May 26, 2015
In today's ultra-competitive multimedia battle for ad dollars, every consumer counts - especially when it comes to measuring the size of an audience. Harker Research Senior Partner Richard Harker was never totally comfortable with the accuracy of PPM ratings, but he couldn't put his finger on what exactly was missing. Now with data compiled by Voltair technology, Harker believes he found out legitimate problems with how PPM measures the listening audience - problems so significant that use of the Voltair box can help radio stations dramatically improve their AQH ratings ... and improved ratings mean higher ad rates. He has been discussing the issue at length on his Radio Insights blog; here Harker offers more details into his perspective on the problems of the PPM and the ramifications of Voltaire technology.
When did you first notice that there may be a potential problem with the PPM?
We've been raising the issue from the early days; it started with the first technical papers we read back in 2004. This was back when we were still putting out printed newsletters - our blog didn't start until late 2007 - and it has been a reoccurring theme since then. The more we saw the numbers roll, especially during PPM's roll-out period when Arbitron did side-by side diary vs. PPM rankings, the more we became very concerned that something wasn't right.
We've never given up on the issue, but the problem was that we never had concrete evidence. There was a lot of circumstantial evidence and strong inference from the research done in other countries, but Arbitron kept assuring us that everything worked. It was until a few years later that we confronted them with evidence that maybe the process wasn't as flawless as advertised. We had nothing concrete until Voltair proved our theories correct. That's why we've hit the subject much harder recently. We finally had concrete evidence that some problems are baked into the PPM.
Did that nonchalance towards your reservations change when Nielsen acquired Arbitron?
Of course, the acquisition of Arbitron is still relatively new, so no conclusions can be drawn. But we were hoping Nielsen would bring more sunshine into the issue. The things is, the people in charge of it now were doing the same thing under Arbitron. We're basically dealing with the same people, whose attitude continues to primarily be one of non-response. We would ask questions, for example, about the U.K. study that found that anywhere from 20-30% of listening was missed by the meter. They dismissed the study outright. We then asked them to provide any research that refutes what the British broadcasters found - and we got no response.
Then there are things like constant listening spans. According to PPM data, everyone listens to radio for around 10 minutes -- regardless of format. Listeners to News/Talk, Rhythmic AC, Country, Urban Top 40 ... everyone listens for 10 minutes. That doesn't ring true. We confronted them with that - another non-response. That defies logic. How could that possibly be true -- unless the meter is truncating listening to 10-minute increments? Now we know that's quite possible, which we've learned through Voltair. That's a much more reasonable explanation.
When exactly did Voltair come into the mix?
The Telos Alliance, a radio and TV tech company, started looking at PPM at the behest of a radio client. Telos was very good at building processors and understood processing. The broadcaster came to them with questions about meter behavior because they couldn't understand how the PPM could accurately pick up listening. Until then, no one had the resources and the curiosity to press on.
Dr. Barry Blesser, who would eventually invent Voltair, wrote a paper once he grasped the significance of what he found -- that the PPM encoding/decoding process had some serious flaws. You can still read that paper online. When we read one of his papers, we started to think about a technical explanation of things that have been attributed to listener behavior.
Is Telos the only company to create an instrument that found lost listening?
Telos has a patent on the process. You can read all about it online to see what they're doing. It's pretty comprehensive. As I read it as a layman, they have a patent on the process of enhancing audio specifically for watermarking - and it's not about PPM exclusively. Voltair, though, was designed for the PPM though enhanced watermarking. They have the general patent, which makes it hard for another company to try the most obvious solutions. Whether there are other alternatives out there that do essentially the same thing, I don't think so - unless Voltair now inspires others to take a different look at it.
My ultimate hope is that Nielsen licenses the technology from Telos to use industrywide. There's no reason why the PPM encoder shouldn't have this process built into it. It's the ultimate solution for everyone's problem. It solves Nielsen's embarrassment over PPM's flaws, and it creates a level playing field for radio, since everyone has the technology.
Another thing a lot of people overlook is that this would help radio itself. What we're seeing so far is that in markets where a lot of Voltairs have been installed, while cume may not grow, AQH ratings go back up. Remember, when PPM was first implemented, cume went up but AQH went down, which cost radio a lot of money. If we can get to point where Nielsen agrees to integrate the technology, radio as a whole will benefit.
Who else should push for making this technology an industry standard - radio groups? Advertisers?
Every stakeholder should take interest in this. Where is the MRC? If this device extracts more listening - and that's essentially what PPM is, a delivery system -- MRC should take an interest and insist that Voltair's technology be put into the encoder.
