To Voltair or not to Voltair, that is the question!

July 30, 2015

In 1981, Charlie Sislen was working at WWDC-FM, and his program director gave him the best insight ever. She said, “I don’t care how many people are listening to my radio station, all I care about is how many diarykeepers think they are listening to my station.” Program Directors have always strived to game the system to increase ratings without breaking the rules.

A lot has changed in 34 years. PPM has eliminated the recall game but, as with all survey techniques, it is not without its flaws. Every survey technique has drawbacks and flaws. It is with this in mind that we weigh in on the Voltair discussion.

Much has been written over the past couple of weeks questioning the impact of Voltair on the ratings. Before some of this perception becomes reality, we thought it would be helpful to look at the facts, figure out what we know for certain and what we still don’t know.

Nielsen held a client-only webinar on July 21st and the press reports focused on a few important findings:

• The Voltair unit did increase the intensity of the embedded watermark. It also introduced “audible artifacts,” meaning people could potentially hear the watermark.
• The biggest impact of Voltair in terms of actual meters capturing actual codes, presumably in a lab, were scenarios where the level of background noise equaled or exceeded the level of the broadcast content (and therefore the embedded code).
• Nielsen expressed a concern that in these situations, the meter might pick up codes where it was not possible for the listener to hear the content.
• Nielsen pointed out that the PPM captures the codes the overwhelming majority of the time, and that extra code captures wouldn’t necessarily translate into increased ratings. This is because as soon as enough codes are captured to give a station credit for one quarter-hour, extra code captures in that quarter-hour don’t matter.
• Nielsen wouldn’t support the use of the Voltair unit, but would be making changes later this year to their encoding technology to increase the “density” of the embedded watermark.

However, based on those reports Nielsen did not address whether Voltair was having an effect on ratings.

We did.

We have looked at the pre and post-Voltair meter exposure for several stations. We have seen more, sometimes considerably more, exposure by meter keepers after Voltair was installed. This has resulted in significant ratings increases. Is this actual listening that the meter was missing, or is the meter now picking up watermarks for audio that could not be heard by the human ear? We at Research Director, Inc. do not know the answer to this important question.

Our sample size is very small, but in our opinion, there is a disconnect between the findings reported by Nielsen and the actual ratings impact that people in the industry claim to see after Voltair is installed. Nielsen said that the main impact of Voltair is in high background noise situations. If this is the main benefit then a lot of radio listening is happening in noisy situations. That seems unlikely since we know that a large portion of radio listening happens in the car (and we don’t think road noise qualifies here). The other possible answer is that Voltair does have some other incremental benefit in order to give the results that stations claim to see.

One caveat: It is extremely challenging to measure the impact of Voltair on a station’s ratings. Ratings can change dramatically from week to week without Voltair, so attempts to look at ratings before and after Voltair is installed and isolate its impact may not be conclusive. Still, where there’s smoke…

Randy Kabrich did a masterful job evaluating a huge amount of data to come to the conclusion that Voltair is much ado about nothing. One of his premises is that the total number of AQH Rating Points across all 48 PPM markets is basically flat from June 2014 to June 2015. This is an interesting attempt to use the data to prove a point, but there are two problems with using AQH Rating points for this purpose:

1. It takes a significant increase in AQH Persons to raise your AQH Rating. The main culprit is the fact that AQH Ratings Points are rounded to one decimal place. If you looked at an AQH Persons Ranker you could see that 5 stations are tied with a 0.2 AQH rating, but the AQH Persons across those 5 stations can vary from 20% to 90% because of rounding.
2. As was stated earlier, AQH ratings change for many reasons. By using all stations rather than just stations that installed Voltair, the effect of Voltair (if any) is being watered down by the stations that did not install it.

Let’s be clear, in the end Randy might be right. But there is enough empirical and anecdotal data to say that the investigation is ongoing. There is too much at stake for our industry. An independent, third party review of a lot more data is needed. More transparency about the analysis that was performed is needed. More cooperation is needed from stations that have installed Voltair. In order to foster that cooperation, Nielsen needs to do what it takes to allow stations to feel comfortable about admitting they are using Voltair.

In short, let’s not let the conversation end here. Let’s continue working towards the best audience measurement system that is technically and financially feasible for our industry.

Then we’ll be able to get back to what we do best, which is creating great content to draw an engaged audience, then creating ways to connect that audience to get results for our advertisers.

-Marc Greenspan, Partner

• Tweet