Apple’s WWDC 2017 announcement that iTunes would provide analytics on podcast listener behavior brought a new hope to advertisers: hope for clear, actionable data such as actual listening (vs. downloading), unique listeners, and listen-through rates. In other words, a system similar to a Nielsen, for a media space that major brands want to be in but can’t justify without reach and frequency numbers. Moreover, some even hoped that there would be a behavioral and geographic component, showing exactly who is listening and where, to help advertisers better understand exactly what they were getting when they purchased a placement.
However, almost a year since the announcement, the developments have been less than revolutionary. On the advertiser and agency side, there is little visibility into networks’ analytics, not because the networks and podcasters are holding the data hostage (as some will share if requested) but rather because Apple has not considered the advertiser’s interest significant enough to build out an agency or advertiser portal. On the podcaster and network side, the results are underwhelming. While many articles hype podcasting as the ultimate advertising unit because of its listener engagement, we know there is more to the story and we’re hopeful Apple will deliver meaningful data to give us true insights.
The reality is that the analytics we do have prove what we already knew—that podcast listeners stick around for the majority, if not all, of a given podcast (including the midroll advertisements)—and none of what we didn’t: how many are listening, what else they’re listening to, and what else they are interested in.
Apple analytics data does not track from the source (the RSS feed). Rather, it tracks from the distribution point (iTunes). Therefore, the analytics tracked are only those that are relevant for iTunes (whether on a desktop or via the Podcast app in iOS). According to a recent callout from Libsyn’s “The Feed,” 62.5% of Libsyn (one of the major podcast hosting services) podcasts are streamed or downloaded through iTunes. That leaves 37.5% of the podcast audience unaccounted for and untracked.
Additionally, the 62.5% of users listening through the Podcast app need to be on iOS 11—the most recent upgrade to the operating system underlying all iPhones—in order to be tracked. As of 4/22/18, 76% of users had adopted iOS 11. That’s another 24% of the 62.5% being tracked that are simply omitted.
Only 47.5% of overall podcast listeners are actually being tracked by Apple’s Podcast Analytics.
However, an interesting development may be in Apple requiring RSS feeds to be in HTTPS format, where the last ‘S’ refers to a secure layer (SSL) encrypting the data. While this is most likely a move to continue end-to-end security around Apple’s entire infrastructure, there may be the added benefit of helping standardize the source of podcasts in an effort to, eventually, become an overarching third party that can make statistical claims on more than just 47.5% of the audiences. SSL technology in no way enables this function, but it could be laying the groundwork for an innovation in the medium-term future that could perform this role.
With only half of the audiences accounted for, agencies are not using this data in order to inform purchasing decisions, as its nuances are still fairly uniform across networks’ individual podcasts. An advertiser-facing portal would allow for more meaningful studies that could link a show’s audience response with anything from the minuscule differences in listen-through, the audience size variance from episode to episode, or the advertisement skip rates.
As one of the top three podcast agencies in the world, Oxford Road is in a unique position to conduct our own research. We have learned that drop-off in midroll units (regardless of position within an episode) stays steady at around 5%, give or take a few percentage points. Furthermore, completion rates for most major shows surveyed have maintained at approximately 80-90%, which is much higher than expected. With that, performance advertisers currently invested in the space can breathe a sigh of relief that their $20 CPM’s aren’t effectively $40 CPM’s due to only half the audience staying tuned in, as some predicted.
Against the backdrop of data-driven digital advertising and traditional media types with legacy audience ratings, podcasts are still a shadowy world for potential advertisers. Current advertisers know, however, that regardless of what limited analytics exist, smart podcast buyers breed highly efficient conversion. As the space continues to scale and advertisers recognize that podcasts could be a main acquisition driver in their portfolio, the demand for more robust analytics will only be more pronounced. With rival platforms like Spotify (6.18% listener share, according to Libsyn) encroaching on distribution, it’s Apple’s chance to maintain advertisers’ attention by actually delivering on analytics. If not, savvy advertisers and media buyers will find a more robust source.
The saving grace of podcasting was not delivered by the Gods of Cupertino. Instead, we, as performance advertisers, are still left wanting and relying on the same metric that got us to where we are: using cross-client performance data to determine the value of a podcast advertisement. Not audience metrics, not CPMs, and not clicks or views; instead, the only thing that matters to us and our clients—conversions per spot. Apple has the power to change that for now, but the matter at hand is still a question of if, not when.