This August 2023 article opens with an anecdote about an unnamed mobile gaming company spending nearly $1M a day in 2015 and pausing its campaigns in February. The story prompts a useful question about incremental contribution, but without the underlying design, comparison and data it is not a reproducible experiment or proof that half of every advertiser’s spend is wasted.
Table of Contents
The objective? To understand the actual value obtained by all marketing activities, before restarting channel after channel, in order to see what the actual incremental contribution each channel brings, and based on this, adjust the level of ad spend, or completely cut off channels which bring no value.
So, what did this “strategic hiatus” yield?
After trying this bold experiment, the results were eye-opening. The company realized that half of their advertising money was being wasted. With this knowledge, they shifted their funds to more effective campaigns and partnerships.
With each passing year, the CMO reiterated this experiment, consistently uncovering similar findings. They recurrently trimmed its ineffective ad spendings by half. This analytical approach resonates with John Wannamaker’s iconic statement,
“Half the money I spend on advertising is wasted; the problem is, I don’t know which half.” 
Image source
John Wanamaker (1838-1922) was an American merchant and advertising pioneer
Fast forward to the age of user tracking:
By 2015, brands and ad platforms had a lot of access to detailed user data. But things started to change in 2016 when Apple introduced Limited Ad Tracking on iOS. Google’s similar feature didn’t have much impact until 2022.
During this time, everyone thought they had the perfect tools to measure digital advertising success.
Mobile tracking platforms became very popular, with some even claiming to be better than big names like Facebook and Google. But a big surprise was coming.
Unexpected attributed conversions or suspected fraud deserve investigation, but a reported anecdote is not a benchmark for another app. Review the underlying evidence, billing, event records and independent comparison before concluding that a channel is fraudulent or that a fixed percentage of spend has no value.
Understanding Incrementality in Marketing:
An incrementality study estimates what happened because of a marketing intervention compared with a suitable counterfactual. Pausing and restarting channels is not, on its own, a valid design: seasonality, carryover and other changes can affect the result. Define the intervention, eligible comparison, outcome and observation period before changing spend.
The goal in such a study is to create an experiment by creating a significant change that should produce a significant outcome, in comparison with a hypothesis. It was a painful process, making some performance marketing managers shake in their seats, as there was a true empirical test, showing the actual value of marketing activities.
What does our expert say about incrementality?
“In an era where traditional attribution is fading, incrementality seems to stand out as the best option, and maybe the only in the future. It provides an alternative lens into ad impact, ensuring we’re not just tracking, but trying to understand how each cent is doing.”
Pablo Perez | Performance Marketing Director @ Admiral Media
Attribution: A Double-Edged Sword?
Imagine you believe half of your sales come from paid ads and the other half from natural, unpaid sources. You’d probably think that if you stop all ads, your sales would drop a lot, maybe by 80%, in a few weeks. But what if you stop the ads and your sales stay the same? This highlights the difference between tracking individual users’ actions and the older method of looking at the overall increase in sales.
In the chase for simplicity, user-based attribution may have oversimplified a complex puzzle. By allocating credit solely to the final user click before a conversion, it undermines the multifaceted customer journey involving multiple touchpoints.
With the dwindling ability to track users, primarily due to new restrictions from Apple and Google, advertisers all over entrenched in user-level data faced anxiety. However, those versed in broader measurement methods flourished.
Future-Proofing Measurement:
Traditional ways of tracking were too simple. Now, we need a better way to measure success.
Using methods like incrementality or new tools helps brands get a clearer picture, instead of just relying on one possibly incorrect piece of information.
If evaluating a measurement tool such as INCRMNTAL, ask how it constructs the comparison, which assumptions it requires and how uncertainty is reported. A vendor choice does not remove the need to reconcile data or separate attributed credit from incremental impact.
