors, but if you have a CEO who thinks like a capital investor, you are fucked up: he is good for selling your company to someone to buy it, but he is not good for running it for long term. If you wish to run your company for long term, try to ditch your CEO or at the very least let him read this text in case it would help (right, fat chance?).

I suggest that you just term it as “new users” and hide all of the gory complexity that lies underneath the number, otherwise your board is likely to resist using a number they will have some difficulties of understanding.

In theory new users number should be higher for old games with large active player population, but since players switch so fast to new games leaving old games; this number is sufficiently equal in this sense.

The basic mechanism naturally is that if you don’t have new users, you will not have future paying customers. If you have a huge group of players, you should be able to generate revenue from them. Freemium games can be profitable if they are on niche market, but as the goal is to make big paybacks instead of just barely profitable, then feeding the funnel metric must be the most valued metrics.

Here is my suggestion for the measuring:

• Some games are played more on weekdays and some are on weekends, so it is best to calculate logged players from a full week. Also this mostly prevents drastic miscalculations in case a server has been down or someone inside your company is tinkering these numbers through external or internal marketing campaign. But on the other hand you should have fresh data, so don’t measure by default for more than one week.

• Then choose only relatively new users only: do not calculate less than 7 days because that will benefit games that have huge fuzz, but are not able to remain their players. If you have advanced setup, you can make a setting request that you want to calculate only logged in users that had created their user accounts 7 – 34 days before the login day.

• If your setup is relatively simple, you can choose fixed days for the login creation: So for example measure logged in users from days 1 – 7 to past, whose accounts had been created 14 – 41 days in the past.

• I have assumed here that you have a meeting once per month or every four weeks. So 34 comes from 7 + 28 – 1, 14 comes from 7 + 7 and 41 comes from 14 + 28 – 1. Naturally it is best to calculate same durations every time, so forget calendar months and use four weeks instead.

Because we want these numbers to be equal between games, I believe you can’t find any better choice than logged in players.

Note that we do not wish to calculate players that have tested the game and abandoned it. That value may be useful for calculating marketing success, but not so valuable for calculating game success. One week is a good length of required “gaming” time.

It is best to show this value in two categories: new users without marketing and new users thanks to marketing. This helps management to ignore partially the marketing effect. In case you are afraid that this gets too complicated, then you would better to just show numbers where marketing has been removed.

2.     Profit (metrics & estimation)

You can spend ages to calculate how many whales, dolphins and minnows you have and what is the percentage between them, but in the end it does not matter much. What matters is whether your game makes profit. So you need to calculate income and decrease development and maintenance. For games that are already active you should calculate estimate for the next three months. People are by nature either optimistic or realistic, so this is one point where official numbers should be made by a single person after he has discussed them with analysts and product owners. I recommend you to rather use an analyst for this than a Director as I expect analyst to have higher accuracy and to have less emotions involved with them.

As you can learn from economics basic course past spending is irrelevant at this point, so you will ignore everything in past and estimate only either the next three months (exceptions to duration are allowed).

There will be people who are going to argue that this metrics should be the most important and not on the second place. Those people have analytical mind, but they are not dreamers. And it is the dreamers who make the greatest successes, so don’t let them to win the argument and concentrate primarily on this metrics only.

Profits = revenue – maintenance – development – marketing

3.      Income € / marketing € (metrics & estimation)

Metrics/estimations #1 and #2 primarily tell which products you should invest to both in development and marketing. This estimation is quite difficult to calculate, but the attempt is to first estimate how valuable marketing investments are for this game. As the game ages these estimations get more accurate.

This estimate is based on previous two equations:

• If you get x number of new players, how much profit and expenses they will bring

• Marketing effectiveness is quite fast to calculate and you can reuse the active after a week number for that.

• Then in addition to this you need a complex Excel similar to what Nicholas Lovell has presented.

Naturally this value should be used for marketing budget:

• The higher the number, the more money you should invest on marketing

• And those games that have lowest numbers and less than 1.5 you probably do not wish to market at all

And if you do not wish to market the game, do you wish to develop it either?