Many of advertisers’ top questions about podcasts are about metrics. Understandably, they want to know how we measure impressions and impact so they can evaluate the return on their investment.
While podcasts are digital media, podcast ads are not tracked like digital display ads. However, the unique native-like podcast ads that Midroll show hosts create deliver results that are superior to display, and better than traditional ad platforms, like radio. We know this because our clients report back on their successes, and we invest in studying and measuring outcomes ourselves.
This FAQ is intended to answer the most common and important questions about podcast advertising.
- How are impressions measured?
- How do downloads compare to audience ratings?
- How do we know that a download resulted in an impression?
- Isn’t there a pixel technology for tracking podcast impressions?
- How do we know that podcast ad campaigns are effective?
- How do podcast advertisers measure the success of their campaigns?
- What are best practices for effective podcast ad campaigns?
- How do we know that podcasts work for brand campaigns?
- Do podcasts work for brand lift?
1. How are impressions measured?
Ad impressions are measured by downloads; specifically, how many times an episode that contains an ad is downloaded. An ad in an episode that is downloaded 200,000 times can be considered to have 200,000 impressions.
2. How do downloads compare to audience ratings?
The advantage to a download count is that it is first-party data reflecting observable user behavior. Audience ratings, like those for radio or TV, are based on sampling methodologies that are subject to a margin of error. By comparison downloads are accurate and actual counts.
Midroll maintains the integrity of this measure by working only with podcast hosting and measurement services that use industry standard methods to ensure the accuracy of their reports.
3. How do we know that a download resulted in an impression?
Truth be told, we don’t know 100% for sure. And we would be skeptical of anyone who says otherwise. We acknowledge that listeners do not always listen to every episode they download. However, the success of most podcast ad campaigns combined with our own research indicate that listening rates are impressively high.
In our survey of 168,500 podcast listeners, 88% said they listen to most or all episodes of the shows they subscribe to. Moreover, according to Edison Research, 73% of podcast consumers listen to an episode within 48 hours of downloading it, and 88% listen within a week of that download.
Therefore we are confident that downloads as a metric are at least as accurate as broadcast audience ratings, which are based on statistical estimates.
It’s important to understand that no podcast ad platform can offer a guaranteed measure of ad impressions. That’s because podcasts are MP3 files, and the vast majority of listeners simply download them for playback on their smartphones. There is no communication between the podcast app and the podcaster’s servers, so no listening data is reported.
It is true that a small percentage of listeners use a web based player or a podcaster’s own app, either of which theoretically could deliver more precise listening data. However, Apple’s iTunes and Podcasts app still deliver about 65% of all podcast downloads, while most other Android and iOS player apps — like Stitcher, Pocket Casts, Overcast and Google Play — make up the vast majority of the rest. Together they greatly outweigh web players and podcasters’ apps.
4. Isn’t there a pixel technology for tracking podcast impressions that is similar to tracking pixels used for display ads?
We’ve heard that claim, but the reality is that the technology isn’t there. The reason is the same as why downloads are podcasting’s primary metric: the major listening apps do not support any listener tracking at all. Apple Podcasts, Google Play, Overcast, Pocket Casts, and Stitcher–none of these support pixel tracking or any other tracking mechanism.
Moreover, for a pixel to track audio impressions, the podcast listening apps would need a mechanism to load those pixels. But podcast apps simply play MP3 files. While an MP3 file may contain a cover art image, the MP3 audio specification has no means of embedding an image that is dynamically delivered by a server, eliminating pixel tracking as an option.
Getting more technical, when a podcast ad vendor claims to support pixel tracking, here’s what they actually can do. Podcast subscriptions work by using an RSS feed, which is a plain text file that points to each MP3 episode of a podcast. In theory, the URL to that RSS feed, and/or the URL to the audio files linked within the RSS feed can include a redirect. So if the episode’s address is www.earwolf.com/cbb1.mp3, you might instead tell the RSS feed to point to www.earwolf.com/pixel-tracking-redirect.php?file=cbb1.mp3.
Doing so means that when someone goes to download the cbb1.mp3 episode of the show, they’re first redirected through the pixel. But what does that pixel mean?
In reality, all that pixel tells us is that someone attempted to download the MP3. Not that they played it, not that they heard it, and certainly not that they got to the ad unit that’s being tracked. It is essentially a second, less accurate way to count downloads — which is the key metric we are able to measure.
5. How do we know that podcast ad campaigns are effective?
The first podcast advertisers were direct response companies that carefully monitor their ROI. Simply put, if they can’t attribute sales to ads on a platform, then they don’t continue using that platform. Dozens upon dozens of successful campaigns are why advertisers like Squarespace and MailChimp stick with podcasts year-after-year.
