For hosts
How Much Do Podcasts Make?
Every answer to this question multiplies downloads by a CPM, and both numbers are private. There is a third number in the same equation that is public, and it decides more than either: how many episodes a show releases in a year. Across 279 still-publishing feeds we read, the median is 52.
8 min read
Podcast revenue is three numbers multiplied together: how many episodes you publish, how many people download each one, and what an advertiser pays per thousand of those downloads. Every article about podcast earnings gives you the second and third as ranges somebody else assumed, and skips the first entirely. That is backwards. Downloads are private, CPMs are negotiated, and the episode count is the one input that is published, checkable, and almost entirely under the show's own control.
So we counted it. Between July 16, 2026 to August 26, 2026 we read the public RSS feeds of a podcast directory; 279 of those shows were still publishing when we read them and carried enough dated episodes to measure a rate. Together they release about 26,576 episodes a year. The median show releases 52, 68% publish weekly or more often, and only 4% publish less than once a month.
That median is the useful frame for the whole question. 52 episodes a year with one host-read slot in each is 52 sellable impressions-blocks a year, and at the $25 to $40 host-read CPM benchmark our sponsorship guide sources, each 1,000 downloads an episode is worth $1,300 to $2,080 a year. Put your own download number over 1,000 and multiply. That is the honest answer, and it is smaller than most people expect.
How much do podcasts make?
Most make nothing, and the ones that make something make it slowly. The arithmetic is fixed: a year of ad revenue is episodes a year, times slots per episode, times downloads per episode divided by 1,000, times the CPM. At the median pace we measured (52 episodes a year) with a single host-read slot, a show with 1,000 downloads an episode grosses roughly $1,300 to $2,080 a year before anyone is paid for making it.
Read every row as "per 1,000 downloads an episode". We hold no download figure for any show and never will from a public RSS feed, so this table is denominated in a unit you supply rather than an audience we guessed. A show averaging 4,000 downloads an episode multiplies these figures by four. The pace rows are this corpus's own measured bands, not round numbers chosen to flatter the arithmetic.
| Episodes a year | At $25 CPM | At $40 CPM |
|---|---|---|
| 12 (monthly) | $300 | $480 |
| 26 (fortnightly) | $650 | $1,040 |
| 52 (weekly, our median) | $1,300 | $2,080 |
| 104 (twice a week) | $2,600 | $4,160 |
Three things that table makes obvious and a CPM range on its own hides. Publishing pace is a straight multiplier on revenue, so the gap between monthly and weekly is a factor of four on identical audience and identical rates. A show under 1,000 downloads an episode is not a business at any pace. And the number most articles present as the answer, the CPM, is the input you control least.
The number nobody publishes: how many ad slots a show actually has
Here is the distribution behind that median. These are the 279 shows that were still publishing when we read their feeds, banded by the rate their own recent episodes imply, over a median window of 28 days.
A rate, not a count of a calendar year: it is measured from the gaps between the newest episodes in each feed and annualised, which is long enough to rank a corpus and too short to judge one show. 13 feeds published faster than the window could bound and are held at a ceiling of one episode a day. "Share of all episodes" is what each band contributes to the 26,576 a year the whole pool publishes.
| Episodes a year | Shows | Share of all episodes |
|---|---|---|
| Fewer than 12 a year | 11 | 0% |
| 12 to 25 a year | 27 | 2% |
| 26 to 51 a year | 52 | 7% |
| 52 to 103 a year | 98 | 22% |
| 104 or more a year | 91 | 69% |
The distribution is top-heavy in a way that matters for anyone reading an industry average. The top band alone, "104 or more a year", is 33% of the shows and 69% of the episodes; the busiest 28 feeds, a tenth of the pool, publish 36% of every episode in it. An "average podcast" computed over feeds like these is dominated by a small number of near-daily formats and describes almost nobody.
At the other end, 11 shows (4%) publish in the "fewer than 12 a year" band. For them the ad arithmetic barely runs at all: at $40 CPM, 12 episodes a year is $480 per 1,000 downloads, which will not cover a hosting plan for most of the year. That is not a reason to publish more for its own sake. It is a reason to stop pricing a hobby as a business.
