Research: Attention measurement focuses on creative and media
October 3, 2023
Most attention measurement companies (61 per cent) focus on both creative and media as attention drivers and to measure these factors, they leverage eye tracking (50 per cent), survey-based ad recall (46 per cent), tuning duration/dwell time (42 per cent), and facial coding (42 per cent). Notably, when combining all various survey measures, surveys become the number one method for measuring attention.
These findings come from Phase One of the Advertising Research Foundation’s (ARF) Attention Validation Measurement Initiative, which is an empirically based evaluation of the rapidly developing market for attention measurement and prediction. This phase included a comprehensive Literature Review, and an in-depth survey among attention measurement providers, resulting in a compendium that encompasses the attention definitions, methodologies, KPIs and profiles of participating companies.
The survey also reveals a noteworthy diversity in how companies define attention, with most emphasising concepts that align closely with their specific tools and methods. Furthermore, though companies vary in their emphasis on attention, viewability, engagement, and emotion, these aspects are commonly portrayed as interconnected in a non-linear fashion, often comprising a sequential process marked by feedback loops and interdependencies.
“In the ever-evolving realm of media consumption, understanding and quantifying audience attention has become more significant for content creators, advertisers, and media executives,” said Paul Donato, Chief Research Officer at the ARF. “By summarising the unique definitions, perspectives, methodologies, and technological innovations of these measurement companies, this report sheds light on the various approaches being offered in the market. We also hope that the Literature Review will furnish advertisers with a foundational reference for understanding attention’s role in advertising.”
Additional findings from Phase One of the Attention Validation Initiative include:
- Most of the companies in the attention landscape that measure both media and creative are part of a larger private corporate structure and use a mix of old-school and new methodologies. However, there are also relatively new, small, privately-owned, attention-specific companies called “attention natives,” that tend to specialize in newer measurement tools like eye tracking and facial coding.
- Almost all companies (92 per cent) collect data from human subjects at some point in the process, however the extent and type of human data collection varies widely – from online surveys and panels to qualitative and biometric testing. In some cases, proxy signals and AI models are trained to replace direct human measurement or to predict attention without actually exposing the advertising to humans.
- Most companies try to control for other variables as they measure attention. When companies test for creative, they generally control for media context. Similarly, when testing for attention to media, companies often control for creative and device. Exercise of other controls is more similar between creative and media testing, but also less often for both. Isolating these variables is a challenge for attention measurement companies.
- Across all companies, CTV, video, social media and programmatic predominate the media platforms measured. Notably, AR/VR/Metaverse is measured by only a small minority of companies even though this is the most complex and costly creative to create.
The ARF’s Academic Advisory Committee and a Brand Steering Committee, which include senior researchers and practitioners from Coca-Cola, LinkedIn/Microsoft, MARS, McDonald’s and The Attention Council, among others, devised the RFI questionnaire used to collect information from the attention measurement ecosphere. The second phase, Validation of Attention Measurement in Creative Testing, is currently underway. The ARF is assembling a set of common creative assets and inviting creative testing vendors to analyse them as they would for any client. The ARF will then analyse the degree of convergence and divergence in the results and evaluate those results in the light of the information collected in Phase One.