My good friend and Northwestern alumnus Pierre Bouvard, whom many of you know from his many years a top executive at Arbitron, recently joined Cumulus Media and its Westwood One division as their Chief Marketing Officer.
In that capacity, Pierre published a blog piece a couple of days ago called “Did You See What Netflix & Pandora Did Last Night? How Streaming Stacks Up to Radio & TV” that purports to show that Pandora and Spotify are nowhere near as strong in their fields as Netflix is in its field.
You can read the piece in its entirety HERE.
The problem is, Bouvard’s analysis doesn’t compare apples to apples.
On the video side, Pierre looked at cumes, in tight demos, against individual brands, which made Netflix look good.
On the audio side, he compared shares, among total audience 12+, against the entire medium, which made Pandora and Spotify look bad.
To be more specific:
On the TV side, using data from Ipsos MediaCT’s TV Dailies service, Pierre shows that In the P18-34 demo, Netflix has a 42% past-week cume, which is higher than ABC (30%), Fox (27%), FX (22%, thanks I assume largely to the wonderful TV show “Archer”), NBC (22%, despite the absence of the wonderful TV shows “Community” (given up to Yahoo) and “Unbreakable Kimmy Schmidt” (given up to Netflix)), and Comedy Central (19%).
On the radio side, using data from Edison Research’s “Share of Ear” study, Pierre shows data that in the P13+ demo, Pandora has a 7% share of all audio usage (which is a bigger pie than “radio” usage; if you subtract out slices of the pie chart that aren’t radio, you can see in that Pandora has about a 10% share of radio usage), which is lower than the shares of all AM/FM stations combined.
Based on this comparison, he concludes “Despite its notoriety, Pandora’s actual share of listening is quite small.”
But actually, on the radio side, if you looked at the P18-34 demo and compared Pandora’s cume against the cumes of specific brands of radio (note you’d have to do this in a given market, since radio is comprised of local rather than national brands), I imagine you’d probably see pretty close to the same pattern of performance as you see with Netflix on the TV side.
Conversely, if you looked at Netflix’s share of TV viewing time among P13+ viewers, you might see a share that’s somewhat larger than Pandora’s share of radio listening, but not a lot larger.
Elsewhere in the piece, he notes that broadcast TV viewing is declining — among P18-49s, in Q1, ratings down 10.5% this year over last — but he doesn’t choose to look at whether the same phenomenon is happening regarding broadcast radio listening.
Finally, he observes, “Netflix has a handful of competitors (Amazon Instant Video and Hulu), but Pandora has a whole bunch of competitors, and more emerge by the day.” That may be true, but I’ll bet you that Netflix’s share of video streaming viewing (particularly if you also throw YouTube into the mix) is smaller than Pandora’s share of Internet radio listening.
In both the TV and radio worlds, the truth is that broadcast consumption is declining and streaming consumption is increasing — and at approximately the same rates. It’s broadcast industry wishful thinking to imagine otherwise.
And I would continue to argue that the best move for broadcasters is to compete in this new space (just as Cumulus is doing with its equity position in Rdio!).
ADDENDUM: According to this recent article in Re/code — http://recode.net/2015/04/23/netflix-eats-into-tv-ratings-with-help-from-the-tv-industry/ — Netflix, with 10 billion hours of video streamed last quarter, represents “close to 6%” of total TV viewing in the U.S. Assuming that Edison’s “Share of Ear” study is largely correct (and it seems to be supported by Triton’s Webcast Metrics numbers compared to Nielsen Audio’s radio PUR ratings), Pandora is garnering about 10% of all “radio” listening in the U.S. So in fact the data suggests that Pandora is having a bigger effect on radio than Netflix is having on TV. Not smaller, but bigger.