In Your Head. BBC StoryWorks, the branded content division of BBC Global News, has released the results of an intriguing and slightly terrifying study into the effectiveness of podcast advertising. This piece of work, called Audio:Activated, was “carried out across four continents by neuroscience researchers at Neuro-Insight”, which measured participants’ “second-by-second brain activity as content is consumed” to test the cut through of podcast advertising and branded partnerships.
The headline results of this study are positive, at least for those in the commercial side of podcasting. The engagement levels with podcast advertising were shown to be much higher than with advertising on radio (5 percent less) and TV (22 percent less). As an explanation for this higher benchmark, the study turns to the customary argument about the connection between podcaster and listener: “The intimate and conversational nature of the podcast environment creates an elevated state of engagement for brand mentions.”
Another finding that I suspect will be welcomed by those with pitches to write comes from the fact that the participants for this study were largely “ad avoiders” and already habitual podcast listeners (there’s more on the precise methodology here). Although a lot of the anxiety about skipping/fast forwarding the ads was allayed by the arrival of Apple’s in-episode analytics back at the start of 2018, further data is also nice, I guess.
For this aspect of the study the researchers also looked at engagement level with branded elements when the listener is doing something else, and found that secondary activities like driving and doing chores actually increases receptiveness to a brand’s messages: “The state of activity means that the podcast content is being taken in through ‘low-involvement processing’ – a much more ‘fuel-efficient’ process that has a lower cognitive load. As such, engagement with the content can last much longer.”
A final point to pull out from this — the study also found that “listeners create subconscious associations with the brand, based on words they hear in the podcast”. Which I suspect is an effect that we’ve all observed (try thinking of Serial S1 without “mailkimp”, for example), but it’s interesting to see it surface in a neuroscience-based study like this.
I would caveat all of this by saying that the sample for this work was small and dispersed around the world, and it would be premature to make any predictions or decisions based on it. But as a point of comparison for existing beliefs that the audio ad market operates on, it’s worth perusing.
Related (sort of): Vice has profiled the creator of AdBlock Radio, an adblocker for live streams and podcasts. Apparently it uses “speech recognition, acoustic fingerprinting, and machine learning” to detect when an ad is playing and block it. One thing that stands out as super problematic to me right away, along with the obvious and fundamental objection that ads help to pay creators: “He notes that the algorithm isn’t perfect, and that hip-hop music, for example, is often misidentified as an advertisement.” It also doesn’t work on host read spots, of course, which sound like the rest of the show.
Upfronts Cometh. I’ve banged on a lot in the past about how slow the UK podcast market has been to mature, and the prevalent idea I come across here a lot that there’s one big step change that needs to happen for the whole thing to blow up. I don’t really subscribe to this view — boringly, I think that things change in an iterative fashion, and that the path to a wider and deeper pool of profitable podcasts here will be paved with lots of tiny alterations — but here’s something worth noting in this context.
The IAB in the UK is wrapping up its week of digital upfronts with a day dedicated to podcasts for the first time, meaning that there will be several showcases in London on 3 October analogous to, although much smaller than, the Upfront event in San Francisco earlier this month. Acast, Audioboom, Dax, Spotify and the Guardian will all be taking part, with celebrity hosts like Jessie Ware and radio personalities like James O’Brien appearing.
Surveys galore. DiscoverPods, a podcast recommendation website, has put out the results of its third annual trends survey. Again, be mindful that this is a small survey of a very podcast literate group — 82.4 percent say they listen to 7 or more hours of podcasts a week — so I’m not sure this is something that there can be much read-across from, although the report is a worthwhile read. One stat that jumped out at me was the one about how “55.6 percent of respondents said they purchased an item after hearing it advertised on a podcast”, which is actually lower than I thought it would be, given the dedicated listening going on among this group.
Forget “Netflix for podcasts”. An app called Synth is pitching itself as the “Pinterest for podcasts”. It apparently auto-generates clips and quotes from episodes. Here’s how they describe it: “We think of our podcast clips as units of meaning (or the new atomic unit of podcasts). They are generated using our own machine learning algorithm that isolates topics of conversations within long-format recordings and surfaces their meaning.” Hmmm.