Chapter: buying
How podcast advertising is actually bought
Four doors into the same market: direct, networks, programmatic, and agencies. Each one trades money for time in a different ratio.
On this page
Buying direct from shows
You email the show, or fill out the sponsor form on its website, and negotiate person to person. On smaller shows you're often talking to the host; on mid-size shows, a producer or a one-person sales operation.
Where it shines: testing a handful of niche shows on a small budget. There's no middleman margin, the host relationship is yours, and small shows will do flexible deals (flat rates, affiliate hybrids, bundled social posts) that no network would paper.
Where it breaks: scale. Every show is its own negotiation, insertion order, copy approval, air check, and invoice. At ten shows this is a part-time job; at forty it's a department. You also negotiate blind, with no view of what other advertisers actually pay.
Buying through networks and sales houses
Networks own or represent catalogs of shows and sell them through salespeople. One conversation can put you on a dozen shows, with professional trafficking, standardized reporting, and dynamic insertion capabilities across the catalog.
Where it shines: reaching big shows at all (most top-tier inventory is only sold this way), packaged buys across a genre, and operational simplicity: one insertion order, one invoice.
Where it breaks: the network sells what the network has. Their job is filling their own inventory, not finding the best shows for you across the whole market. Packages reliably include filler alongside the anchor shows you wanted, so price the package on the shows you'd have bought anyway.
Buying programmatically
Podcast inventory bought through demand-side platforms and audio marketplaces, delivered by dynamic insertion, targeted by audience segment, geography, or show category, and priced at $5 to $15 CPM.
Where it shines: reach and control. Frequency caps, audience targeting, instant creative swaps, and unified reporting next to your other digital channels. For produced-spot awareness campaigns it's the natural buying mode.
Where it breaks: it buys impressions, not endorsements. There are no host reads here, and cheap run-of-network audio can land on inventory you'd never have chosen deliberately. Treat targeting claims with the same skepticism you'd apply anywhere in adtech.
Buying through an agency
A specialist podcast agency plans the campaign, negotiates every buy, manages hosts and copy, verifies the ads aired as sold, and reports results against your goals. Compensation is typically a percentage of media spend or a retainer.
Where it shines: anything past the testing stage. Agencies know the real market price of inventory because they buy it every week, they hold relationships that get their clients better placements and makegoods, and they absorb the operational grind that otherwise lands on your team. On meaningful budgets, the rate difference alone often covers the fee.
Where it breaks: tiny budgets, where the fee can't be absorbed, and mismatched agencies. A generalist media agency that dabbles in podcasts is not the same animal as a specialist. The agency guide covers how to tell the difference and what to ask.
The four paths, side by side
| Path | Best for | Typical spend level | Your time cost | Watch out for |
|---|---|---|---|---|
| Direct | Testing niche shows | Under $25K | High | Blind negotiation, ops load grows fast |
| Network | Scale within a catalog, top shows | $25K+ | Medium | Package filler, catalog-first incentives |
| Programmatic | Targeted reach with produced spots | Any | Low | No host reads, inventory quality varies |
| Agency | Scaling with host reads across the market | $50K+/yr | Low | Specialists only; ask how they're paid |
The first-campaign playbook
-
Set the goal and the math
One action, one target cost per action. If a customer is worth $200 and you'll pay $50 to acquire one, a $15,000 test needs 300 attributable customers to break even. Write that number down before anyone shows you a rate card.
-
Build the show list on evidence
Listen to real episodes. Note how the host treats current sponsors, because that's your preview. Ask for IAB-certified download numbers and audience demographics. A media kit is marketing; certified stats are data.
-
Concentrate the budget
Three to five shows, three-plus mid-rolls each, six to eight weeks. Depth beats breadth in this channel, every time. Resist the urge to add one more show at the cost of frequency on the ones you have.
-
Brief, don't script
Give hosts the offer, three talking points, the required disclosure, and the tracking mechanism. Then let them talk. Insist on hearing the first read before the rest run, and fix problems through the brief, not through line edits.
-
Measure from day one, decide at week six
Codes, surveys, or pixels live before the first ad airs (the measurement guide covers the stack). At four to six weeks: cut the clear losers, renew the winners at better rates, and test lookalikes of what worked.
Expect the first campaign to produce one or two winners out of five shows. That is a good outcome, not a failure. The economics of the channel come from scaling winners and compounding frequency, not from batting a thousand on show selection.
Keep reading
Before you buy, price it and plan to measure
A buying path is only as good as the budget behind it and the tracking in front of it. Line both up before the first ad runs.
Check 2026 rates and budgets →
Then build your measurement stack.