Study to trial randomised control testing for ad measurement

RCT aims to reduce sources of bias in analysis by comparing outcomes for a pre-determined exposed group to a non-exposed group.

The technique is used within the ‘walled gardens’ of tech platforms, but the study will explore how it can be applied in cross-platform measurement across multiple media channels including digital and TV.

The ARF’s proof-of-concept study aims to show how RCT methods can be used to measure the incremental ROI of ad campaigns and complement existing approaches to analysis, including marketing mix modelling and multi-touch attribution.

The ARF is working with measurement firms 605 and Central Control and media researcher Bill Harvey Consulting to conduct the research. It is also recruiting advertisers to participate in the study.

Scott McDonald, chief executive of the ARF, said: “The digital platforms have championed the value of random control tests and have made them a cornerstone of their success. ARF has highlighted the scientific progress that has been made on ‘incrementality,’ but we needed to take it farther to have more extensive impact on existing marketing practice.”

Rick Bruner, chief executive of Central Control, said: “Observational analyses in forms such as retrospective naturalistic quasi-experiments MMM, MTA, single-source and national time series are incredibly powerful in generating hypotheses, but advertisers cannot prove the ROIs without RCT verification. The industry needs all of these methods combined, and cross-platform RCT has been mostly missing until now.”

The study will begin in the third quarter of the year and is projected to be completed by the first quarter of 2021.