"But would the user have purchased anyway?"
For years, this has been one of the most common questions CMOs have asked their marketing teams about their affiliate spend. Those teams, unable to answer the question with much confidence, have instead responded with a shrug: "We're not too sure."
The result: Lacking a good way to prove the incrementality of the publishers they work with, affiliate marketing managers are forced to optimise for margin instead of for volume.
This is a stark contrast from other dynamic channels, which offer tools that lets budget holders hypothesise, test, and prove incrementality. These tools give marketers confidence and proof of value, and those two things are what help them justify their continued focus on volume.
Here's how Tiziana Bacchilega, affiliates lead at Deliveroo puts it:
"Incrementality is key at Deliveroo, and more importantly the ability to rigorously test and prove incrementality to inform marketing spend. It is increasingly difficult to justify investment inactivity if we are not able to measure the return with confidence, and this results in the budget being allocated to channels where incrementality testing is becoming a routine."
In this piece, I'll explain the relationship between a brand's ability to test and their willingness to spend, and dig into how incrementality testing gives affiliates the ability to prove the value they bring to brands.
For marketers, there is usually a direct conflict between margin and volume. The less incremental the channel is perceived to be, the fewer marketers are willing to optimise towards volume and growth. The opposite is also true: The more you can prove your value, the less risky you are, and the more a brand will be willing to spend with you.
In performance marketing, the way you manage risk is by managing margin, and the way to manage margin is by reducing rates.
When spending on a channel that is perceived to be risky, marketers will typically bake in hefty margins — as much as 75% — to give themselves flexibility in the event of any macro changes. (These changes can include their target profit margins, their ratio of new-to-existing users, or their larger corporate strategy.)
For better-optimised channels, brands are willing to run activity at a much thinner margin. This is because they know the value of their ad spend and are able to quickly optimise said spend to compensate for macro factors.
What that means for publishers is that the cost-per-acquisition a brand is willing to spend with an affiliate is significantly lower than they're willing to spend in any other performance channel. To protect themselves, brands will pad their margins and run publishers at a much lower rate than they would deem acceptable in other channels.
In cashback programmes, this results in less money to the user, which leads to lower volumes, which in turn means less revenue for publishers.
If publishers want to increase the amount of money that brands spend with them, then they need to increase confidence in them at the senior marketing level. In my view, the only way they can do that is by regularly, proactively, and willingly articulating the value that they're bringing to the table.
How can they do this? Though incrementality testing. Here's how the process works:
All that sounds good in theory, but publishers invariably run into complications during the testing process. Here are a few best practices:
Publishers and brands should align on what "success" is from the outset. Incrementality isn't a binary thing. Instead, it should be measured with clear KPIs and alignment on what good looks like, ideally before the tests run.
Brands should commit to what happens when an incrementality test is a successful one. One of the biggest reasons why publishers are skeptical of testing is that proving incrementality often doesn't result in an increase in spending. Ideally, publishers should look to brands for a commitment on whether spend will increase when incrementality is proven.
Publishers should make sure that the brand is running the test across multiple publishers at the same time. Another reason why publishers are so skeptical about incrementality testing is that brands will often run holdout groups across 10% to 20% of their users (meaning that these users don't get any cashback) while maintaining existing programmes with competing sites. The result is that these users see they're not getting any rewards, then transact with another competing publisher. A true incrementality test requires that the brand runs the test across competing publishers at the same time. This is especially important for the UK, where we have two dominant cashback players.
Smaller publishers need to make sure they have enough resources for testing. Unless you have the right systems in place, the process I explained above will invariably require a lot of data science hours and a lot of CSV files being sent back and forth. This can be a real resource strain for both brands and publishers, particularly smaller ones, which are unlikely to have the scale, data science, and engineering resources necessary to run these kinds of tests.
The bottom line is this: Marketers today value fewer things more than the ability to optimise and test. For publishers, this means that giving brands the ability to test for incrementality and impact is the surest way to get brands to spend more with them. The publishers that embrace this will see brand investment grow to new heights, while the ones that resist it will continue to decline.
.png)
.png)
