THERE IS PLENTY OF STATION STREAMING AUDIO BEING CONSUMED, BUT YOU OFTEN CAN’T SEE IT

Our chart showing public radio station’s

streaming audio in the April 2019 PPM ratings

Our report on May 22^nd about the apparent dearth of listening to station streaming audio [link], as reported in Nielsen Audio PPM ratings, raised a lot of eyebrows, particularly when we rhetorically asked “Is this all there is, and if it is, is streaming audio a failure.”

In the report on May 22^nd we focused only on Nielsen Audio topline data. This was a mistake on our part. We excluded data from other sources that was relevant. We regret our error and will do better in the future.

Our readers told us that there are more listeners to station audio streams than what is reported in Nielsen Audio top-lines. But, you need to know where to dig and realize that much of the data is proprietary or even hidden.

Often, what a programmer knows about their streaming audio audience is determined by how much they are willing to pay and what use they are making of the information.

FACTORS AFFECTING STREAMING AUDIO AUDIENCE METRICS

• TOTAL LINE REPORTING (TLR)

Nielsen offers subscribing stations the choice of having their listener data broken out by each audio source or having listening to all sources rolled into one data line called Total Line Reporting (TLR). Stations most often want TLR for simulcasts of AM/FM combos, repeater stations in the same market, HD channels that are feeding programming to FM translators and streaming audio.

If a subscribing station requests TLR, and Nielsen determines that the simulcast meets their criteria, listening data from all sources are combined into a single line in Nielsen Audio reports.

An upside for a station to choose TLR is that a station’s estimated number of listeners appears in the ratings as one “big” number. This can be useful in competitive situations.

A downside of TLR is that listening to streaming audio (and other simulcast sources) are embedded into the station’s over-the-air estimates. This means the information is hidden for even the subscribing stations unless they request a rare special “for your eyes only” report from Nielsen to see the breakouts.

According to a February report in Inside Radio, only 6% of Nielsen Audio’s subscribing stations currently request TLR. Nielsen lists the stations that use TLR here.

 In the same story, Inside Radio reported [link] that Nielsen recently changed its criteria for stations to qualify for TLR. Now 95% of the station’s quarter hours need to be simulcast over the course of the survey period – one month for PPM markets and one quarter for Diary markets. There are additional criteria that also must be met. Nielsen Audio makes final decision on TLR requests.

• STREAMING AUDIO MEASUREMENT FROM OTHER SOURCES

There are a number of companies that provide detailed analytics of audio streaming for stations that can afford to pay for it. Perhaps the best known is Triton Digital [link], a global technology and services company that enables broadcasters, podcasters, and online music services to know details about their streaming audio audience.

Sample of Triton Digital’s

Webcast Metics Monthly Rankers

Triton Digital’s Webcast Metics Monthly Rankers [link] is considered an industry standard. The topline data is available for no charge and it is frequently quoted by industry news sources. But, beyond the Monthly Rankers, the paywall goes up. Only clients who have the cash can get the granular, station-level, details.

There are many streaming audio providers. One of the most popular with public radio stations is Stream Guys [link]. They offer a variety of services, software and gadgets including analytics and monitoring of streaming audio listening.

According to the Stream Guys website, public radio clients include WNYC, WBEZ, KUT, KQED and KCRW. Stream Guys is a for-profit company and their audio streaming data is not publicly available.

Plus, some stations that are university licensees use campus resources to obtain the data.  Often it is available at no cost.

INPUT FROM SPARK NEWS READERS

The decision whether to choose TLR for streaming audio data involves a number of factors. Our readers discussed several of them in comments made on the Spark News blog page and on the PRPD Facebook page:

Consultant John Sutton, GM of NPR News/Talk WESA in Pittsburgh, and former Director of Research at NPR, told us:

John Sutton

SUTTON: “The absence of reported streaming data in Nielsen does not mean a lack of streaming audience. The majority of public radio stations have their estimated streaming audience rolled-up into their Total Line Reporting. Their radio and streaming numbers are not reported separately by choice.”

“[Also] When stations do report streaming audio listening separately there is likely duplicated Cume. [Another variable is] that Nielsen [only] reports in-market listening. It does not report out-of-market listening, which can be part of a station’s streaming strategy.”

“We estimate the WESA stream increases our in-market listening by 5% to 6%. Is that worth it? The answer is yes, even if there is no new Cume.”

“Our Triton data shows that about 25% of our online listening occurs outside of our Designated Market Area.”

Steve Allan

Steve Allan, Programming Research Consultant at Research Director inc. [link], verified Sutton’s information and added: 

ALLAN: “For stations that are not simulcasting [over-the-air and streaming audio]. Nielsen requires a separate encoder which means an additional cost to the station.

Certainly, stations know their exact streaming numbers from their server side data but Nielsen does not have access to that. We have seen an increase in streaming numbers in some markets.

However, and to your point Ken, logic dictates it should be higher. After all, stations are running countless promos instructing their listeners on how to set up their smart speakers for listening.

Stephen Steigman

Stephen Steigman, Chief of Broadcast Operations at KCUR in Kansas City added:

Steigman: “I know you're focusing on Nielsen PPM readings, but what about other sources of (better) streaming data like Triton analytics?”

“We rolled up our broadcast and stream Nielsen numbers a while back so we don't get individual stream numbers from Nielsen, but our Triton numbers are to-the-listener accurate.”

“Our streaming numbers (according to Triton) are tracking at about 12-13% of our combined weekly cume (from Nielsen), and we know that close to 40% (yes, you read that right) of our stream is consumed by smart speakers.”

“That - in my opinion - is the future of a station's stream, and as such, there's no reason to believe that streaming is a "failure."

Aaron Read

Aaron Read, Director of IT & Engineering at The Public’s Radio in Providence, Rhode Island, pushed back on the notion that data from vendors such as Triton is more useful than data from Nielsen:

Read: The thing about Nielsen that you can't get from most any other method of audience data in streaming is that Nielsen knows an enourmous amount of data *about the panelists*.

Sure, you can get much more precise data about how many people are listening at any given second. But you can't get a lot of useful info about who those people are. In many cases, you can't even tell if they're really people! (instead of bots)

This isn't to excuse Nielsen per se, just something to keep in mind that server stats from Triton or wherever aren't necessarily the end-all/be-all of listener data.

For what it's worth, I'm pretty sure that our streaming will never show up in Nielsen's reporting. [That is] because our server capacity with StreamGuys is only about 700 simultaneous streams.

To which Steigman replied, via a comment on our site:

Steigman: Although Nielsen has data about the panelists, if I had to pick or choose, in this case I'd rather have accurate numbers than statistical extrapolation based on a small number of panelists. Nielsen certainly can't tell me how they're listening: computer, mobile device, smart speaker. That information is helpful.

Bill Lueth

Bill Lueth, VP of USC Radio summed it up for many readers:

Lueth: “KDFC and KUSC combined have more than 300,000 unique monthly listeners but you can’t see that in Nielsen. I assume there are other similar situations.”

PROMOTIONAL MESSAGE