Measuring ad impact with incrementality
Advertisers face a growing pressure to identify activities that drive growth. They are constantly looking for ways to justify the true impact of their campaigns and their spending. While the attribution model has been the most common and widely used measurement tool, it doesn’t determine the true impact of a single marketing activity. Nowadays, more and more advertisers are moving towards measuring incrementality to help them understand the effect of their media plans and eventually make better informed decisions.
What is incremental lift in paid marketing?
Incremental lift is defined as the likelihood of a consumer converting due to a particular interaction. The interaction could be as simple as watching or clicking on a specific ad. Whichever interaction directly influences a purchase decision is identified as incremental. It also means that a purchase wouldn’t have happened without that exact interaction. For example, Emma was exposed to a display banner that led her to buying a dress. If it weren’t for that ad, Emma would not have bought that dress. That specific display ad has caused an incremental lift.
How is incrementality measured ?
To measure incrementality is to determine the difference in outcomes between an audience that saw the ad and the ones that didn’t. By isolating an ad, we can estimate its true causal effect on sales. One of the reasons why some platforms like Google and Facebook are dominating online advertising is thanks to the measurement capabilities they offer, allowing marketeers to connect the dots between their ads and sales. One of their solutions is the randomised control trials (RCT), a type of experiment that can effectively gauge the value of a marketing tactic.
RCT randomly assigns users to a control and test group. The test group is exposed to the ad while the control group is not. That way any difference in outcome can be linked to the ad, hence determining the incremental lift. This is the fundamental principle of the experimental design that determines causation.
Why incrementality is better than attribution?
Attribution is a measurement method that assigns a certain value to one or several touchpoints, reflecting their impact on conversion. It’s widely used by marketeers due to its simplicity and ease of use. The most common types of attribution is the last click. It attributes the last interaction as the one that caused the conversion. As it’s surely a simple form of estimating causal effects, it has some imperfections. Let’s take the example of someone who just bought a watch, the last touch model implies that the poster they saw when entering the store has driven them to purchase that watch. In reality, the journey towards the purchase decision is a lot more complex than that. They might have watched the ad on TV, read some organic posts and been served a couple of display ads. This model wouldn’t give credit to any of these touchpoints except for the poster they saw at the store. Hence, inaccurate representation of what actually caused the purchase. For that reason, the attribution model might not be very accurate in determining a single’s ad influence.
Contrary to attribution, incrementality measurement can determine the true causal effect of an isolated marketing activity through controlled experiments. Even though it’s not as simple to implement as the attribution model, it does pay off in the long run. This rigorous experimentation approach allows advertisers to make informed decisions thanks to a scientific, evidence-driven process. Based on the outcome of these experiments, marketeers can improve their ROI and direct their spending in a more effective manner.
