Small Shows Hit Different
I’ve been saying this for years. Now I have the data to back it up. Plus a calculator you can use!
Last week I published a breakdown of 25 sponsorship campaigns across 12 brands. Real ROAS numbers. Real attribution gaps. The categories that converted and the ones that didn’t. If you haven’t read it yet, start there. This piece builds on it.
The short version: the campaigns that underperformed almost never failed because of show quality. They failed because the infrastructure to capture what was happening didn’t exist. And the campaigns that overperformed? They almost always came from shows that had no business winning on paper.
One affiliate platform captured less than 8% of the revenue that Shopify recorded from the same campaign. A car dealership in Boulder drove more than $440K in gross transaction volume from a podcast that most major brand buyers wouldn’t have taken a meeting with.
The data isn’t an anomaly. It’s a pattern. And it points to a structural problem in how most brands evaluate podcast partnerships.
The Metric That’s Lying to Everyone
Download count is the metric everyone asks about first. It’s also probably the least predictive number in the conversation.
Here’s why. Download count measures reach. It does not measure trust, specificity of audience, purchase intent, or the degree to which a host is woven into their listener’s actual life. Those things are harder to quantify, so they get ignored. But they’re the things that actually drive conversion.
When Rambling Runner drove 65% of total conversions in a four-podcast campaign for a nutrition brand, despite not being the largest show, that wasn’t luck. That host’s audience runs. They buy nutrition products. They trust the person talking to them. The signal was always there. The infrastructure to see it wasn’t.
This is the mistake I keep watching brands make. They optimize for the metric that’s easy to see and miss the one that actually matters.
Why Affiliate Deals Are a Forcing Function
Here’s something Matt Cisneros at Backyard Ventures said after reading the data piece:
“One thing I definitely suggest is for smaller shows to test with affiliate deals because you can have them share the data. And once you have that data, you can use it to your advantage for further sponsorships. Getting to know your audience as well as you can is a huge advantage when deciding what brand partners are good fits for your personal brand and audience.”
He’s right. And it’s the exact model I’ve been building toward. I just wrapped up my annual audience survey to help me further understand what they care about. I’ll likely share more about that soon too.
A 30-33% affiliate deal with no flat fee does two things simultaneously. It removes the risk from the brand side entirely. And it forces the data to the surface. When a brand knows every conversion is being tracked back to a specific show, they suddenly care about the data. They share it. They get curious about what’s working. That data becomes the evidence that either proves or disproves what you suspected about a show’s audience.
That’s the bet. Run enough of these and you stop guessing about which shows punch above their weight. You know.
I built a calculator to show the unit economics of this structure. If you’re a brand evaluating whether an affiliate deal makes sense, you can run your own numbers here. The math is usually more favorable than people expect, because the frame is wrong. It’s not revenue share. It’s CAC reduction.
At a $60 AOV and 30% affiliate, you’re paying $18 per customer acquired (at 30% affiliate fee). If your current CAC is $35, you just cut it nearly in half. And you only pay when something converts.
That said, the structure only earns its place when the math closes on both sides. You need either enough conversion volume for the percentages to add up to real money for the host, or a high enough AOV that each individual transaction is worth the effort. A $29 product converting 10 times a month means the host makes $87. That’s not a partnership; that’s free awareness with extra steps. The filter isn’t affiliate vs. flat fee. It’s whether the economics actually work for everyone at the table.
The goal isn’t to keep shows on affiliate permanently either. It’s to use the structure as proof. Graduate to a flat fee renewal (perhaps with a lower affiliate percentage) with real data behind it instead of just a pitch.
The Infrastructure Problem
The affiliate model only works if the infrastructure exists to track it properly.
This has been the missing piece in podcast partnerships for as long as I’ve been doing this. Promo codes are a proxy for attribution. They’re better than nothing, but they leak. People forget to use them. They buy directly. They hear a podcast ad on Tuesday and Google the brand on Saturday. The code never gets entered and the conversion never gets credited.
The point I keep coming back to: the future of creator partnerships isn't coupon codes. It's campaigns that are woven into a creator's actual life. A product they use, recommend genuinely, and can link to in a way that captures what happens next. That's a different signal than a promo code pushed during a mid-roll.