One caveat on the pool, because it changes who this describes. Shows that had stopped publishing are excluded by construction: a rate is a projection forward, and projecting one out of a dead feed would be plainly false. So every figure here describes ACTIVE shows, which is the optimistic half of any podcast directory.
Ads are the smallest of the four ways a show makes money
The arithmetic above is the ad business, and for most shows the ad business is the least of it. Four revenue shapes, roughly in order of how many shows they actually work for:
- Selling something of your own. A service, a product, a course, a book. It needs no minimum audience because it is priced per buyer rather than per thousand listeners, which is why the smallest shows on this list are usually the ones doing it.
- Direct sponsorship of a defined audience. A show with 800 people who all do one job is worth more to the right advertiser than a broad show with 8,000, and it is sold in packages rather than at a CPM. Our sponsorship guide has the rate-card math for this.
- Listener support: memberships, bonus feeds, tips. Converts a small share of an audience at a much higher value per person than an ad ever will.
- Programmatic and network ad inventory. The route the table above prices, the one that scales with downloads, and the last one to become meaningful.
The order is not an opinion about what is noble, it is what the arithmetic forces. One consulting client won from an episode can outweigh a year of $25 CPM inventory at the median pace, because the client is not priced per thousand downloads.
It also changes what pace is for. If revenue comes from ads, publishing more episodes multiplies it directly. If revenue comes from what a listener does after the episode, pace matters because it decides how often you are in front of them at all, and the ceiling is set by whether you can fill the episodes with something worth hearing.
What it costs to run, measured against what it brings in
A revenue figure means nothing without the bill beside it. Media hosting runs from about $15 to $99 a month depending on plan, which is $180 to $1,188 a year. At the median pace and a $25 CPM, that is covered by roughly the first 1,000 downloads an episode and nothing more.
So the break-even a small show should actually care about is not "when do I get rich", it is "when does the show stop costing me money". At 52 episodes a year, one host-read slot, $25 CPM: a $15-a-month plan clears at a few hundred downloads an episode, and a $99-a-month plan needs several times that. Our hosting-cost guide works that break-even through plan by plan with the current published prices.
The second cost is the one no invoice shows, which is the hours. If the show books guests, that work scales with pace: across the same pool of feeds, 2357 guest chairs a year are visible in total and the shows that fill any fill a median of 26. A weekly guest show is committing to a booking conversation most weeks, forever, and that is the cost that stops shows rather than the hosting bill.
The other two sides of this arithmetic
Both are our own pages, and both are the source for a figure used above rather than a further reading suggestion.
- Podcast sponsorship rates: CPMs, costs, and a calculatorWhere the $25 to $40 host-read CPM used on this page comes from, with the campaign math and the rate-card formula behind it.
- How much does podcast hosting cost?Current published plan prices from the hosts that publish them, and the sold-downloads break-even for each one.
- How much does it cost to find podcast guests?The other recurring cost of a guest show, priced per booking against the same corpus of feeds.
What this measurement cannot tell you
Four limits, and they bound every figure above rather than decorating it. First and largest: we measure OUTPUT, never audience. Nothing in this corpus knows how many people download anything, so no figure here is a claim that any show earns any amount. Every dollar on this page is per 1,000 downloads an episode, which is a unit you supply.
Second, the pace is extrapolated from the newest episodes in each feed over a median window of 28 days. A show that had a busy month reads fast and a show between seasons reads slow. It is a fair way to rank a corpus and a poor way to judge one show, including your own.
Third, 13 feeds publish faster than that window can bound and are held at a ceiling of one episode a day, so the concentration figure is a floor rather than an estimate.
Fourth, the pool excludes shows that had stopped publishing, and it excludes 22 feeds in the directory we could not read at all. Both exclusions push the same way: this describes the active, readable half of podcasting, and the true median across everything with a feed is lower.
What to do with the number
If you are trying to work out what your own show could make, do it in this order rather than starting from an industry average:
- Count the episodes you published in the last 90 days and multiply by four. That is your slot inventory, and it is the multiplier on everything below.
- Take your average downloads per episode over a consistent 30-day window. Divide by 1,000. That is how many times to multiply the table above.
