I’ve always had difficulties falling asleep at night, and lately, with a lot more reason. As a kid, devoid of options available later in life, my tranquilizer was my transistor radio, tucked safely between two head pillows, tuned either to a late-night ballgame or all-news radio. I gave them 22 minutes, they gave me a sedative.
These days, the combination of my phone and an impossibly comfortable headband with built-in soft headphones have replaced my trusty Panasonic with the 9 volt battery, and an endless cacophony of podcasts have replaced all-news (and, for that matter, all-sports) radio. And since despite the fact that I consider myself technologically above average for someone of my generation, my fat fingers reduce me to somewhat of a klutzy Luddite. So when I was prompted to subscribe, I would often hit “download all episodes” for something I’d discover.
Which has often led me to waking up in a time warp. Did you happen to miss THE DAILY’s five-part series on the challenges facing those running in the statewide elections of–2018?!?! No worries, I listened this past weekend. Those debates about how in the heck the Montreal Canadiens pulled off that upset over the Toronto Maple Leafs in front of crowds of placards during the 2021 all-Canadian (and still isolated) Stanley Cup Playoffs? In my head as my prostate woke me up last night.
Well, apparently, I’ve had the same first-hand experience as many, many others, and one that Apple has now seen fit to nip in the bud. As SEMAPHOR’s Max Tani reported at roughly the same time I was catching up on how that newcomer Ron DeSantis was emerging in Florida:
Apple has quietly tightened its reporting of how many people listen to podcasts, sending shock waves through an embattled audio industry still reeling from the end of the COVID-era production bubble.
The shift, Apple wrote in a blog post, was technical: The dominant podcasting platform had begun switching off automatic downloads for users who haven’t listened to five episodes of a show in the last two weeks.
But while few users noticed the shift, some of the biggest podcasts in the world saw their official listener numbers drop dramatically. Long-running shows that publish frequently were hit particularly hard. A user who listened to a show like The New York Times’ The Daily a few times, subscribed, but stopped listening would continue to count as a download indefinitely. Even better under the old rules: For people who listened to a show, dropped off for a while, but started listening again later, Apple would automatically download every show in between. The arrangement drove big download numbers, a crucial metric for ad sales and a sign of the vast reach of podcasts as a medium.
Guilty as charged. I apparently had downloaded hundreds of old episodes of THE DAILY when I subscribed, and when my current queue would run out it would immediately autoplay the next newest, even if it were months or years earlier.
But as Simon Owens observed in his TECH AND MEDIA NEWSLETTER, the addressing of my quirky overnight behaviors has caused a lot of folks responsible for the creation and monetization of audio content to have more sleepless nights of their own:
With the latest iOS update, those automated downloads no longer get triggered. If your podcast only publishes occasionally, the update probably wasn’t a big deal, but those publishing more frequently saw a seismic hit to their download numbers. In fact, the change was so consequential that the downloads for the top podcasts are down by as much as 20%.
Unsurprisingly, those shows were heavily dependent on the lost downloads, mostly so they could exaggerate their listener base to advertisers.
As a media researcher who has dealt with the reality checks both of skeptical advertisers and demanding executives, I’ve lived through these kinds of existential debates for decades. I once worked for someone who literally screamed bloody murder when I was forced to try and educate him that a website click is not necessarily the same as a view, nor did it necessarily count in the manner that a television viewer was tallied. I wish at that time I had access to Owens’ summative statement of exactly how misinformed he was:
Almost every metric we use, both privately and publicly, to measure the success of our content is grossly exaggerated, and there’s very little we can do to change that.
To start with, many experts believe that upwards of half of all web traffic is generated by bots; and while some of this bot traffic can be detected by your web analytics platform, a good bit of it isn’t, hence why ad fraud is such a massive problem. If inauthentic bot activity were easy to spot, then adtech platforms would simply weed it out, and publishers wouldn’t have to lose an estimated $84 billion every year to ad fraud.
