Here’s a familiar scene: I’m trying to pass the time, so I pull up the Apple Podcast charts to see what the youths are up to. (Ha.) This was my Sunday afternoon, and by that point, I hadn’t looked at the charts in a good few weeks. Part of this has to do with the way I learn about new projects these days: press releases, emails, text messages, phone calls, even a postcard once. But it mostly has to do with the fact that I haven’t found the Apple Podcast charts particularly useful in quite some time. Not for my purposes, anyway.
On Sunday afternoon, this is what I saw:
There’s a scene in The Matrix where that one creepy white dude looks at a cascading wall of code and says, “There’s way too much information to decode. You get used to it, though. Your brain does the translating. I don’t even see the code. All I see is blonde, brunette, redhead.”
With these charts, all I see is: “Doubt, doubt, trust (?), doubt, doubt, Serial.” To my eyes, which hasn’t looked at the Apple Podcast platform in a good while, the charts felt distinctly… broken. But then I thought back to earlier times when I skimmed the charts more often, perhaps every few hours, to see what was trending. What if the charts were always like this, and I just needed some time to readjust?
So I did what I always do: I tweeted.
Turns out, maybe the charts were acting crazier. My DMs lit up, and the complaints poured in. Most messages were sent in private, couched in an apparent unwillingness to publicly express displeasure about Apple Podcasts, perhaps because getting promoted on the platform’s editorial page remains a manual process. Though there were certainly other podcast folk independently raising the question in the open themselves.
Several people expressed suspicions of a dramatic increase in chart manipulation campaigns. They were talking about those scams in which unscrupulous characters use or pay (via marketplaces like Fiverr) for automated scripts that rapidly trigger Apple Podcast platform interactions in order to artificially vault their shows up the charts. (See this and this by Discover Pods’ Kevin Goldberg, or this, for more details.) You can probably guess what’s a suspicious show, and what isn’t, by cross-referencing related data points like “number of episodes in the top episodes list,” as Chartable’s Dave Zohrob did here. Given the increased attention and stakes currently enjoyed by the podcast industry these days, there have never been greater incentives to engage in these efforts. Or at the very least, as Court Junkie’s Iman Jalali pointed out, there are actual human beings who wish to say things like “I have the #1 business podcast and #2 podcast overall on iTunes” at random conferences. “It’s Google Search Engine manipulation of 10+ years ago happening all over again within iTunes,” Jalali said.
This is combined with a vibrant cluster of frustration I heard around large drops in charts placement that seemed to be inconsistent with download behavior. Granted, I was also reminded that these drops, or mass chart reshuffles, periodically happen — it’s almost as if the charts “resets” itself every once in a while. But it feels like it’s happening more lately, some said.
I also heard associated complaints of disappearing ratings and reviews. The most prominent of these, perhaps, came from Radiotopia’s 99% Invisible. Host Roman Mars tells me that somewhere between 500 to 700 of the show’s ratings appears to have disappeared over the past few weeks. As a reminder, Apple has signaled that ratings and reviews are important drivers of chart placement. That’s why you hear those appeals for reviews at the end of so many podcast episodes; it’s deployed in the belief that such ratings can help the show get discovered on Apple Podcasts. If the durability of those ratings are compromised, or if the relationship between reviews and the ranker is in question, or if the chart rankings more generally are ultimately meaningless, then the logic of integrating those appeals don’t actually end up holding much water.
(Meanwhile, things look pretty normal in Canada. It always looks normal in Canada.)
At this point, you could argue: so what? These issues aren’t exactly new. And you would be correct. I’ve seen “chart resets” many times back when I still checked the charts every few hours. Disappearing reviews aren’t unheard of, though they seem fairly uncommon. And finally, chart manipulation scams were around back when I started listening to podcasts in 2008. Scams, after all, are a natural byproduct of platforms and human systems. If something can be gamed, it will be gamed.
By Tuesday morning, the charts seemed to have worked itself back out. But the weekend’s pronounced chaos at the top of the charts have left many podcast folks wondering: have the frequencies of these three issues — the chart resets, the ratings disappearances, the scams, but particularly the scams — greatly increased? And furthermore: are all three issues increasingly interrelated?
