Podcast metrics get a bad rap. There is still a lot of confusion over the metrics and what they mean, and talk about the "Wild West," especially among marketers and advertisers. But in reality, it's not that bad.
Over the past decade, the media and advertising industries have come together to create a clear definition of a download for on-demand audio content. With this metric, championed by the IAB and industry leaders, everyone can speak the same language and transact business in a meaningful way, which has added to the 85 percent growth in podcast advertising revenue between 2016 and 2017 according to the 2017 IAB Podcast Advertising Revenue Study, with even more significant growth anticipated in 2018.
The analogy I use is the magazine industry -- a well-established model that few question. We all know what circulation means, and we all know who the audience for any publication is based on surveys and qualitative data. That's similar to podcasts today: Downloads are the equivalent of the magazine making it to the mailbox. We can guarantee the file made it to a user who requested it.
Since it's hard to accidentally download a podcast, there's great confidence (backed by qualitative data) that people have chosen to subscribe and are listening. Qualitative data can also tell you who is listening. Publishers have great survey data on the audiences listening to their podcasts. At NPR, we've found podcast fans are happy to take surveys and help us out. Beyond publisher-specific data, firms like Edison Research regularly quantify the audience at scale, and brands often conduct their own brand lift studies to measure impact of their campaigns.
So that's the good news. But we're still talking about delivery of a file and not actual listening. Given the digital age we live in, we as an industry can do better.
The technology for delivering and playing podcasts is different from the standard Internet models for video or text content. There are no cookies in podcasting; it's just the transfer of a file. This made a lot of sense when podcasts were developed more than a decade ago, as data was expensive and podcast files were big. Downloading them over Wi-Fi and listening later, without using expensive cellular data, was the norm.
But that's all changed. Most basic data plans today provide ample overhead for even the heaviest of podcast listeners, and Wi-Fi is more readily available than ever.
Listening behaviors are shifting quickly, too. Edison Research, in their Podcast Consumer 2017 research, found that 77 percent of people now simply click on a podcast and immediately listen. With 42 million people listening to podcasts each week, that's a lot of activity.
The time has come to move the metric from downloads to listens. We don't have to settle for knowing the magazine made it to the mailbox. We can know how many pages someone read and even which pages they skipped along the way.
There are no fundamental barriers to real listening analytics on podcast content. It just requires a new model and participation from the industry. But there are things that have to happen to get there.
Technically, none of this is challenging. Why? Because earlier this year, NPR developed a new method to do all of the above and tested it in the NPR One app. We embedded some additional metadata into podcast episodes that marked important points in the file. We also embedded a URL for sending anonymous listening event data back to NPR's servers. Then, in NPR One, a simple model was built to read that data and send it back – a simple, anonymous ping when listening happened at those key points.
It's called Remote Audio Data, or RAD for short.
It worked just fine. Now, NPR can tell how much of the content people listened to, which sponsor messages, and even if people skipped the sponsor messages. And this technology is easy to scale. The problem is that NPR One accounts for just one portion of NPR's podcast listening. Listeners also hear NPR podcasts through Apple Podcasts and dozens of separate and independent platforms.
To move the entire industry to a listening-based metric, we would need a model like RAD adopted far and wide. We're working on that. NPR is preparing to move into phase two of testing RAD and will conduct a wider test along with several of the largest podcast publishers, multiple ad vendors and at least one other large podcast app. It works on both IOS and Android.
In the meantime, other changes are afoot.
Later this year, Apple will roll out Apple Podcast Analytics, a major step in sharing actual listening data with publishers. This is fantastic, but again, it only addresses the 65 percent of our audience on Apple platforms. At least from what Apple has shown thus far at the Worldwide Developers Conference (WWDC) it doesn't do it in a way that can easily become a transactable metric at scale.
We're talking to Apple about RAD and hope they will support it. We're also talking with most larger publishers and most large listening platforms. There seems to be a lot of interest and support.
My hope is that RAD, or something like it, can become an open industry standard in 2018. If we can achieve that, podcasting will become an even more valuable platform. We already know podcast ads are extremely effective at delivering on both branding and direct response goals. Tying that ROI with metrics that give marketers a clear understanding of how budget and listening combine to deliver value will make podcasting one of the most transparent and effective platforms in the modern media mix.
Photo credit: National Public Media.
Click the social buttons above or below to share this story with your friends and colleagues.
The opinions and points of view expressed in this content are exclusively the views of the author and/or subject(s) and do not necessarily represent the views of MediaVillage.com/MyersBizNet, Inc. management or associated writers.