Podcast downloads are a good metric for one job only: telling you whether your reach is trending up or down over time. They are a poor metric for everything a B2B marketing leader is actually accountable for. A download records a file request, not a listen, not a person, and certainly not a buyer. Track it. Never report it as the headline.
That distinction matters more in B2B than anywhere else in podcasting, because your audience is small by design and your revenue does not come from ad impressions. A consumer show with 50,000 downloads an episode monetises attention in bulk. Your show monetises the fact that a VP of Engineering at a target account listened to 34 minutes of your episode on the way to work and replied to your BDR’s email three weeks later. Those are different businesses measured with the same dashboard.
What does a podcast download actually count?
A download is registered every time a client requests the audio file from your host’s server. That request can come from a streaming app buffering the first minute, a listener pulling an episode down before a flight, an app auto-downloading to a phone that never gets opened, or the same person replaying a favourite episode on a second device. One human being can generate several downloads. Several human beings sharing a home IP address can, on some hosts, be collapsed into one.
The IAB Tech Lab Podcast Measurement Guidelines (v2.1) exist to clean up the worst of this. Under v2.1 a download only counts when the same IP address and user agent combination has not already been counted inside a rolling 24-hour window, the request is not from a known bot or crawler, and at least 60 seconds of audio (or one megabyte, whichever is smaller) was requested. Platforms that do not apply this filtering can overstate real downloads by 50% or more.
So before you argue about whether downloads are a good metric, check whether your number is IAB-qualified. Comparing a raw feed-request count to a filtered benchmark is like comparing gross revenue to net. The units look identical and the meanings are not.
Two further things a download does not tell you. It does not tell you whether anyone pressed play, because the request and the listen are separate events and most apps never report the second one back to you. It does not tell you who the person is, which in a market where your total addressable audience might be four thousand people is the only question worth asking.
Why are downloads a misleading metric for B2B shows?
The misleading part is not the number itself. It is what people infer from it.
The default logic runs: more downloads means a better podcast, and a better podcast means more revenue. Podcast strategist Jeremy Enns has worked with hosts driving 50,000+ downloads an episode who struggled to cover the cost of making the show, and with hosts generating seven figures a year off a couple of hundred downloads an episode. One of his clients, running a show with well over 100,000 downloads a month monetised through a major ad network, had a record download month that produced about $1,200 in revenue.
If a hundred thousand downloads a month can be worth roughly the cost of a decent freelance day, the number is not a proxy for value. It is a proxy for volume, and volume is only worth something when you sell it by the impression.
The second problem is a measurement gap that shows up across the industry. In the 2024 Podcast Marketing Trends Report, 40% of respondents said their show’s purpose was to grow or support their existing business, and only 8% cited a revenue-related metric as their primary measure of success. In 2023 the split was 42% and 10.4%. Nearly half of shows exist to build a business, and fewer than one in ten measure whether they are doing it. The reason is not stupidity, it is friction: attribution from listener to customer is genuinely hard, downloads are sitting right there on the dashboard in a big font, and hosting platforms have trained everyone to treat the biggest number on the screen as the important one.
The third problem is that total downloads grows by definition. Publish more episodes, accumulate more downloads. A show with a large back catalogue can post an impressive monthly figure while new episodes land in the modest range, because new listeners are bingeing the archive. All-time downloads is a number that can only go up, which makes it useless as a signal and irresistible as a slide.
Downloads have also been gamed. Purchased downloads are a real practice, which is part of why the metric has lost standing as a measure of a show’s quality.
What is a good download number for a B2B podcast?
Buzzsprout, which hosts well over 116,000 active shows, publishes global platform data. Measured on downloads in the first seven days after an episode is released:
| Downloads in first 7 days | Where that puts you |
|---|---|
| More than 28 | Top 50% of podcasts |
| More than 104 | Top 25% |
| More than 428 | Top 10% |
| More than 1,050 | Top 5% |
| More than 4,763 | Top 1% |
Those figures were updated in March 2026. Sit with the first row for a moment. Twenty-eight downloads in a week puts a show ahead of half the industry. Four hundred and twenty-eight puts it in the top decile. The numbers B2B marketing leaders quietly consider embarrassing are, statistically, above average.
Other benchmark sets tell the same story with different cut points. Under 100 downloads per episode is normal for a new show. Around 1,000 per episode puts you in roughly the top 10% of active podcasts. Both sets exclude the graveyard of shows that published four episodes and stopped, so the real percentile against everything ever launched is more flattering still.
The honest answer to “what is a good number” is that it depends on the size of the audience your topic can reach. A show about procurement software for mid-market manufacturers has a ceiling. A show about Game of Thrones does not. Ask two questions: how many people are interested enough in this subject to consume content about it, and how many of those people listen to podcasts at all. Multiply those down and your realistic ceiling is often in the low thousands. Against that ceiling, three hundred engaged listeners is a very large share of the market.
