Someone emailed me last month asking whether $25 per thousand downloads was a fair sponsorship rate. He had 1,400 downloads per episode. He had done the math, arrived at $35 an episode, and felt embarrassed by the number. He was about to tell a sponsor yes.
I told him not to send that email. Not because $35 is insulting, though it is. Because the whole formula he used was built for a business he is not in.
CPM pricing came out of radio and display advertising, where the buyer wants volume and does not much care who is on the other end. Podcasts got handed that model early because it was the only one anybody had. Then everybody kept using it, and now hosts with 900 real listeners are pricing themselves like a billboard on a quiet road.
The number everyone quotes and where it came from
You have probably seen the standard rates written down somewhere: $18 per thousand for a thirty second pre-roll, $25 for a sixty second mid-roll. Those numbers are real in the sense that big agencies actually transact on them. They are also close to useless if your show is under about 5,000 downloads an episode.
Here is why. At 1,400 downloads, a $25 CPM mid-read pays you $35. A sponsor who pays $35 for a spot has zero incentive to renew, zero reason to give you creative freedom, and no real stake in whether the read works. You have just made yourself the cheapest line item on their spreadsheet. Cheap line items get cut first.
A sponsorship price is not a calculation. It is a position. The CPM is one input, and for small and mid-sized shows it is not even the most important one.
What a sponsor is actually buying from a small show
I edit podcasts for a living, which means I sit with hours of raw audio every week and hear exactly how these reads land. A host reading a script at the top of the episode is worth very little. A host stopping mid-conversation to say "honestly, I have used this for eight months and here is the annoying part" is worth a great deal, and the difference has nothing to do with download counts.
What the sponsor gets from a 1,400-download show that they cannot get from a 90,000-download network placement:
- A named human vouching for them, not a voice-over
- An audience that is narrow enough to be worth targeting precisely
- Creative that does not have to be pre-approved by three people
- Someone who will answer an email in a day
Price that. A niche B2B show with 1,200 engaged listeners in one industry can defensibly charge $500 to $1,500 an episode, and I have seen hosts get it. The CPM implied by that is somewhere north of $400, which sounds absurd right up until you notice that the sponsor's customer is worth $40,000 and they need two of them.
Stop selling downloads. Sell access to a specific room you happen to own the key to.
Three ways to price, and when each one holds
CPM. Works when you are big, when the buyer is an agency, and when the show is broad enough that reach is the point. Above roughly 10,000 downloads per episode you will be negotiated into CPM whether you like it or not, so learn the numbers.
Flat rate per episode. Works for everyone else. You pick a number, you defend it, and the conversation moves from arithmetic to value. I have argued this before in a different context, and the same logic that makes flat pricing beat hourly billing for creative work makes it beat CPM for sponsorships. Both replace a metered input with a stated outcome.
Performance or hybrid. Affiliate codes, cost per acquisition, revenue share. Useful as a supplement, dangerous as your whole model. Attribution in audio is bad and always has been. If you go this route, negotiate a floor. A pure performance deal means the sponsor gets a free ad every time your audience forgets a promo code.
For most shows I talk to, the answer is flat rate with a soft performance kicker. Say $600 an episode, plus $20 per tracked signup over the first fifty. Now the sponsor is invested and you are not gambling your rent on a discount code.
The five inputs I would actually use
If you want a formula, use this one instead of a CPM multiplier.
- Audience value. What is a customer worth to the sponsor? A $9 app and a $4,000 consulting retainer should not pay the same rate for the same audience.
- Concentration. What percentage of your listeners are plausibly in the sponsor's buying group? A show where 70% of listeners are the exact target beats a show ten times bigger where 3% are.
- Read quality. Are you doing a scripted 30 second pre-roll or a two minute integrated segment with a story in it? These are different products and should have different prices.
- Exclusivity and term. Category exclusivity is worth real money. So is a four-episode commitment. Sell those separately instead of giving them away.
- Everything else in the package. The clip you cut from the read, the newsletter mention, the show notes link, the YouTube version.
That last one is where most hosts leave money on the table. If you are already producing vertical clips for LinkedIn and social, a sponsored segment cut into a standalone clip is a deliverable with its own audience and its own value. Charge for it. Do not staple it on as a freebie because it felt easy to make.
What I see hosts get wrong most often
Quoting a rate card before they understand the buyer. Ask what the sponsor is trying to do first. If they want brand awareness, sell reach and repetition. If they want signups, sell integration and a landing page. The number should come third, not first.
Discounting instead of unbundling. When a sponsor says the price is too high, the reflex is to cut it. Do not. Cut the package instead. Drop the mid-roll, drop the clip, drop the exclusivity, and re-quote. Now you have taught them what each piece is worth rather than teaching them that your prices are made up.
Selling sponsorship when they should be selling something else. Not every show should chase advertisers. If your audience is two hundred perfect prospects for your own consulting practice, a sponsor is a distraction and a tax on your credibility. This is close to the argument in when a podcast is the wrong channel, and the same honesty applies here. Sponsorship is one monetization model, not a graduation ceremony.
Ignoring what the episode sounds like. A sponsor is buying a slot inside your show. If the first ninety seconds of that show are a music bed and a mission statement, you have already lost a chunk of the audience the sponsor paid for. I have written about why podcast intros lose listeners at some length, and the retention curve is not subtle. Fix the intro before you sell the pre-roll.
Rate cards, media kits, and the paperwork nobody enjoys
Keep the media kit to one page. Downloads per episode over the last six months, audience description in two sentences, platform split, three placement options with prices, and a link to a sample read. That is the whole document.
Do not put a testimonial in there that you wrote yourself. Do not round 1,380 up to "nearly 2,000". Sponsors talk to each other and the podcast advertising world is smaller than it looks.
Two practical things worth doing:
- Charge for the production, not just the placement. If you are cutting a custom segment, mixing a bed under it, and delivering an alternate version for YouTube, that is work. It belongs in the price.
- Sell in packs. A single episode sponsorship rarely moves the needle for the sponsor, which means it rarely renews. Four episodes gives the message time to land and gives you predictable revenue. I price my own editing packages the same way for the same reason.
If the sponsored episode is going on YouTube, the packaging matters as much as the read. Our sister studio ClickTheory does thumbnails for exactly this reason, because a sponsored episode that nobody clicks is a sponsored episode you will not sell twice.
The taste part
There is a version of this post that is purely spreadsheets, and it would be wrong. Where you place a read, how long you let it breathe, whether you cut back to the conversation cleanly or let a beat of silence sit there, all of that changes how the read performs. It is the same thing I keep coming back to in editing as a taste problem: the decisions that matter are judgment calls, and no automated tool has an opinion about them.
We watch every episode we cut, all the way through, with a human on it. Part of the reason is that a sponsor read is the easiest thing in the world to ruin. Cut it too tight and it sounds like an ad. Leave it too loose and people skip. The right answer changes per host, per sponsor, per episode.
Where I would start on Monday
If you have never sold a sponsorship, do not build a rate card yet. Do this instead.
Write down who your twenty ideal sponsors would be. Not the ones you think you could get, the ones you would actually recommend. Work out what one customer is worth to each of them. Pick a flat episode rate that would look reasonable next to that number, not next to your download count. Then go have a conversation, not a negotiation.
Your first sponsor will probably pay less than your fifth. That is fine. What you are buying with that first deal is a case you can point to, a read you can play for the next buyer, and a price you never have to quote apologetically again.
If you want the sponsored segments in your show to sound like they belong there, that is the part we handle. Email me at samy@pixelsamy.com and send an episode. I will tell you honestly whether the read is working.