Whether listening on the morning commute to work or while doing mundane tasks, podcast listeners love the flexibility which audio content continues to afford them for both entertainment and education and advertisers have taken notice. According to Statista, podcast ad spend in the U.S. is expected to total $534 million in 2020.
However, there appears to be a growing trend in podcast ads that could lead to a breach of privacy for listeners. Namely, many podcasts are now tracking listener locations and in some cases behaviours across the web in order to serve more relevant ads. This privacy threat has grown to the extent that popular podcasting platform Overcast recently introduced notifications that allow listeners to see which podcasts are tracking listener data.
We’ve spoken to podcast platforms and ads experts to determine whether podcast ad personalization could be entering dangerous territory in regards to customer data and privacy.
How Do Podcast Ads Work?
Many podcast listeners will be familiar with podcast ads read by the host before or during the podcast. According to Stas Tushinskiy, CEO at interactive audio advertising company, Instreamatic, “The most popular ad format is host-read ads. This is when a podcaster shares his/her experience with a particular brand, or just reads a sponsorship message.”
But that’s not the only way to deliver a podcast ad. “There are also traditional audio ads run via podcasts: some podcasters connect to audio ad exchanges and, when a user downloads a podcast, an audio ad is pulled from one of the ad exchanges and stitched into a podcast. Usually, podcasters enable this capability for older episodes when a host-read ad becomes irrelevant,” he added.
Chris Frantz, VP of Marketing at Australia-based online video maker Biteable dives a bit deeper. He explains the nuances of ad purchases for podcast ads and where you can purchase ads from a publisher, broker or channel. “The publisher is the creator of the podcast and always the most cost-efficient purchasing option. The publisher will typically offer lower rates than a broker (such as Midroll) or a channel (such as Spotify) since there's no intermediary party. The content is often evergreen due to the ad being embedded in the stream directly, as opposed to a CPM bidding model which brokers and channels are moving to at the expense of performance gains for the buyer,” he explains.
Similarly to brokers in other industries, this group acts as the middleman between publishers and channels. “Brokers have connections and agreements with publishers all over the podcast landscape and were previously the only option prior to more ads being integrated into purchasing with channels. Typically, broker's will require a minimum of $10,000 initial purchase and they will recommend a package of publishers,” said Frantz.
Finally, there are ad purchases done at the channel level and the popularity is rising. “Channels are becoming the most popular place to purchase podcast ads. Spotify has begun to purchase and consolidate podcast content, with the recent purchase of the Joe Rogan show, podcast creation app Anchor, and the prior purchase of the popular Gimet podcast network. The benefit of purchasing through a channel like this is a smaller entry price and a DSP (demand-side platform) interface to optimize and plan.” said Frantz.
Should Podcast Listeners Be Worried About Their Privacy?
Podcast advertisements are booming business but could privacy become a concern for listeners? According to Stephen Smyk, SVP of Podcast and Influencer Marketing at performance audio agency Veritone One, “it depends on what privacy risks one is referring to, but the only data shared in an RSS feed with a podcast hosting company is the IP address that the request came from. Therefore, there are no cookies or any other tracking mechanisms in podcasting. Given that IP addresses are usually shared and are dynamically assigned, there is very little actionable data to use for targeting or other purposes. There are some who use the IP address to gather aggregate geographic data and to try and target ads based on that information, but it is not down to an individual level,” he said.
Tushinskiy believes that the privacy risk associated with podcast ads is unfounded. “In my opinion, no digital ad carries a privacy risk because ad networks and advertisers don't have access (and don't need access) to personally identifiable information like names, addresses, credit card info, etc. They are only interested in a user profile, (gender, age range, interests, etc.), which doesn't provide much from a privacy risk perspective,” he says.
Frantz, however, believes that there are indeed privacy issues that need to be discussed. “Your platform choice will carry the most implications of the risk to your personal data. Inherently, podcasts are just an RSS feed that can be streamed from any server, without any need to volunteer personal information or install an app. However, channels like Spotify create unified advertiser ID's that may combine user behavior with app behavior, which offers similar privacy risks to other digital ads,” he explained.
One thing is for certain, at this moment in time, podcast listeners have no way to approve or control the way podcast producers and platforms collect, store, and use their data, which is without doubt a worrying threat.
The Future of Podcast Ads
So what does the future hold for podcast ads if there is a clampdown on the supposed privacy issue? It could lead to ads which aren’t beneficial for the user.
“Fighting ad tracking would result in only one thing: more ad loads, as ad rates would decrease and publishers would have to compensate for the loss. And all of the ads are then going to be completely irrelevant,” said Tushinskiy.
The future of personalization is also likely to change. “There continue to be new ways to personalize ads, and one key way is through registration information. In the event that you were to listen to podcasts on a service rather than through anonymous RSS feeds, the show that is serving you the content can work with the service to deliver personalized ads based on whatever information that service has on the user. The industry is already moving in this direction, but the inability to use the primary method of delivering ads (host read, in content) along with the very small levels of inventory once you start diving into these small audience segments has reduced the enthusiasm of the advertisers to use this approach heavily,” said Smyk.