There is a thought experiment worth borrowing. If you ran a podcast about sustainable fashion and your only listener was the head buyer at a major global retailer, would that be a failed show? Exactly who is listening matters more than how many.
What should you report to the board instead?
Report the numbers that would change a decision. A useful test: if a metric will not change your editorial or distribution choices in the next fortnight, and it will not change a budget conversation, it does not belong in the board pack.
Here is the reporting stack that survives a CFO’s questions.
Start with pipeline touched. Which named accounts have appeared in the show, listened to it, downloaded a clip, or been introduced through a guest. In B2B this is the number that justifies the line item, and it is available to you because your show is doing something a consumer show cannot: putting your buyers and their peers in the room. When a target account’s VP spends 45 minutes in conversation with your founder, the episode is a sales asset with a distribution channel attached.
Then engaged listening, not downloads. Apple Podcasts counts an engaged listener as someone who listens to at least 20 minutes or 40% of an episode. Spotify counts a stream at 60 seconds. Between them these two platforms account for roughly 70% of podcast consumption, and both expose listen time in their dashboards, so you can build a decent sample even without complete coverage. Completion rate above 70% is strong for episodes under 30 minutes; above 60% is strong for episodes over 45. Below 40% usually means your title and intro promised something the episode did not deliver.
Then unique listener growth over 90 days, which tells you whether you are reaching new people or recycling the same audience. Downloads at three times unique listeners means people are re-listening. A ratio near 1:1 means they stream once and do not come back. Neither is automatically bad, but they point to different fixes.
Then first-seven-day downloads per episode, expressed as a percentile rather than a raw figure. “This episode was in the top 10% of active podcasts in its first week” travels much further with a board than “we got 430 downloads,” because nobody in the room has an intuition for what 430 means.
Finally, the qualitative evidence that no dashboard produces. Replies from listeners. Guests who became customers. Prospects who mentioned an episode on a discovery call. Sales reps sending episodes into live deals. A single screenshot of a buyer quoting your show back to you in a deal thread carries more weight in a budget review than a chart of monthly downloads, and it is the kind of proof a B2B podcast agency should be helping you collect systematically rather than by luck.
So should you stop tracking downloads entirely?
No. Downloads still earn a place in the stack, for three reasons.
They are the only metric that works across the whole open RSS ecosystem. Every other number is trapped inside one platform’s dashboard. If you want a single trend line covering Apple, Spotify, Overcast, Pocket Casts and everything else, downloads is what you have.
The criticism that downloads are mostly automatic is weaker than it used to be. Auto-downloading has largely fallen away in modern apps, so most RSS listening now involves someone actively requesting an episode. Justin Jackson of Transistor makes the point that a download is not obviously worse than the high-level numbers on centralised platforms: a YouTube view is someone watching for a set number of seconds regardless of video length, and a Spotify stream is 60 seconds of listening. They are all rough proxies for attention, with the same weaknesses.
Direction of travel is genuinely informative. Shows with momentum show downloads climbing episode over episode, and when you check Spotify streams and YouTube views for the same shows, those climb in the same shape. As a health check, that is worth having. As a business case, it is not.
The practical position: use downloads as a trend, at episode level, over the first seven days, IAB-qualified. Never use total all-time downloads. Never open a board update with a download figure. Never set a download target as the show’s objective, because the fastest way to hit a download target is to make the show broader and less useful to the two hundred people who can actually buy from you.
How do you build reporting that connects listening to pipeline?
Start by deciding what the show is for before you decide what to measure, and write it down in a sentence that includes a business noun. “Create demand in mid-market manufacturing” is measurable. “Build brand awareness” is not.
Then instrument the paths a listener can take out of the show. A dedicated landing page per episode, a tracked link in the show notes, a named offer read out in the episode, a follow-up sequence to guests and to the audiences those guests bring. Attribution in podcasting will never be clean, and it does not need to be: what you need is enough surface area that some listeners identify themselves, so you can see the shape of the audience even if you cannot see all of it.
Add self-reported attribution to your demo form. A single “how did you hear about us” free-text field catches podcast mentions that no analytics platform will ever attribute, and it is the cheapest measurement upgrade available to you.
Then run the guest list as a target account list. Booking a buyer, a champion inside a target account, or a person your buyers trust turns each recording into a meeting you would otherwise have had to earn with cold outreach. That mechanism does not appear anywhere in your download chart, and it is frequently the largest commercial return the show produces.
One more discipline: report the same three or four metrics every month, in the same order, with the same definitions. Metric-shopping, where the number you highlight changes with whichever one looks best, is how programmes lose credibility internally. Pick your decision metrics and hold them.
Downloads measure reach. Completion measures whether the content lands. Unique listener growth measures momentum. Pipeline touched measures whether any of it was worth doing. Report all four and the download number stops being the argument, which is exactly where it belongs.
