In short: we don’t know.

There have been several pieces that speak of the presumed importance of podcasts in the 2024 US Presidential Election. Even Elon Musk says as much.

But I don’t think it’s as simple as that.

Let’s start with a few figures.

47% of Americans listen to podcasts every month, and 34% listen every week. That means 53% of Americans aged 12+ don’t listen to any podcasts in a given month.

But 47% is still a big number, and Trump went on the biggest podcast of the lot, The Joe Rogan Experience.

But the nature of podcasting means that podcast listeners themselves have a vast range of choice. And a significant majority of podcast listeners do not listen to the most popular podcasts.

This slide from the Edison Research Podcast Consumer 2024 report shows what I mean:

The yellow are at the top shows that 35% of weekly podcast listeners hear any of the top 10 podcasts in the US. That means 11.9% of Americans will hear at least one of the top 10 podcasts in a given week (35% of 34% of Americans). So perhaps Joe Rogan alone reaches 5% of Americans if we’re generous (there isn’t a publicly available precise number).

5% is a big number. But a 5% share is also roughly what an average Sunday episode of CBS’s 60 Minutes achieves (to be clear, a 5% share means 5% of households watching TV at that time see the show, not 5% of all adults in the US).

But beyond Rogan, the returns from every additional podcast a candidate appears on diminish quickly. 52% of American podcast listeners don’t hear any of the top 25 podcasts. And 39% of regular podcast listeners don’t hear any of the top 100 podcasts. 23% of regular podcast listeners don’t hear any of the top 500 podcasts, while 18% don’t hear any of the top 1000 podcasts!

You would have to be a guest on an awful lot of podcasts to reach every American podcast listener!

Then there’s the lack of available metrics on podcasts.

Forbes says Trump appeared on 13 podcasts, and they’ve tallied up the YouTube views each podcast has achieved, because there isn’t an open source of podcast audio listens. But YouTube is probably as important as podcast audio in this context since all the podcasts Trump appeared on also have YouTube versions of those episodes available. But as Peter Kafka notes in his piece on Business Insider, a YouTube view might represent watching for as little as 30 seconds. Whereas TV ratings commonly average the number of viewers across the entire duration of show – so it’s an utterly different metric.

Oh – and those publicly available YouTube views include plenty of non-US views (E.g. Rogan also has the most popular podcast in the UK), and reaching non-Americans doesn’t help you increase your share of votes.

We also haven’t really accounted for the overlaps between both similarly themed podcast/YouTube feeds. Given the nature of the titles Trump chose to go on, he is likely to have reached the same people multiple times.

Other things to take into account are the “preaching to the choir” element. Recent Edison research showed that Rogan listeners were already likely to vote for Trump. So how many Undecided or Harris voters switched as a result of his appearance?

But we also need statistics surrounding who actually got out and voted. Younger men, the natural audience of someone like Rogan, are historically less likely to actually go out and vote.

Maybe this time around, Rogan et al really did get people off their sofas and into polling booths. But I’ll wait for the statistical findings that show this before accepting any simple narratives.

Should political candidates go on podcasts? Of course. You get a very engaged audience if you do so.

But you need to be clear that there isn’t such a thing a single “hit” that you can do on one podcast to reach a really significant audience – Rogan notwithstanding. Really, podcasts/YouTube just need to be another part of the media mix a candidate needs to employ to reach a large audience.

But it’s important to consider the listeners to those podcasts and their inclination to actually cast their ballots. And that’s the proof that we’re missing at the moment.

Image above generated using Photoshop AI with the prompt: “a photo realistic ballot box sitting on a table in a white room”