What's the maximum you should spend making an ad?
One of the most vital parts of performance marketing profitability is understanding the maximum you should spend creating an ad
Teams often lionise every pound spent on Meta ensuring its efficient and effective.
But there’s a huge area that is practically ignored by almost everyone I know – that is, what is the ideal cost per ad creation?
If you’re an in-house team, you may not even be calculating this. Time often gets ignored inside startups (I know in all five startups I’ve worked inside, pre-Ballpoint, I never once saw time tracking in play).
If you’re paying an agency or freelancer, no doubt you’ll have a sense of it. But I imagine most of the time decisions come down to gut (“That creator is charging how much?” or “we want to make this look really special, we’re happy to invest in it”).
It feels like a nice-to-have. Just the cost of doing business on Meta. But in fact it’s one of the most important figures you can have as a chief growth officer or founder. If your agency isn’t talking about this with you, share this post with them and go and ask why.
This might sound like a nice-to-have question, but actually it’s the difference between your Meta ad account working or not working.
Why is this?
If you spend too much creating each ad, they won’t on average ever pay back.
To understand the answer to this we’re going to learn today a little bit about probability, ‘expected value’, and why poker players make excellent performance marketers.
Topics like this can easily stray into deep theory.
So fear not, I’m going to keep this concise.
And there is a huge practical output.
Background context
A small number of ads have hugely outsized returns
Open up Ads Manager, organise all ads by spend, and it’ll be dominated by just a few ads.
Meta's auction pushes spend towards whichever ad it thinks is most likely to convert
The result is a distribution curve that looks like those above from a couple of our clients. A huge longtail of ads that get effectively no spend, and a small selection that get a lot of spend.
Ads diminish in impact over time
Anyone who has managed ad accounts for a long time also knows that ad impact diminishes over time. But if Meta’s signals indicate it’s still the most probable convertor, it’ll continue to spend into it.
Some ads last longer than others, but at their core every ad has a lifespan and it diminishes beyond that.
And when you work back from there it explains why you need to create high volumes of ads.
This is why a high volume of creative is important
Looking at those two client charts above and the top 10 ads are spending 2,000x and 200x the median. One client has 1,500 ads in there, the other over 2,000.
Distribution is heavily, heavily skewed towards a select number of ads.
If the strength of an account is pinned on a small selection of domiannt, hero ads, but the number of those that reach that status is tiny, then that demonstrates quite how hard it is to predict success.
Our entire process is about improving the probability of that successful outcome, but at the very best it is still a coin toss whether an ad will go on to beat the median or not.
What this means for thinking through testing
As a result, we have to think of ads as a version of placing bets.
We don’t know what the result will be. No amount of data or insight can gaurantee that outcome.
And so one of the biggest levers we have is the volume of bets we make.
After all if one in 140 ads goes on to be a ‘top 10 ad’, then not only is volume vital, but so too is understanding the cost per creative.
Cost of ad creation becomes the straw that can break the camel’s back
Creating 140 ads is one thing.
But the cost of creating those is one of the big unspoken of things in performance marketing.
And this is where understanding a little about probability and expected value (EV) comes into the fore.
At a simple level think of it like this:
You should see the overall cost per test as the media spend + the creative cost
Say you do £300 on media, but £500 on creative (time, shoots, editing), then your cost per creative is £800
Identify the “expected value” of each ad
Identify the maximum you’re willing to pay for each creative.
What is expected value?
Expected value is calculated by multiplying all possible outcomes by their probability of occurring.
A simple EV calculation is this:
Let’s imagine we’re playing a betting game on a coin flip.
A coin flip pays £100 on heads, nothing on tails.
Expected value is each outcome multiplied by its probability, added together.
(50% × £100) + (50% × £0) = £50. The EV of this flip is £50.
That £50 is the most you should ever pay to play. It’s your break-even.
Pay £40 per flip and you make £10 per flip on average, even though you lose half of them.
Pay £60 per flip and you lose £10 per flip on average, even when heads comes up.
The skill isn’t predicting the flip. It’s knowing the maximum price worth paying to be in the game.
With ads, the price of the bet is media plus creative. The payout is what a winner contributes over its lifetime
Use our Maximum Creative Cost Calculator
We’ve built a calculator so that you can forecast these figures by yourself.
Here are the figures you need:
The share of your ads that become winners, workhorses, and duds
Expected lifetime spend for an ad in each bucket
Revenue per £1 of spend (aMER) for each bucket
Your gross margin before marketing costs
What you currently spend creating each ad
Josh Lachkovic is the founder of Ballpoint. We are a creative-first growth agency built for the brands looking to become the dominant players in their categories. If your goal is to triple your ad spend this year, while maintaining efficiency goals, then get in touch.