The Bumper Score
The Bumper Score is a Bumper-derived metric that measures how well your show delivers ads to a Verified Audience. Scores range from 0 to 200, with 100 representing the average of all participating shows. Scores above this 100 benchmark signal stronger verified reach and higher value to potential sponsors.
It exists because of a gap that has held podcast advertising back for years. When a media buyer asks "how many people actually heard my ad?", the honest answer from most publishers has been "a lot, probably." Downloads are real, but a download is a file request — it does not prove that anyone pressed play. The Bumper Score is our effort to put a number on the part of your delivery that can be confirmed, using first-party data from the platforms where your audience actually listens and watches.
What the Bumper Score measures
The Bumper Score answers a single question:
If an advertiser buys this show, how likely are they to reach an audience whose presence can be independently confirmed?
That is a question about measurement coverage, not about audience quality.
What a low score does not mean
A low Bumper Score is not evidence that your audience is fake, inflated, or absent. It cannot be. The score is built from the data platforms are willing to report, and some parts of the podcast ecosystem report nothing at all.
A show can have a devoted, highly engaged audience and still score below 100, simply because that audience listens in apps that do not publish playback data. The score measures what we can prove — and the inability to prove something is not proof of its absence.
How is the Bumper Score calculated?
In essence, it looks at the number of audience members we can confirm were present at the midpoint of an episode, divided by the total number of times that episode could have been heard or seen.
Bumper Score ratio = audience confirmed at the episode midpoint
────────────────────────────────────────────
every download or view the episode received
We calculate that ratio for your show, then compare it against every other participating show.
The denominator: everything that could have been heard or seen
This is your show's total delivery — every download and every view, across every platform, whether or not anyone can confirm a play took place. It includes:
- Downloads reported by your hosting provider, broken out by the app that requested them
- Views reported by YouTube, where there is no download to count
- Spotify delivery, either from your host or from Spotify directly, depending on how your show is set up
- Downloads from apps that report nothing back — Overcast, Pocket Casts, Castbox, direct RSS subscribers, unidentified Android apps, and many others
The numerator: audience confirmed at the episode midpoint
This is the number of audience members we can confirm were present partway through the episode. We build it from first-party retention data published by three platforms:
- Apple Podcasts publishes an episode retention curve, which we apply to your confirmed Apple audience
- Spotify publishes a retention curve covering the period immediately after publication
- YouTube publishes an audience retention curve as a percentage of views
These three are summed. No other platform currently publishes the data required to contribute.
Which episodes count toward your score
The score considers the recent past: episodes published within the past three months that have been in market for at least seven days. The seven-day minimum gives platforms time to report retention data.
We add up confirmed audience and total delivery across all of those episodes and divide once, rather than averaging each episode's individual ratio. This keeps a small episode from carrying the same weight as one that reached ten times as many people.
How your score compares to other shows
The Bumper Score is a benchmark, not an absolute measurement. We rank every scored show on the Bumper Dashboard by its ratio, then map that rank onto the 0–200 scale so that the median show scores 100.
This has a consequence worth understanding: your score can move even when your show does not change. As other shows are added to the Dashboard, and as their data improves or degrades, the distribution shifts underneath you.
What is considered a "good" Bumper Score?
| Score range | What it means |
|---|---|
| 150–200 | Significantly above average verified delivery |
| 100–150 | Above average |
| 100 | Industry median |
| 50–100 | Below average |
| 0–50 | Significantly below average — large gap between delivery and verified audience |
Why is my score lower than I expected?
A score can be low for genuinely different reasons, and they call for different responses. Your ingredient breakdown in the Dashboard will tell you which one applies.
Your download count runs well ahead of your confirmed audience. Podcast apps routinely download episodes automatically for everyone who follows a show, whether or not that person ever listens. Feed crawlers, bots and pre-caching add more. All of it inflates the denominator without adding anyone to the numerator.
A large share of your downloads come from apps that report nothing. Your audience may be entirely real and deeply engaged — there is simply no first-party record of it.
Your audience drops off before the midpoint. This is the only one of the three that is a genuine engagement signal. If confirmed audience at the halfway mark is much smaller than the number who started, people are leaving before a mid-roll would play.
The first two are measurement artifacts. Only the third says anything about your content.
What the Bumper Score cannot tell you
- It cannot prove the absence of an audience. It can only report how much of your delivery is independently confirmable.
- It cannot see platforms that do not publish data. As more platforms offer consumption data to podcasters, more of your delivery will become verifiable — without you changing anything.
- It is less stable for shows with few recent episodes. With a short back catalogue in the window, a single unusual episode has an outsized effect.
Which shows can be scored?
Scoring requires per-episode, all-time download totals broken out by app. Not every hosting provider publishes that breakdown, and shows on providers that do not cannot be scored at all. See our FAQs for which sources are supported.
How often is the Bumper Score updated?
The Bumper Score is updated regularly on an ongoing basis. As episodes fall in and out of the three-month window, as delivery and Verified Audience for those episodes grow, and as other shows join the Dashboard, your score may fluctuate.
How can I increase my Bumper Score?
Your Bumper Score rises when the ratio from impressions to Verified Audience increases, and with higher episode-level listen and watch time.
Will advertisers be able to see my Bumper Score?
You have full ownership of your data, including your Bumper Score. Advertisers and agencies may request your show’s score, or you may choose to share it in order to increase their confidence in your show’s ability to deliver ads to real audience. However Bumper will never release this information without your explicit permission.