Innovations

NPR’s move into podcasting analytics raises privacy concerns

December 18, 2018
 

Until recently, it hasn’t been easy to measure the success of podcasts, beyond how many people download them—a metric so broad as to be almost useless, much like page views are on the web. In the past year or so, Apple and Spotify have both come out with analytics on their platforms that tell producers how long listeners spent on a podcast, but their systems are proprietary, which means they only share the data they choose to share. So NPR, a podcasting giant in its own right, decided to develop its own open-source podcasting measurement standard, one that anyone can use.

In a blog post on December 11, developer Stacey Goers introduced the software, called  Remote Audio Data (RAD) technology, as a way of sharing listening metrics directly to publishers with what she called “extreme care and respect for user privacy.” The RAD standard inserts an audio “tag” into the podcast stream at various pointspodcast producers can choose where and when to add themand then keeps track of how many listeners made it past a specific tag. This would theoretically allow podcasters and marketers to know how many people listened to an entire podcast, how many bailed partway through, how many skipped the inevitable Blue Apron ad, and so on.

RELATED: NPR kills journalist’s piece over her accent

NPR said in a news release that the new standard is being incorporated by podcasting distribution services such as Panoply, Radio Public, and Podtrac, and that creators and publishers including The New York Times, iHeartRadio, Google, and ESPN have supported the development of the technology.

But not everyone in the podcasting business is happy about the new standard. While NPR makes the argument that the data will make it easier to convince advertisers and sponsors to commit to backing podcasts, and possibly allow creators to bring in more revenue, skeptics say it will also make it easier for creators and advertisers to track and identify individual listeners. Marco Arment, who developed both the original platform for Tumblr and the document-saving service called Instapaper, created a podcast distribution service called Overcast. He says he has no plans to integrate the NPR standard into his product, and he’s not in favor of others doing it either.

“I understand why huge podcast companies want more listener data, but there are zero advantages for listeners or app-makers,” Arment said on Twitter after the NPR announcement. “I won’t be supporting any listener-behavior tracking specs in Overcast. Podcasters get enough data from your IP address when you download episodes.” Arment went on to say that he sees no reason for any podcast app to support that kind of tracking and reporting: “There’s no clear benefit to the app makers (in fact, it requires more work) and it’s a privacy violation and a GDPR liability that users would likely object to if they understood it.”

Podcasting purists are also afraid that the possibility of granular tracking of users could turn the business into the same kind of wasteland of terrible direct-response ads that much of the web has become—inevitably leading to less revenue, not more. “Let’s look at the web, the best and most recent example of a mass-market, highly diverse, mature, fully-analytics-capable medium,” Arment said. “How did that turn out for all sides involved? Is web publishing a healthy business while minimizing consumer/privacy abuse?”

RELATED: Battle heats up for morning daily podcasts as NPR enters the fray

Goers and NPR colleague Joel Sucherman, a vice-president of digital strategy for the broadcaster, say that the standard anonymizes the data that is tracked and retained about users. However, those who understand the technology say it would be easy for someone to take the information they get from the download of the podcastwhich includes an IP address and other related infoand figure out who someone is (and where), and combine that with the tracking of their listening behavior. While NPR might not want it to be used that way, the likelihood is that someone somewhere will choose to do so.

Sucherman says the benefits of knowing in a more granular way how much of a podcast people listen to is worth the risk. “It’s fine to sit in judgement and say [podcast] publishers are evil somehow because they want to understand more about people’s listening habits, but that’s not living in the real world,” he says. “In order to produce better shows and better content, I think it’s a fair tradeoff.” Goers says one of the groups NPR really wanted to help was smaller independent podcast creators, who might not have the brand to bring in big names. “Think about smaller podcasts being able to attract sponsors and show results,” she says.

NPR One tried the same kind of tagging with its own podcasts and it had immediate results, they add. The NPR Politics podcast used to open with every panelist introducing themselves, and analytics showed a lot of people stopped listening before they even got to the main topic. So the data team took that to the editors and they pivoted the show so that it got straight to the point much earlier. “After we changed that, it became an immediate hit,” Sucherman says. “We could see the exact moment when people started ditching out, so we took it back to the newsroom and they were fascinated by it. It allowed us to make decisions that are impossible to make with traditional analytics.”

Arment and others noted that Pocket Casts, a distributor that NPR bought a stake in earlier this year, has not said yet whether it plans to incorporate the tracking standard (Update: In a post on Twitter, Pocket Casts said it has “no plans to implement RAD at this time”). But Sucherman doesn’t think that lack of support is a message about whether the company agrees with the technology. “They’re an independent company, so we don’t tell them what to do, we just own shares in them,” he says. “Also, they just had a big new release, so I wouldn’t read too much into [them not supporting RAD] just yet.”

As rational as the criticisms of Arment and others might be, if history is any guide, advertisers and marketers are likely to push for more analytics and data on podcast users in order to justify the money they are putting into the market, and if NPR’s standard can provide that, then it will likely find some support.

ICYMI: Judge makes unexpected decision in Cardinal sex abuse case, affecting journalists across the world

Mathew Ingram is CJR’s chief digital writer. Previously, he was a senior writer with Fortune magazine. He has written about the intersection between media and technology since the earliest days of the commercial internet. His writing has been published in the Washington Post and the Financial Times as well as by Reuters and Bloomberg.