- Multiply: episodes a year, times downloads over 1,000, times $25 for a conservative host-read rate. That is your gross ad ceiling with one slot an episode, not your revenue.
- Subtract hosting and anything else you pay for. If what is left is not worth the hours, ads are not your revenue model and the list in the previous section is where to look.
- Only then ask whether more episodes helps. Pace multiplies ad revenue directly and does nothing for an audience that was not going to grow.
The reason to start with your own inventory rather than a benchmark is that the benchmarks in this market are built on the top decile. In our own pool the busiest 28 shows publish 36% of all the episodes. An average that includes them describes a daily news operation, not the show you make.
Sources and benchmark date
The publishing-rate figures are derived at build time from our own read of 362 public podcast feeds between July 16, 2026 to August 26, 2026. The CPM benchmark is not ours and is cited rather than restated. Method, limitations and the raw data are public:
- Which Podcasts Actually Book Guests? (our study)
Original analysis of 362 public podcast RSS feeds from a 384-show directory. Source for the publishing-rate distribution, the concentration figures and the guest-chair counts on this page. Method, limitations and the full dataset are published with it under CC BY 4.0.
- Podcast sponsorship rates: CPMs, costs, and a calculator
The source for the $25 to $40 host-read CPM benchmark used in every dollar figure above, which that page sources in turn from Acast's published advertiser and creator guidance.
- How much does podcast hosting cost?
Current plan prices read from each host's own pricing page, with the sold-downloads break-even the cost section above refers to.
Frequently asked questions
How much do podcasts make?+
It depends on three numbers multiplied together, and only one of them is public. A year of ad revenue is episodes a year, times downloads per episode divided by 1,000, times the CPM. We measured the first: across 279 still-publishing shows whose feeds we read between July 16, 2026 to August 26, 2026, the median is 52 episodes a year. At the $25 to $40 host-read CPM benchmark, that median pace with one slot an episode is worth $1,300 to $2,080 a year per 1,000 downloads an episode. A show averaging 2,000 downloads doubles it. Most shows make nothing from ads, and the ones that make something usually make more from selling something of their own.
How many episodes does the average podcast release in a year?+
Across the 279 shows that were still publishing when we read their feeds between July 16, 2026 to August 26, 2026, the median is 52 a year and 68% publish weekly or more often. Only 4% publish less than once a month. Be careful with the word average here: the busiest 28 shows in the pool publish 36% of all the episodes, so a mean is dragged upward by a small number of near-daily formats. Note also that the pool excludes shows that had stopped publishing, so this describes active shows.
How much can a podcast with 1,000 downloads an episode make?+
At one host-read slot an episode and the $25 to $40 CPM benchmark, 1,000 downloads an episode is worth $25 to $40 per episode of gross ad value. Over a year that is $300 to $480 at monthly, and $1,300 to $2,080 at the median 52 episodes a year we measured. That is gross media value before anyone is paid for making the show, and it roughly covers a media hosting plan and little else, which is why shows this size usually monetise through their own products or direct sponsorship rather than CPM inventory.
Does publishing more episodes make more money?+
For ad revenue, directly and proportionally: pace is a straight multiplier on the number of slots you have to sell, so weekly is four times monthly on an identical audience at an identical rate. Of the three inputs to podcast ad revenue, it is the one most under a show's control. It does nothing on its own for the other revenue shapes, where what matters is what a listener does after the episode. And it is not free: at 52 episodes a year a guest show is committing to a booking conversation most weeks. Across the same feeds we read, only 23% of live shows visibly fill any guest chair at all.
What is a realistic podcast CPM in 2026?+
A practical planning range is $25 to $40 for a host-read sponsorship and lower for prerecorded or programmatic inventory. Those are Acast's published benchmark ranges, cited on our sponsorship rates guide rather than restated from a blog. Two things a CPM range will not tell you: it prices delivered, sold impressions rather than total downloads, so unsold inventory earns nothing, and small direct deals are frequently sold as flat packages because strict CPM math on a small show produces a price too small to cover the work of selling it.
More on what it costs
The hard part of publishing weekly is filling the weeks
Pace is the multiplier on every figure on this page, and for a guest show the thing that caps it is the booking. List your show free, say what kind of guest you want, and get introduced to people whose topics match the episodes you make.