Setting a standard for measuring podcast audiences

Note: Part 2 of "Mapping The Podcast World" will appear here later. This discussion of audience measurement felt pressing to publish first.

Nicholas Quah of the Hot Pod newsletter recently posed the question of whether there are not enough advertisers to support podcasts, or not enough podcasts that are attractive enough as advertising venues. One of the reasons for this problem, Quah says, citing several sources including the Wall Street Journal and technology business news site The Information, is the lack of an accepted standard measure of podcast audiences.

Quah adds that such a measure would stabilize podcasting as an industry and gain the medium more respect with advertisers, but stops short of a solution, saying this makes now a good time “to play around and figure out what works.”

That seems like a vague prescription, however, for podcasters who need a reliable standard to earn the trust of advertisers. A possible avenue could be something modeled on CPM (cost per thousand views), the longtime standard used for broadcast and print, which also mutated to mean cost per thousand impressions from online advertising on text or image pages (excluding audio or video streams).

Such a benchmark still leaves a lot of definition questions, though. What constitutes effectively serving an ad? A complete listen to a whole podcast (especially if that ad might be halfway through an episode)? A mere download of the podcast episode? Or the equivalent of actual “click-throughs,” which for podcasts, would mean someone using an offer code to try the advertiser’s service, or otherwise indicating when surveyed by the advertiser that they chose that product or service based on a particular podcast ad?

For its podcasts, a group of public radio broadcasters have issued an attempt at a standard, but this largely consists of technical specifications to prevent multiple counts of the same download (which can happen if requests are being counted, since sometimes multiple requests are needed to complete a download), or ways to make sure downloads are indeed correctly counted if they happen to be coming through the same IP address.

Either way, this presumes counting, correctly, the number of downloads, is the correct measurement. That may be so, and this effort is an attempt to ensure consistent standards within the public radio podcast world, to make its podcasts competitive with commercial podcast advertising networks like Midroll and Panoply. That doesn’t mean, however, that this standard is consistent with the way those commercial networks do their measurement to justify their pricing.

Still, download numbers may be the easiest measure, but to convince more advertisers of the maturity of podcasting as a medium, the industry not only has to get consistency on how to measure those numbers, but also arrive at accepted measures for engagement with the advertising, or response rates.