Work / our own app
Ludo Arena
Meta advertising for our own game, judged on what players do after the install.
A multiplayer game with 30,000 players a day, acquiring more through Facebook and Instagram advertising, where an install is the cheapest thing to buy and the least useful thing to count.
The figures
- $0.03
- cost per installmeta ads
- 1.2M
- events a dayour app
- 30,000
- players a dayour app
In order
What we asked, recorded, learned and changed.
What we asked
Which advertising brings players who stay, rather than players who install? What does a player who comes back tomorrow do today that the one who leaves does not?
What we recorded
About 1.2 million named events a day across 30,000 players: a daily reward claimed, a private table opened with friends, a coin table joined, a tournament entered, a match finished. Each install carries the campaign and the creative that produced it.
What it told us
Players who claim the daily reward and play at least once with a friend in their first days behave differently from everyone else. Campaigns can be judged on how many of those players they bring, not on installs alone.
What we changed
Campaigns are optimised toward the actions that predict a returning player, and creative is tested against the same measure rather than against installs. An install now costs three cents, which is only a useful number because we can also say which of those installs stayed.
The product
The screens the events come from.
The lesson
An install is where the question starts, not where it ends. Three cents an install means nothing on its own; it means something once you can say which of those players claimed a reward, played a friend, and came back. Advertising a product you measure lets you buy the customers who stay and stop paying for the ones who do not.