Advertisers are a trickier situation. Advertisers never want to pay more; they want to pay less. If radio's AQH levels go up, by rights radio should be able to charge more money for spots. Advertisers won't want this. But everyone should want accuracy; that's fundamental. You need to rely on a ratings service. Also keep in mind, Nielsen could very well use this opportunity to improve TV ratings; Voltair offers a whole different level of accountability into that mix, too.
Now we know why the PPM has been the least favorable to spoken word formats, because encoders fail to encode every time someone stops talking. If there are a number of pauses -- which happens when the people aren't talking over each other -- the station can get no credit at all. With Voltair technology, those pauses aren't a problem, so spoken word formats stand to benefit the most. Remember, Talk radio took a huge hit when PPM rolled out; nationally it lost two to three shares. Now we know that could very well be due to the encoding process.
Think about all the newscasts and sportscasts, both on radio and TV. You'd think Nielsen has to take interest in this; I don't know why they haven't responded as well. For now, there are a lot of anxious radio broadcasters who'd like to know what's really going on. Think of the implications!
All that being said, Voltair technology compensates but doesn't entirely solve the problem. PPM as implemented will never fully credit all listening for all formats. Voltair just gets us a little closer.
Who's responsible for balancing everyone's interests here?
Good question, especially when you look at the gains some stations have made by installing the Voltair box. We're about to post the most detailed and accurate determination of what Voltair has done for two stations. For one station, Occasions went up 26% almost overnight. For the second station, Occasions went up over 60%! Recall that from day one Arbitron said that the key to winning with PPM was increasing Occasions. Now we know how!
Predictably, stations that currently have Voltair don't want a level playing field right away. Why? Because they get more listeners with the box, which more than pays for itself in a large market. If I'm a broadcaster who has it, I'll drag my feet to make as much money as I can before the other broadcasters in the market get it.
Look across the country. All Access covers the PPM ratings. Look at the big jumps some stations are getting; how many of them have record ratings? The question naturally raised is okay, what happened so that these stations suddenly have record ratings? Certain AC stations in major markets are doing better in the first quarter of 2015 than they did playing all-holiday music over the Christmas break. That's unheard of ... huge success stories that defy explanation ... unless those stations have something nobody else has. It's obvious that many of those success stories are a direct result of those stations using Voltair.
The only reason more stations don't have them is because Telos can't make them fast enough. They're turning them out 24-7, but they're a small company that can't make them fast enough. Broadcasters are frustrated that they can't get them sooner. They've seen enough. At this point, the floodgates are about to open.
So in a sense, the playing field levels as soon as everyone has a chance to buy one.
The marketplace will do that. When you're watching your competitor move ahead of you and nothing else is going on -- other than thinking that they have box and you don't - you're going to buy a Voltair.
Beyond that, this is an issue for not just station management, but for everyone in radio who reads All Access. Because if Voltair is right about the problems with PPM in encoding, then basically everything we understood about how to win in a PPM market is wrong ... everything! Everything was predicated on the fact that since the PPM was accurate, we needed to act on what we learned from it. Such as shutting up jocks. Radio has been depersonalized in terms of personalities because we saw the numbers go down and thought listeners didn't like jocks. Now it turns out that the jocks weren't getting credit for hours because the PPM dropped out during momentary pauses and soft-volume moments. End result: A lot of high-profile jocks are out of work because we thought they talked too much and listeners didn't like it when they did. Now we'll have to completely rethink it.
The same goes for certain formats such as Smooth Jazz. The PPM pretty much destroyed Smooth Jazz; 46 stations have abandoned the format. If you look at last few Arbitron books before the PPM took over, Smooth Jazz stations may have not been top 5 or 7, but they had good numbers and were doing just fine in salable demos. That's why a lot of most medium-sized markets had a Smooth Jazz. Now in the PPM era, you can't find them at all because that music doesn't encode well.
You can go right on down the line with this. The whole "minute by minute" programming decisions PPM supposedly enables you to make? It turns out big chunks of listening are disappearing from minute to minute because it's not being encoded well. So how can you do accurate "minute by minute" programming with that?
The bottom line is that we have to get an accurate measurement of what's really going on with our listeners. We need to make sure PPM is 98-99% accurate, not 99% accurate one minute and 40% or less accurate the next. Once Nielsen finally acknowledges this issue and signs off on \n encoding process that fixes the problem, we may be looking at a very different radio environment. Everything we look at may be very different; we may see some different formats and more spoken word variations. You won't have to be a Top 40, Hot AC or Country to succeed.