And these days, this sort of measurement denialism has been amplified by a streaming industry so petrified to actually tell the world how inconsequentially viewed a a majority of the content they have created and curated is they continue to go to great lengths to hide their actual numbers except in cases where they are either demanded by advertisers or are favorable enough to create a compelling narrative. Peacock has been eager to share their Nielsen numbers on this past weekend’s Chiefs-Dolphins football game because they defied the expectations of most pundits and still continue to frustrate those who believe it was some sort of ethical violation to force the majority of those 23 million viewers to pay one red cent for the privelage of seeing it. They were quite eager to pass along the datecdote (copyright: Entertainment Strategy Guy) that TED also set a record for viewership of a Peacock original–yet the specific numbers, like almost every other one of their competitors’ remain a trade secret.
And as Owens notes, if you applied the same sort of logic and filtering to that which the podcast industry adjustment has seen, the actual number of people who indeed saw enough of TED to pass judgement was even less than their internal records might be showing:
Even in cases where a human being actually visited your content, they probably didn’t consume very much of it. I’m always reminded of this when I open my YouTube dashboard and see the average watch time is five minutes. Given that my videos usually span about 45 minutes, that means an enormous portion of my viewers are opening a video, watching the minimum 30 seconds needed in order to be counted as a “view,” and then closing the video not long after. Web articles also have a depressingly low completion rate.
Everyone, from my old boss to the currently flailing leaders of an industry chasing the holy grail of Netflix with increasing futility, loves bigger numbers. We now report linear viewership in millions or billions of minutes, not only to “level the playing field”, but to make it sound like it’s still a number that can be statistically significant. Outdated measurements of linear live-same-day adult 18-49 ratings rounded to one decimal place are annoyingly redundant. When everything, as master strategist Preston Beckman often notes “regresses to the mean”, a sea of 0.1s and 0.2s to compare each other to is irrelevant and difficult to spin.
There are encourging signs out there that given these reality checks on quantitative size some companies are developing different strategies to prove the relative worth of their content. THE VERGE’s Amrita Kahlid recently reported on efforts from Spotify’s newly recruited ad sales topper Dave Byrne to help quell some concerns that have hampered the growth of the podcast industry’s monetization: Kahlid asked Byrne a pointed question that elicited a telling response:
I know Spotify announced a partnership with Integral Ad Science (IAS), or the major software that brands use to make sure their ads are appearing on “brand safe” or appropriate content. Can you walk me through Spotify’s offerings for brand safety and what went into their development?
We’ve been making sure to do heavy investment in contextual targeting. A lot of old-school brand safety capabilities and technology are very much keyword-based. For example, let’s say I were to say something like, “I had too much to drink last night.” You know the context behind that is related to alcohol based off the way that I’m saying it. [An alcohol brand] may even anti-target [the keyword drink].
The problem with keywords is that you may miss out on other great things, like health podcasts of people drinking water, pre- and post-workout. You could also be missing out on great content from food critics and that kind of thing. So we decided to invest in the contextual aspect of things to understand the full breadth and scope of what’s being said and how it’s being said and then create brand safety solutions that kind of go around that.
IAS was an early partner of FOX’s networks when we were developing a narrative that people actually paid more attention to ads when they were surrounded by compelling content, which we, of course, thought was pervasive across our entire portfolio. That wasn’t always the case, but it did help a network like FX, which some people believed was a toxic cesspool of overly adult content, win many debates about exactly how well-received and better-recalled and how little negative association was applied to brands who did choose to advertise in those shows. The same sort of skittishness is now rampant in a world where the likes of Joe Rogan, Ben Shapiro and Alex Jones are household names. Not every podcast is headed by those sort of personalities, and not every listener is going to do something drastic as a result of listening to them.
Trust me, if that were even remotely the case, I’d probably have led an insurrection or two by now, given all of the episodes I’ve downloaded.
But, then again, I’ll readily admit I’ve slept through a lot of those playbacks, and probably wouldn’t pass IAS’ litmus test. But at least Spotify is trying something. And they need to. Despite having a valuation of $38B, they’ve been cutting jobs and reporting losses. And they’re arguably best in class.
A lot more jobs can be saved–and perhaps some could be resurrected–if the companies hell bent on goosing numbers as high as possible without any context would invest in the methodologies and expertise that could better help tell the story that the investment community, let alone advertisers, would respond to.
We all have to start listening a lot more carefully. And selectively. Otherwise, more of us might be up at night, and we’ll have to download more different shows to put us to sleep.
Until next time…