Facebook and Instagram ads that pay past the install
An install is the cheapest thing to buy and the least useful thing to count. We buy the people who are still there next week.
The cost per click looks brilliant. The cost per result looks brilliant. Revenue is flat, and nobody can explain the difference.
What changes
What you can say afterwards.
Which creative sells
Judged on customers rather than on thumb stops. The best performing video by engagement is very often not the one that brings buyers.
What an audience is actually worth
Measured after the click, so an audience that converts cheaply and churns immediately stops looking like the winner.
Where Meta and Google overlap
Both platforms claim the same sale. We count it once, and the honest split usually changes where the next pound goes.
The difference
Counting attention, and counting customers.
We ran this on our own game before we ran it for anybody else. Ludo Arena buys players through Facebook and Instagram, and an install told us almost nothing.
Optimised for results
- Cost per install falls, and the team celebrates
- Creative is picked by watch time and comments
- Lookalikes are built from everyone who ever converted
- Browser pixel only, so blocked traffic is invisible
- Budget follows whichever ad set is cheapest today
Optimised for customers
- Cost per player who is still there on day seven, which is a different league table
- Creative is picked by what the people who stayed watched first
- Lookalikes are built only from customers who paid, and refreshed as that list changes
- Conversions API alongside the pixel, so blocked traffic is still counted
- Budget follows the ad set that produces customers, even when its click cost is higher
What is included
The monthly work.
- Account and campaign structure that does not fight the algorithm for no reason
- Meta Conversions API, with values sent server side and de-duplicated against the pixel
- Customer lists built from people who paid, refreshed rather than set once
- A creative test running at all times, with one variable at a time and a written result
- Instagram and Facebook placements judged separately before being merged
- A weekly note: spend, customers, cost per customer, what changed and why
We buy players for our own game at three cents an install. The number we actually manage to is what those players do on day seven, because the first number is easy to make look good and means nothing on its own.
Ludo Arena, our own game
Questions people ask
The ones that actually arrive.
Do you make the creative?
We write and cut what is needed to test, and we are honest that we are a measurement and performance studio rather than a film crew. Where a brand needs production beyond that, we will say so and work alongside whoever does it.
Our audience is not on Facebook.
Sometimes true, and worth checking rather than assuming in either direction. If the arithmetic says the channel cannot reach your buyers at a price that works, we will tell you and point the budget somewhere else.
Is the Conversions API really necessary?
If a meaningful share of your traffic blocks the pixel, yes, because every sale it misses is a sale the bidding never learns from. We measure the gap before recommending the work.
How much creative do you need from us?
Less than you think to start. Two or three honest assets beat a dozen polished ones, because the first job is finding out which message works, and that is cheaper to learn roughly.
Read next
Your turn
Bring one question you cannot currently answer.
30 minutes. We will tell you whether it is measurable, roughly what it would take, and whether we are the right people for it.