Connected cars help measure the effectiveness of radio advertising for consumers on the move

Listening in car is still of paramount importance to radio, even in the pandemic a major part of listening takes place while driving. A new, innovative pilot study from the US used connected cars to understand the impact of radio ads for consumers on the move. It proved the efficacy of radio ads and showed the way for in-car attribution. This study heralds a new way through which radio broadcasters can partner with advertisers and automakers to prove radio advertising works.

Building a study that captured consumer habits on the road

Taco Bell is an American quick service restaurant (QSR) chain for which radio is an important advertising channel when it comes to enticing people in their cars to take a break and grab a bite to eat at one of its restaurants. They partnered with auto maker General Motors (GM), to conduct a test study that measured the impact of radio ads using information from connected cars, mixed with other data inputs to gain a detailed picture of how drivers respond to their audio messages. The study took place over two months between July 14 and September 14, 2019 in Columbus, Ohio where Taco Bell has a large number of restaurants and GM had a significant number of vehicles that operate in the area.

Over the two-month span of the study, 185,000 GM connected cars drove through Columbus, the sample size was derived from only those cars that spent a majority of time in the market to eliminate those drivers just passing through. 

After defining the general area, GM manually created a series of polygons of 78 different Taco Bell locations to create a geo-fence and measure a visit to the store when a car entered the polygon.

Blending advertiser and connected-car data

GM’s connected cars were able to measure location data via GPS and in-car radio listening data while Taco Bell provided GM with their radio ad logs for the analysis. On top of that GM has the broad demographics data of each car owner which added an extra layer of data.

General Motors was thus able to identify what cars pulled up to Taco Bell branches, and what the drivers were listening to when they made the decision to stop. At which point, it wasn’t difficult to go back into the data and pair off the QSR’s local ad logs with the moment of decision. A control group was established from drivers that never heard a Taco Bell commercial. For the study, a seven-day attribution window was used, hence, if a person heard a Taco Bell ad and visited the QSR within one week, it counted.

To give the Taco Bell marketers a very real index of what worked, GM also broke down its findings in to day parts – weekday vs weekend and morning, midday, afternoon, night as well as overnight drive.

To evaluate which kinds of messages were most persuasive to consumers on the road, GM split delivery environments into three genres - 30s spot, on-air personality sponsorship and Taco Bell promotions for news/weather/traffic reports.

Results

Among the findings from the pilot study were that the typical 30 second radio ad was the most effective in driving restaurant visits. Voiced (personality sponsorships) and NWT (News/Weather/Traffic) ad units generated synergies when combined. The combination of the 30 second standard ads with voiced personality enhanced campaign frequency. In terms of day part, mid-day recorded the highest driving activity.

The authors of the study stress that this was only a pilot with a small sample from only one market and one advertiser. They, however, see great promise for future studies that can take place on a bigger scale.

Data privacy

All drivers enrolled in GM’s connected services provide consent and, similar to smartphones and other connected devices, can control such information as location settings. In terms of data protection, all data remains within GM and is not shared with any other company. GM conducts the analysis which is an aggregate total of a broad group, not individuals. In addition, GM shares with its partners the findings of the study, but not the data.