+6 votes
I often find there is this chicken-egg situation when creating assets for performance marketing. Do you have a specific number of assets you aim to make per $X on media spend? Or do you find that too constraining? Is it just more about keeping the process running, aiming to come up with X amount of assets each week following your weekly creative review?

We are currently evaluating our creative resourcing and we could scale up endlessly, so I'm looking to find a benchmark of some sort to work towards, but of course we'll always test and adapt the process as we go.

FYI we have 9+ mobile games running at scale, so we're already churning out a lot of content, my assumption is that we are under-producing though.
by (190 points)

3 Answers

+1 vote
Creatives ads are one of the most crucial aspects of performance marketing. It's important to adapt on the go and do not stick to a specific amount of X creatives per $Y in media spend.

It very much depends on which stage you are in soft launch, are you trying to maintain the spend or scaling up? How fast are your creatives are downgraded or need to be refreshed? how many channels you actively promoting? Is the 15K daily, weakly, monthly budget?

I'm very hesitant to state a number of X creatives per $Y without context, however, without a performance baseline, I'd recommend starting with around 5 concepts and create variations of the best-performing ones.
by (180 points)
thank you! This is for a mix of soft launch, games at scale and products in growth phase - so more of a general question / average but I totally hear what you are saying. We were hoping to back out recruitment plans to hard numbers, but per Eric's response I think the focus should be more on creative fatigue rate.

Thank you! I appreciate your insights!
+1 vote
This is probably not the best way to think about how to make creatives. I would mostly recommend the multi-armed bandit approach. You start with a couple creatives, run campaigns on those, keep the good ones, and continue iterating on why those are good.
by (1.2k points)
Thanks! That's definitely our current approach to it, but as we scale up and incorporate more titles into our portfolio we're trying to staff the creative team accordingly, so looking at # of assets x time taken to build = number of artists.... obviously we'd build in slack to their schedules so that additions / emergencies / additional iterations on the fly can be factored in.
Was hoping to be able to back out a staffing plan based on numbers, but may need to be more of a guesstimate :) thank you though!
I think you could continue to operate this way. Basically when your creative teams are working around the clock to deliver solid creatives you know you need to hire more and that they are operating at max capacity. Every studio is different in the way in produces optimal content.
0 votes

I'm not sure that "X creatives per $Y in spend" is the best way to look at the creative production cycle.

There definitely is a cold start problem with creative at launch: you need some set of creatives to launch a new app with, and you want enough of those to ensure that you've explored / experimented with a number of different themes and don't get stuck at a local maxima with respect to performance. But after that -- once you have some baseline performance data and you are either trying to maintain spend or scale it incrementally month over month -- then you can intuit the volume of creative you need on a weekly basis just by watching campaign performance.

I wrote about the ad creative lifecycle and the performance "half life" of ad creative; if you need to assign a volume of ad creatives by $X spend for planning purposes, you can back into that number just by watching for how quickly your ad creatives start to degrade and using that to come to the half life value. But the reason I'm reticent to say it's useful to think of creative production in dollar value terms, at least as a generalization without any context, is that I've seen some creatives for very broadly appealing apps stay performant at meaningful spend for a very long period of time but then degrade very quickly when spend was doubled or tripled. So how do you capture that in a X creatives for $Y spend heuristic? Just like overall campaign performance, creative demand tends to change non-linearly with spend.

In general, though, I do think marketing teams underinvest on creative -- it's probably the biggest source of performance frustration I see when talking to marketing teams. 

by (8.2k points)
thank you! we were hoping to back out a year long recruitment plan based on solid numbers, but you're absolutely right that creative output doesn't entirely linearly correlate to spend.
I'll dig back into our data to look for average creative fatigue times and factor in at least 30% downtime for emergency creative needs (which I know we will fill!). Thanks again