5 Steps To Put OOH Into Your Media Mix Models

By Jeff Casper

Modelers prior to 2010 had a very good reason not to include OOH into their media mix models: There were no audience ratings. Modelers could correct for but could not directly account for out of home’s impact on an advertiser’s goals. The result, predictably, were that models could not recommend out of home.

But now OOH can be and should be a part of media effectiveness models. In 2010 the out of home industry, through the Traffic Audit Bureau, released a full complement of ratings and impressions for each and every TAB advertising location. TAB Ratings are now widely used across the country to buy and sell media, but they still are not widely used in assessing ROI.

The hurdles are there but surmountable. While the product is available, the process to integrate OOH into models is often not defined. The parameters around how to measure OOH need to be set, as does getting the data in order.

Here are five steps that can be and should be done today with your modelers so that OOH can be separately recognized.

1. Agree on the Role OOH is Meant to Play

Agree ahead of time what the communication goals are for your OOH plan. Is it to increase awareness? Extend the message? Impact store visits? Whatever the goal, a comparable metric to judge its effectiveness should be used that matches the goal.

2. Agree on the smallest market possible

OOH is both a national and a hyper local medium. Many models attempt to compare media across a DMA. This often is not ideal to get a proper read on OOH impact because it is so location sensitive., unlike other media. It is imperative that modelers understand that the smallest geography possible must be used to correctly measure impact. A trade area or a group of counties is much more telling than an entire DMA. This may mean performing multiple models.

3. Agree and set up a proper delivery input platform

One of the major hurdles is providing the input data for modelers to use. Buyers must take on the added responsibility of providing detailed spreadsheets of what was bought, how much, and when in an agreed upon format. Creating this regular process and taking ownership of it is key.

4. Agree on how to adjust OOH GRPs

Typically, OOH is purchased in 4 week flights at an even number of GRPs per week. While this is the official buy, it is not reflective of how audience is actually delivered. Applying a frequency distribution to the results is one way to effect a more realistic audience delivery. A frequency distribution is required to calculate reach and frequency. It is an estimate of how many people see your campaign once, twice, three times, etc. over the course of the campaign. The frequency distribution can also be applied

to the ROI delivery to estimate true audience delivery over the course of the campaign. This is just one way to adjust GRPs, and there will be other opinions. The important thing to understand is that an adjustment must be made.

5. And finally, write it down! Gather your thoughts and record best practices.

Out of home requires special attention to model. Many of these steps can be automated. The most important thing to understand is that including OOH as a separate input means modelers will be able to explain more of the variance. This means the entire model is stronger as a result. Including OOH not only helps understand OOH impact, it leads to better quality work.