Has the time for an accountability approach arrived?
After the recent news of Comscore’s Chief Executive Bryan Wiener and President Sarah Hofstetter stepping down, we started thinking about the future of media measurement and reporting. As a category, the currency for trading is suffering; with investment dollars increasingly shifting out of TV and into digital, companies like Comscore & Nielsen are heavily impacted.
What’s affecting the industry isn’t only a loss of relevancy, but a loss of credibility in the measurement system. The way people deliver, buy and consume media has changed – but how performance is measured hasn’t. The challenge goes back to the inability to track a consumer within and across media channels. If we could, that would give us a consistent, comparable metric i.e. we could build true multimedia reach curves. Despite multiple efforts, however, the industry has failed to achieve this.
With traditional currency disintegrating before our eyes, how should we continue to measure performance? Has the time for an accountability approach in our measurement arrived? If so, what constitutes this better currency and how can we make the shift?
Marketing Mix Modeling is often compared to providing an exchange rate given its ability to equivalize across channels on an ROI scale. This is exactly what is needed for setting budgets across media, however utilization of MMM is not feasible for trading in those media since it does not provide a media-centric metric to compare the underlying reach and frequency of delivery.
Is there an underlying metric that is equivalent in both interpretation and scale across all media channels?
At Marketscience Truesight, we have been working on a solution using the best available data and developing a proxy approach to equivalizing reach & frequency for each media channel. While this approach doesn’t account for how consumers consume media across channels, it does put all channels on an equally scaled metric and allows us to solve for potential synergies across channels within the modeling process. Having equivalently scaled media inputs to the modeling process is critical to avoid bias in the estimation of channel performance, especially in multiplicative models.