Sure seems like it. I’ve reached out to Apple for comment or insight on the matter, and I’m still waiting to hear back. You’ll know when I do.
Let’s do counter-arguments: Who cares? Do people still look at the Apple Podcast charts? Do the charts still matter? And don’t we all know that the charts are weird and kind of broken anyway?
I’ll confess: for the longest time, I’ve never thought these issues to be significant problems per se, because I’ve generally held a low estimation of the Apple Podcast charts’ reliability and took it as a given that the podcast industry would recognize, understand, and work around that. One producer voiced an opinion that summed up my position pretty well: “Sincerely, I’ve operated thinking [the charts] were all garbage and that that was the industry consensus. Are substantial decisions actually made based on them?” To put it another way: Forget it, Jake. It’s Chinatown.
However, there’s an inequity with those counterarguments, and with my original position. An Apple Podcast Charts that doesn’t work the way it should might have limited consequences for bigger shows and publishers — which already have the audience base and budgets to diversify their marketing efforts — but it disproportionately impacts smaller or newer teams looking to authentically access the top layer of the charts, where they can possibly be spotted by the demographic most likely to try them out. Even if the actual number of people checking out the charts is generally low, it’s still a meaningful number to that class of show.
Are substantial decisions actually made based on the charts? Probably not, as far as major podcast publishers, advertisers, and various business folk are concerned. After all, they are already naturalized, informed citizens of the industry. But the ones who do make decisions based on them are the ones the industry needs: those dipping their toes into podcasting waters for the first time. Listeners, yes, but also: new advertisers, investors, talent, and so on. Perhaps even unfamiliar reporters, the ones who are more inclined to give attention when they hear the word “Top 10 on iTunes,” as Joseph Fink points out, and the ones who use the word “chart-topping” when writing about a podcast when that concept, in fact, means very little.
As my dad would say (shouts to him), a bad first impression is money left on the table. What does it mean when the top of the Apple Podcast Charts, which is still believed to be among the first touch-points for many newcomers, feature more scams than authentic entries? What signal of values does the chart project to those experiencing their first glimpse of the wider podcast universe?
There are also broader technical implications of an unreliable Apple Podcast chart system. As Flash Forward’s Rose Eveleth points out, there’s a possibility that the effectiveness of the platform’s “You might also like” discovery feature might be compromised by errant chart behavior, and that a good deal of third-party podcast apps rely on Apple Podcast listings to populate their catalogue and power their own discovery features. (We saw a version of this dependency flare up during the Alex Jones de-listing episode.) I say “possibility” because, well, we simply don’t know how the Apple Podcast platform works. All previous scholarship on the matter, including my own, is purely speculative. It’s a black box, it changes when it wants to, and it almost never signals why.
And then there was the weird bug that happened last week. On Thursday morning, US-based podcast publishers woke up to find that entire episode listings went missing for certain shows, including heavyweights like The Daily, My Favorite Murder, and Serial. The common denominator appeared to be podcasts that published new episodes that day, though it should be noted that not all such shows were affected. Apple fixed the bug later that day, but the experience left many publishers feeling incredibly unsettled and powerless.
Which brings us to another major question: what does it mean when creators invest a good deal of resources — time, talent, budgets, blood, sweat, tears, hopes, dreams — into this crucial marketing, discovery, and validation channel that is ultimately opaque and unpredictable?
Let me check myself here. Would that it were so simple for me to baldly state that none of this should stand, that it’s ridiculous for the charts to be this bonkers, that Apple should do better. The reality is that managing something as complex as the Apple Podcast platform is very hard. And the further reality is that podcasting remains a minor concern for Apple. Their historical status as a benevolent ward of podcasting’s growth and open publishing values can be interpreted as… well, somewhat unexpected. Given these two realities, I don’t feel particularly legitimate in pressing on the question of whether Apple Podcasts has a responsibility to provide its podcast constituency with systems that are more reliable. With perspective, I’m just happy to be here.
But something has to be said about what’s on the horizon. It’s fall of 2018, and there are enough rumblings to suggest that Apple may not be podcasting’s de facto discovery hub for very long…