Game developer must read metrics

Mobile gaming is a huge global opportunity, with output reaching 9 billion dollars in 2012 and continuing to grow.

With the 2015 smartphone user growth doubling on today's 1 billion basis, and the game taking up 66% of all application benefits, the potential geometry is known.

Game development continues to have a bright future, but it is limited to those who can develop profitable games.

The pursuit of profitability has now become a science, and monitoring analytics and ongoing A/B testing has become one of the key elements of game development. In fact, Zygna, the gaming company, is often classed as a large data company as a contributing factor to the use of analytics, if not an introduction.

We can think of metrics as "joystick", where the game developer creates an ideal result by pushing or pulling. Ultimately, the goal of a game developer is to find the best control mix to get the best financial results at the lowest cost. Creativity, novelty, and fun all have to be included in this prism of analytic learning: As long as the "analytical science" continues to improve, these three elements will have a sustainable development space. This is the basic philosophy behind the development of modern games.

For those who want to learn more about game Analytics, I set up a simple set of analog numbers to compare game performance with predictive retention and propagation. This is a simple spreadsheet form that you can click to download. I'll list the assumptions that generate these virtual values, and I'll explain some of the notable results from the numbers.

Suppose a game studio has 6 games:

The 1th game as a contrast, the other game is basically only one indicator of the difference with the 1th game. For example, the number 2nd game differs from game 1th in terms of average player playtime (compared to other indicators). I used an average CPA (user acquisition cost) of 1.3 dollars to calculate the cost of all the games. Finally, in order to prevent the definition of terminology from being vague in my following article, the explanatory description can be found in the spreadsheet.

Now let's look at the general conclusion of the simulation model.

1 The longer the average player returns in the game, the higher the DAU (daily active user) value.

Since Dau is the approximate measure of a player's participation, it also has a direct correlation to the revenue generation, and the average player's return can be seen as having a significant impact on the game's potential for gold uptake.

From the column in the Excel table downloaded above, named relative income (comparative Revenue), data from game 1th and number 2nd show that, compared to game 1th, the higher average player's game time for the 2nd game has brought extra dau and benefits.

2 game D2 (2nd day) The higher the retention value, the higher the Dau value.

As explained above, the Dau is directly related to the yield. Therefore, it can be inferred that the D2 retention value has a very obvious effect on the gold absorbing potential of the game. Also for this reason, many game companies run A/b test to optimize the retention value of the game early in the release cycle. At the same time, since we don't normally optimize games with D2 retention values under 10%, games with a lower retention value will soon weaken. Let's take a look at the comparison between game 1th and game 3rd in the table: the latter's higher D2 retention value makes its dau value better, so it certainly brings more revenue.

3 large advertising budget will not improve the profitability of the game.

That is to say, if a game is not doing well among a group of players, it is not helpful to advertise for more similar users to improve profitability. This is also the game company will hit the heavy sums to attract more eyeballs before, the first optimization of the game indicators of the reasons, but also explains why some games have not been completely online on the halfway.

Interested readers can look at the performance of game 1th and number 4th in the "Relative benefits" column.

4 The higher the spread of the game, the better the profitability.

A high degree of communication will bring more ad-free users to the game, with the consequent reduction of an important expense: The user gets the cost. A similar effect can be achieved by having a fixed player network that can be used to cross-sell other games at minimal cost. In the previous case, the spread of the entire player network to achieve their own expansion.

Overall, the ability to get users without cost is extremely important to the financial health of any gaming company, and it is no surprise that Mark Pincus in his recent performance briefing that the promotion of Zynga's player network will be the cornerstone of the company's future strategy.

As mentioned earlier, avid data enthusiasts can look at the comparison between the 1th and 5th games in the "Relative benefits" column, and see that the obvious difference between the ECPA of the two games (the cost of effective user acquisition) is the distinct K factor ( The percentage of users who are attracted to the ads in other games and who eventually install the game.

5 High user monetization value will bring higher profitability.

The result is simple and straightforward, but it can be more than that. That explains why gaming firms flock to 43-year-old housewives or 28-year-old male players, why is operator billing so popular in the rising markets of South Asia, why real money-line gambling is on the rise, and why game of Candy smash Legend (Candy Crush Saga) on the cross promotion platform.

Let's analyze game number 1th and 6th with a relatively high arpdau (average daily active user earnings) value. The difference between the total benefits of these games is a good illustration of my point of view.

This includes the results in my electronic simulation form. A lot of the results are very intuitive, and then look at my virtual data, there will be a deeper understanding of the basic game analysis. After all, for a game developer, they should strive to create a game with high average player returns, high retention, high transmission, and high aprdau values.

So the key to this table is not to identify the obvious facts, but to let people know that nothing can really ensure that a game achieves the desired target. What the game company can do is set up internal programs and channels, like the following, to achieve the best in a game with an ideal index.

Rapid prototyping and game testing: this is critical to quickly measure the retention value of a game before trying to build a mature product. Otherwise, you'll end up with a lot of game design that's not worth putting into practice.

A large number of A/b Tests: High strength, a large number of A/b tests are important in the life cycle of a game, because even small bugs can have a direct measurable impact on profitability in the analysis.

Frequently updated approach: a stable "content update path" for the increase in average player game time is indispensable. Once the game company is committed to creating a game, it is necessary to view the game as a permanent project in the process.

Big-game companies are already following the basic principles mentioned above-usually small studios insisting on informal methods. This is not a good thing, because informal methods not only can not help small studios catch up with large companies, but will be the gap between the two widening.

The vast future of the mobile game market is constantly moving the nerves of the rich. For today's game developers, the most fundamental principle is to implement the overall architecture of the basic data analysis (rather than trying to cater to some of the targets of illusory blindness). Their job is basically to create digital entertainment products that activate the maximum number of highly-propagated users every day and extend the product cycle as much as possible.

(VIA TC Note: The author of this article, Hassan Baig, is an entrepreneur who manages a gaming company founded four years ago in Pakistan: White Rabbit UBM.) )