Skip to main content
ABC News
But First, A Word From 100 Podcasts’ Sponsors

This story was produced in collaboration with Marketplace.

If you spent the beginning of February listening to the latest episode of each of the top 100 shows on the iTunes “top podcasts” chart, you’d have heard “Stone Cold” Steve Austin talk about the best way to deliver flowers. You’d have heard the Freakonomics crew spend 10 seconds telling you it was brought to you by Goldman Sachs. And you’d have heard 18 different voices give 30 different spiels for a website that helps people build websites.

We know because we did it. FiveThirtyEight data-reporter intern Hayley Munguia spent two days1 doing a fast-forward version of this aural marathon, recording each ad in a spreadsheet. The data is imperfect2, but then again so is all data about podcasts. While TV and radio have Nielsen ratings, podcasts have only numbers that the shows themselves gather and volunteer — usually the number of downloads but not the number of listens — and demographic information acquired using voluntary online polls. The “top 100” spreadsheet is a small flash of light in a fairly dark corner of media. It offers a glimpse at the strategies, categories and wild variations of a stubbornly niche market — and the handful of signals that suggest it may be on the brink of change.

Superficially, the February snapshot confirms what any avid listener will suspect:

  1. The typical podcast advertiser is the 2015 version of a mid-tier dot-com. Out of 186 ads, 162 — roughly 87 percent — were for products or services that acquire customers online: Web-based services for businesses (e.g., Web-based services for consumers (e.g. Squarespace) and Web-ordered physical products (e.g. Dollar Shave Club).
  2. The typical podcast advertisement is the ur-example of the low-budget “native ad.” It is a podcast host or podcast producer speaking into a microphone. There was only one ad-agency-produced radio spot with voiceover and music in 186 ads.
  3. The typical podcast ad doesn’t just tell you about the product, it also tells you how to become a customer and gives you a coupon code that’s both an incentive to shop and a way to track the ad’s effectiveness. You’ve heard this type of ad. It’s the one that says: “Go to’sName for a non-neglibile percentage off your next purchase.” These are called direct response ads, and Midroll, a company that sells ads for hundreds of podcasts, says the percentage of this type of ad has actually increased in recent years — from roughly 60 percent to roughly 80 percent — though they expect that percentage to decrease in the coming years. Fully 89 percent of ads on Midroll-represented shows in the February snapshot featured a direct response coupon code.

There are three commonly given reasons for why podcasts are dominated by native, direct-response ads for digital products: small audiences (“Serial” aside), antiquated technology, and the fact that native ads are just the way podcast advertising has always been done.

Podcast listener? Subscribe to FiveThirtyEight’s shows here.

Podcasts as we know them today are fundamentally a holdover from an earlier era. Subscribe using your smartphone or computer, and the device will download the latest audio file when it’s released. The multistep process creates a modest technical barrier to listening and an even higher barrier to gathering any data on listeners: A podcaster can tell a file was downloaded, but knows little about the downloader and nothing about what they do with the file afterwards. Ad sellers put a positive spin on the barrier to listening, saying it yields a more tech-savvy listenership, ripe for digital advertisers.

The barrier to gathering data makes it a natural fit for national advertisers (since you can’t target listeners based on their location) that use direct response methods. “Direct response advertisers are fine with rough measurement,” says Erik Diehn, vice president of business development for Midroll. “They’re going to measure for themselves. But for brand advertisers, who don’t measure direct impact of the ad, they tend to be very, very concerned about delivery.”

As for why these ads sound like endorsements, the usual explanation is that that’s just how the medium started — advertisers moved to podcasts from commercial talk radio. “It turns out it’s an incredibly engaging ad format and gets better results than some guy screaming in your ear,” says Diehn.

In other words, if it ain’t broke, don’t fix it.

Sidebar: Anatomy Of A Podcast Ad

Read more about how a podcast ad comes together here.

But the data also suggests that it is a little bit broke.

The top five radio advertisers are Home Depot, T-Mobile, Geico, Comcast Cable and AT&T. In contrast, the top five podcast advertisers in our February snapshot were Squarespace (30 ads on 18 shows), (12 ads on 11 shows), Audible (11 ads on 7 shows), MailChimp (8 ads on 7 shows) and Dollar Shave Club (5 ads on 5 shows). Those 66 ads represent 35 percent of the ads Hayley heard. A similar proportion of ads (31 percent) came from advertisers that appeared on only a single podcast in the top 100. In other words: Podcast ads aren’t all Squarespace and Audible — those are just the big fish in a pond full of minnows.

