When a podcast works for a client, we lock it in for the year (or longer if they let us). But the question is always asked, “if we advertise on a podcast so regularly, won’t we oversaturate the audience?” While the simple answer of watching the performance on any given show over time and optimizing accordingly is acceptable, data shows that podcasts may have much longer shelf-lives than one may think.
Let’s say a network reports 100,000 downloads per episode for ”The Giles’s Cheesy Nugget Podcast”. These 100,000 downloads are NOT the same 100k people every week because not all listeners will download the next episode.
Assuming the overall download number stays the same each week, the size of the total audience a podcast reaches is much larger than the downloads reported. People stop listening to shows and new listeners take their place.
By estimating the amount of audience reach of a podcast over time, we can project the amount of incremental reach in any given week. Recently, we have been able to make some simple projections of how the original audience decays (“audience retention”) using data published by a network partner. The data suggests, for example, that approximately 20% of the listeners in a given week will NOT download the show following week.
Our best estimate of the decay is shown in the following images:
As the number of original listeners declines, the number of new listeners increases as shown in the graph below.
Based on this data, a show with 100k downloads will reach over 180k people within a given year.
This is one of the reasons why separating podcast integrations across multiple weeks may work so well for advertisers. For example, let’s say you advertise on a podcast once every three weeks. When you advertise again in week 4, an estimated 42% of the audience is BRAND NEW! By the time you run your next insertion, another 14% of the audience is also new. Again, this data is assuming the total audience is flat—a conservative assumption in many cases at a time of rapid growth in the space.
However, two drops in a 100k download show doesn’t mean 100k people have had 2 ad exposures. Since about 20% of the 100k are new listeners (20% have dropped out from the first week), you have 200k impressions and 120k audience = an average frequency of around 1.7 per listener. The actual amount of ad frequency will depend on how frequently you run ads as illustrated in the charts below.
Based on this data, to generate 3 ad exposures (on average) you’d need to run 9 weeks at a frequency of every other week (i.e. your 5th insertion). So if you’re running on a podcast at the recommended frequency of every 3 weeks, after an entire year, the average listener has only been exposed to your ad 6 times!
The implications are huge.
First, it’s unlikely that you’re oversaturating the audience of a particular show. A decline in performance is more likely due to other factors such as audience mismatching, rate increases, or decline in overall listenership. But even if you have hit a point of oversaturation and are seeing diminishing returns, spacing out your flighting or taking a short month or two hiatus from a show may give it a good chance of working again.
Second, given its relatively low average frequency, podcast creative doesn’t need to be refreshed as often as higher frequency mediums like radio.
Third, the data suggests that the industry may want to rethink how we look at CPMs, given the reach is not truly equal to the downloads. We could look at metrics like cost-per-thousand-reached, for example, but ultimately, you’re getting more than you pay for.
Finally, when planning your podcast campaign or evaluating performance, consider the fact that you’re not bombarding listeners with your message nearly as often as you think, and your message is reaching far more people.