At this time Midroll works with over 200 client companies, ranging from emerging start-ups to established direct response advertisers and large national brands, like Dunkin’ Donuts and Procter & Gamble.
Midroll clients attest to the effectiveness of podcast ads:
- Online menswear retailer Mack Weldon now spends a full quarter of its ad budget on podcasts, based upon results like those from one ad on “Comedy Bang! Bang!” that doubled sales. Read the full case study.
- Kate Huyett, VP of marketing for online footwear retailer Bombas reports, “Podcast ads gave us more room to tell our story in a more compelling way… All the data we have suggests that they are profitable.” Read the full case study.
- After the company’s first 18 months of running podcast ads, Fracture CMO Herb Jones found, “What came to light is that podcasts work, and work well.” Read the full case study.
Furthermore, a comprehensive series of unaided recall studies conducted by Midroll found that recall rates with podcast ads is high. 80% of listeners could name at least one brand advertised on a podcast episode. Learn more below.
6. How do podcast advertisers measure the success of their campaigns?
The measurement method depends on the type of advertiser and campaign. These are the most proven tracking methods for the three most common types of campaigns: direct response, branding, and tune-in.
Direct Response: Offer Codes
Direct response advertisers often use an offer code that is tied to a bonus or discount. We recommend that a unique offer code is used for each podcast so that responses are easily attributed.
Direct Response: Custom URLs
A custom URL has the benefit of being one fewer thing for the listener to remember. Often, advertisers with vanity URLs have those custom links trigger a discount just like an offer code would. These are also useful for direct response advertisers, but may also be used by any campaign that has a web landing page.
Direct Response: Check-Out Survey
This method is often used when an advertiser doesn’t have a special bonus or discount to offer. Including ad copy asking listeners to fill out the survey helps improve accuracy.
Branding & Tune-In: Social Mentions
Increases in social mentions and discussions are strong indicators of branding campaign success. look for an overall increase in mentions during the time that your campaign runs, adding a few weeks after it ends to account for listeners who don’t listen to episodes right when they’re released.
Branding & Tune-In: Web Hits
Website and landing page visits are another good indicator that listeners are engaging with a brand or product. Advertisers look for increases during a campaign period, plus a few weeks. Even with a custom landing page advertisers should look at overall site traffic, since listeners may only remember a company’s main URL or use Google to find a site, which won’t be directly trackable.
Learn best practices and pro-tips for measuring and optimizing podcast ad campaigns in our Definitive Guide to Podcast Advertising.
7. What are best practices for effective podcast ad campaigns?
Longer campaigns outperform short ones. Midroll’s brand recall study demonstrates that campaigns that run ads on at least five episodes of a podcast generate the best unaided recall. This repetition helps to reinforce your brand message. Yet, because each live ad read is original, listeners don’t tune out like they do when hearing the same commercial repeated on the radio.
Advertisers new to podcasts often want to test the waters with just one spot on a single episode of a show. However, the results of such short campaigns often do not reflect the full advantage of podcasts. In fact, Midroll’s recall study found that campaigns with ads on five or more episodes result in 39% more listener recall than those with just one spot. Running spots on just four episodes improves unaided brand recall by 27% compared to running a single ad.
It’s also important to write great ad copy that gives hosts the right talking points and language to represent your brand and sell your product. At Midroll we like to provide hosts with either traditional ad copy or bullet points that address your key messages. In general, we prefer bullet points so that the host can make the spot his or her own, but if you need to tightly control your message a full script is fine. Midroll clients can rely on our podcast ad experts to assist with creating the best copy.
Learn more best practices and pro-tips for effective ad campaigns in our Definitive Guide to Podcast Advertising.
8. How do we know that podcasts work for brand campaigns?
Midroll conducted a comprehensive study investigating unaided recall by surveying listeners who heard podcast ad campaigns for four different national brands. A total of 11,123 listeners took a survey. 80% were able to name at least one brand advertised within a podcast episode, unaided.
By comparison, an Undertone and Ipsos ASI study found that full-page takeover display ads result in a 45% recall rate on mobile, and only 35% on desktop. The same study found that only 33% of respondents remembered even seeing banner ads.
When prompted with a brand name, 67% of listeners were able to name an actual product feature or specific promotion mentioned in a podcast ad.
Do podcasts work for brand lift?
Yes. In Midroll’s unaided recall survey of 11,123 podcast listeners, 51% said that hearing a podcast ad made them somewhat or much more likely to make a purchase from the advertised brand.