Out-of-home media has a measurement problem. A new data standard for MMM models may help brands to better understand its impact, writes Alex Tait.

For anyone working at a media owner, it can be easy to forget just how much you are across as an advertiser. Your brief, or at least part of it, may be overall media effectiveness. But between agency, budget and stakeholder management, the reality is that you often don’t have time to get into the weeds on how a specific media can work effectively.

That’s why for me, as a former client-side marketer who spent a significant amount on out-of-home (OOH) media, the project we’ve been working on over the past few months with a wide brief to maximise its impact for advertisers has been particularly satisfying.

This isn’t an industry PR piece for OOH, as you’ll see when you read the study. Our guiding principle has been to focus on key interventions that advertisers and the industry can make to ensure media work even more effectively. With some of the recommendations we are advocating, in effect, for a change programme for the industry.

Our ground-breaking project has come up with two findings important for any brand investing in media:

  1. When OOH is measured correctly, it will likely perform better in MMM models;
  2. By measuring it correctly, you will be able to optimise plans to further increase its effectiveness.

We’ve also put together a checklist of simple steps advertisers can take to maximise the impact of OOH.

OOH is usually not measured correctly in MMM or econometric models

As the project progressed, we focused increasingly on the area we believe can affect the most positive change: MMM measurement. We’ve worked independently with all participants in the OOH industry (including 15 MMM agencies, key media owners, OOH trade body Outsmart, and advertising trade body ISBA) and have ultimately produced what we believe to be the most comprehensive study into OOH measurement to date.

A fairly staggering statistic from our study is that eight out of ten MMM companies responded that they didn't get the data they need to model OOH optimally.

We wanted to make all our findings as practically oriented as possible. To address that particular point, we’ve compiled an MMM data standard for the industry to use. You can download it for free in our wider recommendations on the Entropy site.

You can also watch one of the advertisers who contributed to the project – Money Supermarket’s Head of Marketing Strategy and Insight, Jonathan Wood – discuss the MMM data standard below.

If your brand needs help assessing whether the data going into an MMM model is optimal and meets the standard, we are partnering with Outsmart to provide a data health check service, which you can also request at the above link. The service will assess the data inputs for OOH against the gold standard, with a RAG scorecard against each element, and liaise with the various parties involved in each brand’s modelling to provide the feedback.

However, our longer-term recommendation for the industry is to build a portal to make the data as easily available for modelling as it is for TV with BARB or digital via ad server impression data.

Two other key recommendations we are making for advertisers are:

  1. Plan with measurement in mind

This is a media 101, but is often not implemented.

What we are recommending is sufficient OOH scale, so the channel can be picked up in the model; intelligent phasing to avoid collinearity with other media; and a thought-through test and control set-up so that we can compare exposed areas with non-exposed areas.

  1. Systematically test OOH creative prior to use

Intuitively, one of the factors – indeed, perhaps the most important factor – in an OOH campaign is the quality of the creative. As with digital display, average dwell times for OOH are somewhere between one and two seconds, so you need to make an impact.

However, despite creative testing being very cost-effective – depending on the tool used it costs from several hundred to several thousand pounds – interviewees in our study demonstrated there is still much more scope for using them.

Download our full recommendations here. If you’d like to be walked through the recommendations, we'll be happy to oblige. We are also looking for advertisers to work with on the next stage of the project, which we could discuss at the same time.