MMMs, Advanced Radio, & Pixels - Lessons from the CAO Attribution Forum

The Media Roundtable is back! We recently welcomed dozens of Chief Audio Officers for an Attribution Forum at Audacy's SoundSpace in Soho, NYC. ICYMI, we’re sharing highlights and wisdom straight from the CAOs.

Giles Martin, (EVP of Strategy & Insights, Oxford Road) hosts James Ingrassia (EVP, Client Service, Oxford Road), Emmy Jadoff, (Senior Manager of Integrated Media, HelloFresh), and Shane Pittson, (Founder, So&So) for a deep dive into attribution.

The team is talking: MMMs, Radio Measurement, Pixels, and more. Let’s jump in.

“Whether it's an MMM or even a pixel-based model, it's a view of the world that’s created through the math, but it's not the truth. It's not reality. It's giving you an estimate.”

Get in the Mix 🥣 - MMMs can be wildly helpful (one CAO’s anecdote showed podcast’s impact was 5x bigger than they thought). The problem? It’s not the end all, be all. One recommendation: if you’re using an MMM, advocate to be involved in the overall process. The MMM is only as good as its inputs – your expertise is a key part.

Next Level Radio 📻 - One tactic that got a lotta love? Geo holdouts. But as with any attribution, details make the difference. Instead of picking random pairs, look for closely correlated markets (and control for seasonal/regional impact). There are 3rd party partners available for this – and possibly your Data Science team if you’re lucky. Whatever the attribution challenges may be, we go further when we go together.

Pixel 🔎 - Pixels: they don’t solve everything, but they can sure point you in the right direction. Caveats abound (better for website traffic than foot traffic, YouTube is an issue, etc). Since we still don’t have perfect solutions, one approach might be: which two solutions give you a complete enough picture for you to make better decisions (Pixel and survey is a dynamic duo snagging a recommendation).

As British Statistician George E. P. Box said; “All models are wrong, but some are useful.” To get the most out of your approach to attribution, tune into the full episode linked below.

Bianca Gorodinsky