The platform we're pushing to integrate across the Long Run Labs network this spring is built around exactly this problem. Better attribution infrastructure means the data that was always there; conversions happening days or weeks after an episode aired, purchases that never got a promo code entered; starts getting captured. That's the data that tells you whether a show actually works. It's also the data that makes every future partnership negotiation easier, because you're not asking a brand to trust your audience on faith. You're showing them what happened."
What This Actually Looks Like
Two shows from the same wearables campaign tell the story better than any argument I can make.
Ally Brettnacher hosts Finish Lines & Milestones, a show within both the Long Run Labs Network and SandyBoy Productions. Not a large show. The kind of show most brand buyers decline without a second look based on download count alone. Her fee for the entire campaign was modest.
Her return: 3.4x on spend. Every dollar of it invisible to the brand; the affiliate platform captured nothing from her show. If the brand had evaluated her performance from the affiliate data alone, they would have concluded she drove zero revenue. She didn’t drive zero revenue.
For The Long Run is also not a large show by conventional media metrics. Same campaign, same product category. FTLR drove 86.5% of total Shopify revenue across all six shows; 6.4x more than the other five shows combined. An 11.5x return on ad spend.
Across the entire campaign, the affiliate platform showed the brand 7.7% of what Shopify actually recorded. The other 92.3% was invisible.
Two small shows. One campaign. A combined return that no media buyer would have predicted from the download numbers. The data was always there. The infrastructure to see it wasn’t.
That’s not a fluke. That’s what happens when hosts have genuine affinity with a product, a specific audience, and the trust that comes from years of showing up consistently for the same people. That affinity isn't manufactured. One host in the network responded to a Go Brewing opportunity this way: "I stopped drinking alcohol a couple of years ago after a challenging journey with it. I now drink a lot of NA beer, including Go, which I like. This is easy for me to promote." That's the signal. Not the download count.
The Piece That’s Still Missing
The challenge I keep running into isn’t convincing brands that small shows can work. Most marketers intuitively get it. The challenge is giving them a framework to evaluate it that isn’t just “trust me.”
That’s what the affiliate batch is designed to solve. A set of campaigns structured specifically to surface the data. Brands that are willing to run a de-risked structure in exchange for transparency about what happened. Shows that have the audience and the trust. Infrastructure that captures what actually converts.
When you have 50 campaigns worth of data at that level of specificity, you stop having an opinion about which shows perform. You have evidence. And evidence is a different kind of leverage in a sponsorship conversation. The next layer is matching. Scrape a brand's reviews, build a customer persona, map it against show audiences. That's when the network stops being a directory and starts being a recommendation engine.
Matt’s point lands: the affiliate deal isn’t just a pricing structure. It’s a data acquisition strategy. The brands that get that first are going to have a significant advantage in how they evaluate and build creator partnerships going forward.
Where This Is Headed
The network is 35+ shows. 700,000+ monthly downloads. But the number I care most about isn’t any of those. It’s the number of campaigns where we have clean attribution data that actually tells the story of what happened.
Right now that number is 25. The goal is to get it to 50, then 100. At that scale the patterns become predictive. You can look at a show, an audience, a product category, and say with real confidence: here’s what this is likely to return.
That’s the business I’m building. Not a podcast network that sells on reach. A network that sells on evidence.
If you’re a brand exploring podcast partnerships and want to run the unit economics on an affiliate structure, the calculator is here (with some screenshots below).
And if you just want to follow along as the data gets messier and more interesting, this is the newsletter for that.
You just read 1,500 words about measuring return on ad spend. If that’s relevant to your job, the paid tier ($80/year or $8/month) is probably worth it; we publish campaign-level data, attribution analysis, and network insights that don’t make it into the free posts.
The math tends to work out.
Jon Levitt is the founder of Long Run Labs, an endurance and outdoor podcast network, and the host of For The Long Run. This newsletter covers the business of creator partnerships, sponsorship strategy, and what the data actually shows, in addition to a weekly article from that week’s Long Run Labs Podcast.