Solutions
Teams
Built for your whole team.
Industries
Trusted by all verticals.
Mediums
Measure any type of ad spend
Platform
Use Cases
Many Possibilities. One Platform.
AI and Automation
The Always-on Incrementality Platform
Teams
Built for your whole team.
Industries
Trusted by all verticals.
Mediums
Measure any type of ad spend
Use Cases
Many Possibilities. One Platform.
AI and Automation
The Always-on Incrementality Platform
Now that the user level tracking is becoming a faded memory, advertisers are in the look out for alternative measurement methods. Incrementality is gaining popularity, but the methods Advertisers were used to required geo-testing.
Geo-testing requires Advertisers to make active changes in a region in order to compare the performance of that region vs. another. Conducting geo-testing experiments forces advertisers to stop channels or activities in one country or region, and if one something happened the day after the geo-test was done – advertisers would need to, again, stop campaigns in a certain geo to run the test.
Geo-testing was the “gold standard” during the 1960s, during a period when Advertisers had only sporadic campaigns, running across only TV, Radio, and Print, where the granularity of campaigns, channels and ad groups did not exist.
But in today’s day and age – Advertisers are “always on” – and require their measurement to be always on too.
Real time user-level attribution provided some sort of solution, but given the recent privacy regulations imposed by Apple and Google, user-level attribution is no longer possible at scale. Using user-level data, programmatic platforms were able to offer some sort of incrementality measurement using audience groups, but this capability, biased as it was, is also becoming a thing of the past.
Geo-testing for incrementality is a compromise for Advertisers, and often, will provide partial insights if any at all. Geo-testing does not provide insights in a granular level, but only at the level of a channel – at best.
Geo-testing for incrementality is a hassle. It is complex, has a large opportunity cost involved, and more importantly – provides inconclusive results as an outcome.
Advancements in technology and algorithms provided better methods for incrementality measurement – fit for today’s needs.
Always-on incrementality measurement allows Advertisers to measure the incremental value of their ad channels, specific campaigns, and even specific ad groups without the need to run experiments. By detecting micro changes in marketing activities (each time a channel was opened / paused / closed , every campaign budget increase or decrease, and every bid change) and utilizing those as tiny retrospective experiments, gives data science teams the ability to run millions of time series data points in order to learn the causality behind sales performance and conclude if it is incremental or not.
Always-on incrementality does not operate in the domain of impressions and clicks, but uses costs and revenues as the inputs – therefore, the changes in the attribution space have no relevance or impact over always-on incrementality measurement.
In the last years, causal inference as a method of marketing measurement has been proven to be effective in academic studies by University of California and MIT. Platforms like Facebook, Microsoft, and Google created open-source projects to support this areas. Large advertisers attempt utilizing causal inference successfully, but only at the level of channel or medium. To create a more granular use of causal inference in marketing, an algorithm needs to be trained by an enormous number of various scenarios. Similar to how a Tesla network allows cars to “teach” each other new skills.
Always-on incrementality measurement is becoming more available to marketers. Up to recently, this method was only offered as a service by niche technological consultancies and agencies.
As access to 3rd party user-level data is disappearing – many cling to tracking as means of measurement by utilizing some form of fingerprinting in some form or another. But in the long run – only those who adapt to a more technological method of true measurement will be the ones who prosper.
Incrementality has been a hassle so far, due to the need to run experiments. It’s time to drop these pointless experiments and move on to AI and Machine Learning based causal inference methods that offer advertisers an always-on measurement approach that works across any medium, every channel, if it is digital or not
INCRMNTAL is an always-on incrementality measurement platform. Our software allows marketers to test out incrementality with a push of a button, and get results within seconds, with no need to run any experiments that require you to stop your campaigns. If you want to learn more, visit INCRMNTAL or book a demo today!