Friends Before Bots: Podcast Listeners Don’t Trust Algorithms, Survey Finds

I don’t know about you, but my go-to source for finding new podcasts has always been to ask friends for recommendations. In fact, I don’t think I’ve ever risked a new show unless it’s had the stamp of approval from someone I trust.

So when I saw the results of our upcoming Podcast Discovery Survey, I wasn’t hugely surprised to see a lot of listeners do the same.

But here’s what’s interesting. As algorithms, AI and advertisers become increasingly embedded into how podcast discovery works, our survey suggests listeners are turning even more towards the people they know and trust for recommendations.

Here’s what the survey data tells us about how listeners look for podcast recommendations in 2024 and how you can capitalise on this trend to attract more new listeners.

Listeners Prefer Recommendations Over Algorithms

The results were relatively broad when we asked survey respondents to think about how they discovered the last podcast they listened to.

But more than half (a combined 56.5%) of respondents said they found their most recent podcast through a human recommendation. Only 19% found their last show by browsing their podcast app, and a measly 3% thanks to an algorithm.

As you can see from the graph, all the top options (except for ‘browsing a podcast app’) involve human recommendations—and not money-backed recommendations, either.

The combined options in “Other” weren’t included in the graph because they got less than 3% of the vote. These included “mentioned on TV, radio, or news media” and “roundup or ‘Best of’ list.”

This suggests listeners want podcast recommendations from real people they trust. They prefer this over blindly jumping into a show that an algorithm, advert or chart recommends.

We dug down further to find out how respondents use their podcast app to find new shows. Only 6% said, “I see what the app’s algorithm recommends based on what I already listen to”.

Rather than trust an algorithm recommendation, they prefer to type a topic into the search bar and do their own research.

Listeners Less Trusting of People They Don’t Know

Some interesting comparisons arise when you compare this year’s survey with data from the same survey we ran in 2020. One important takeaway is that listeners are now much less likely to trust recommendations from people they don’t know.

In 2020, 80% of respondents said they took a recommendation from influencers or celebrities online, and only 20% said they took one from people they know.

In 2024, the numbers have flipped. 60% said the recommendation came from someone they know, and only 40% were from strangers or “influencers” online.

As the podcasting industry becomes more saturated with brands, influencers and advertisers all looking to make a buck, it seems listeners are becoming more cautious of commercialisation.

As one respondent said:

“I’m always on the lookout for podcast recommendations from people and sites I trust. The lists of ‘top 20’ podcasts are usually not very helpful, as they tend to be just based on what’s already popular, and that tends to mean it’s a tie-in with something else or trading on the name of a famous presenter rather than something genuinely interesting.“

As we know, a sponsorship deal or a place in the Apple Podcast charts is no guarantee that a show will be your cup of tea. The beauty of podcasting is that creators (thankfully!) don’t need to pitch or wait for commissions to make the shows they want to make. This means that, unlike TV, a lot of our favourite podcasts have never, and will never, make it to the charts.

AI is Having Very Little Impact On Discovery

Last year, I wrote a very hopeful article predicting the impact AI was about to have on podcast discovery. But while AI has had an undeniable impact on podcast production, the mark it’s made on discovery has been… overwhelmingly underwhelming.

It’s frustrating because, unlike content creation, discovery is one area of podcasting that has problems AI could solve. I think it’s fair to say you’d get very little resistance from listeners and creators if AI were to take over more of the discovery process.

The reason AI doesn’t feature in any of the graphics—and particularly the first one in this article about the ways people find new podcasts to listen to—is because that option didn’t get a single hit.

Algorithms barely hit 3%, and recommendations from AI tools got 0% of the vote.

Across the board, barely any respondents selected options that involved AI in the survey. But this is also because AI hasn’t really innovated discovery yet. There’s still hope that it could do in the future.

How to Capitalise On This Trend

So, while the rest of the content world strategizes on how to work with algorithms to increase visibility, podcasters face a different challenge.

How do you get more people talking about and recommending your podcast?

For a start, you can make sure you’re doing the following…

• Nurture your community. Reply to feedback and make genuine connections with your listeners so that they feel like they’re part of something.
• Ask listeners to recommend your show to others. If you don’t ask, you don’t get! Sometimes, listeners just need a nudge to tell their friends.
• Collaborate with other podcasters. Whether through guest appearances or episode swaps, collaborating with other podcasters allows you to tap into a new audience that already trusts the host’s recommendations.

And most importantly of all, keep making great content people want to shout about! Take care of your listeners and they’ll take care of you.

We’ll be releasing the full report of our Discover Survey later this month. Subscribe to the Podcraft Perspectives newsletter to get a first look at the data!