Still, more than a third of podcasts — 38 out of 100 in the February snapshot — had no ads at all. And of the remaining 62 shows, the median number of ads was two — which for many shows is a less-than-ideal sell-through rate (industry lingo for the percentage of available ad slots that were actually sold). Midroll’s head of ad sales, Lex Friedman, says a typical show for them might have six ad slots — two at the beginning, three in the middle and one at the end. “If a show gets 50 or 60 percent sell-through, we’re happy with that,” he says. “But we’re not satisfied with that.”

This is one reason all the podcast-ad salespeople I spoke with for this article talked about reaching beyond the stable of coupon-code-quoting dot-coms, just as it’s one reason WBEZ, WNYC and NPR held the first “podcast upfront this week. They want to tap the brand advertisers that are the biggest spenders on radio and TV. “If you can get Coca-Cola or car companies or studios, they have deep pockets, they have big budgets,” Friedman says. Those big budgets let them buy in bulk, spending less for each individual ad but filling more of those empty ad slots.

For the shows that sell 100 percent of their ad slots (which Midroll says is 5 percent of its shows), that may not be an attractive proposition — but for the vast majority of podcasts that don’t, brand advertisers carry the promise of a new, more stable revenue stream.

The February snapshot suggests, though, that big brands’ foray into podcasting is still in its infancy. Big-name brand advertisers made up less than 5 percent of ads in our database. It’s a short list: Geico, Goldman Sachs, Subway, “tune-in ads” for WWE and “American Experience” on TV, and — depending on what you count as a big name — tomato sauce companies Muir Glen and La Famiglia delGrosso. (While not exactly “household names,” neither of the tomato sauce companies sell products directly to consumers from their websites.) While they may have the budgets to buy in great volume, they don’t appear to be exercising this power yet — none of these advertisers ran more than a single ad in our snapshot.

Geico ran one ad on “The Adam Carolla Show.” The company is one of the top advertisers in radio, spending more on radio ads each year — upwards of $63 million in 2014 — than the entire podcast advertising market is worth, by most estimates. While the company wouldn’t discuss its advertising strategy, Geico’s overarching method of advertising everywhere, incessantly, is clear enough to anyone with a radio or a television. To them, podcasts may simply be another channel, one where they can reach people who may not be watching TV. “‘Here’s an audience that’s hard for us to reach every other way; let’s torture them this way.’ That’s how I view those ads,” says David Reibstein, professor of marketing at the Wharton School at the University of Pennsylvania.

Goldman Sachs has a different strategy. It ran a brief scripted-read on the public-radio podcast “Freakonomics.” But the investment bank doesn’t plan to be involved in podcast advertising for long. “The traditional, sponsored, ‘Brought to you by X, Y and Z’ — long-term, we’re not interested in doing that just to do that,” says Jake Siewert, head of corporate communications for Goldman Sachs. “We’re doing that more as a way of raising our visibility to consumers of podcasts as we build a market for our own podcast.”3

A company called The Media Kitchen was responsible for buying Goldman Sachs’ podcast ads. Derek Lu, a senior strategist there, talked up the possibilities of the format, but also admitted that none of the agency’s other clients — a group that includes Vanguard and Victoria Secret’s PINK — were doing podcast advertising, in part because the poor data and small scale of the medium were obstacles. “A lot of clients are focused on: ‘What am I going to get out of this? How much sales am I going to get? How many conversions? How many signups? How much traffic?’ This is not that area,” says Lu. “It’s hardly a space for everyone.”

“I feel like what a lot of people who write about podcasts want is for it to be a big advertising industry,” says Squarespace’s Ryan Stansky. “I don’t think it’s going to be.”

But the people who sell podcast ads nevertheless remain focused on the big brands. “We’d like to see more major studios on the entertainment side. We’ve only dipped our toes in with automotive,” says Midroll’s Friedman. “If you hear their spots on national radio, we want them.”

Asked for the key to cracking these new categories, he had a one-word answer: “Persistence.”

Listen to the Marketplace version of this story below:


  1. February 9 and 11.

  2. The “top podcasts” are not like the Billboard 100: It’s a list of shows, not episodes, and they are ranked by an opaque algorithm believed to rely on factors including subscription numbers, rate of increase in subscription numbers, user rating on iTunes and — depending on your level of paranoia — the arbitrary favoritism of someone in Cupertino. Also, human error could have played a role in shaping our spreadsheet.

  3. Goldman’s podcast is called “Exchanges at Goldman Sachs,” Siewert is the host, and it debuted in December.

Stan Alcorn is a reporter for Marketplace.