A sponsor asks whether your podcast can deliver 5,000 unique US listeners, reach verified finance decision-makers, and run only in the United States. Your dashboard shows something else: 750 downloads per episode in the first 30 days, 70% of one month’s downloads attributed to the US, and a voluntary survey in which 55% of people who answered one question said they make or share a software purchase decision.
Those figures may all be useful. They are not interchangeable. Combining them would turn downloads into people, one month’s geography into episode-level targeting, and self-selected responses into a census.
A credible podcast media kit does not make a small show look large. It lets a buyer qualify five things quickly:
- Editorial fit: What does the show consistently cover, and for whom is it made?
- Measured reach: What exactly was counted, over what window, for which episodes?
- Available inventory: Which dated placements can actually be booked?
- Creative and measurement scope: What will be read, approved, disclosed, reported, and licensed?
- The next step: What must the buyer provide before either side treats the opportunity as a fit?
That makes the kit a qualification document, not an audience census, a rate card, or a promise that sponsorship will work.
What useful real podcast media kits show—and what they leave unanswered
Public media kits vary widely, but several reveal useful building blocks.
| Public example | Useful disclosure | Important missing qualification |
|---|---|---|
| Change Happened, Then What? media kit | Presents modest figures directly—“200+ downloads per episode” and “200+ listeners in 10+ countries”—then names pre-roll, mid-roll, post-roll, and custom options. | No measurement period, provider, survey basis for age and gender, dated availability, spot length, or reporting definition. |
| Footnoting History media kit, June 2024 | Separates 3.86 million total downloads, 16,000 average monthly downloads, and 3,800 average downloads per episode in 30 days. It also names Podbean as the source of its geography. | Does not identify a cohort behind the episode average or define sponsor inventory, creative terms, or reporting. |
| The Art of Online Business 2025 media kit | Dates its download snapshot, reports seven- and 30-day averages for the last 20 episodes, keeps monthly back-catalog downloads separate, and gives a campaign next step. | Its demographic and business-profile statements do not disclose a survey method, sample size, or question wording. |
| Trail Correspondents Season 4 media kit | States a dated per-episode figure, caps each episode at three one-minute ads, distinguishes pre-roll from mid-roll, and sets a three- to seven-episode commitment. | The download window, current availability, reporting terms, and present-day validity of the 2023 offer are not stated. |
A current sponsor page can add another useful pattern. The Art of Network Engineering reports first-seven-, first-30-, and first-90-day downloads separately, then distinguishes baked-in episode ads from a dynamic catalog campaign, newsletter placement, sponsored articles, category exclusivity, and short-form video. The structure is helpful. Some audience-role language is still stronger than the publicly disclosed method, so it should not be copied as verified purchasing authority.
The lesson is not to imitate one kit. Combine the strongest disclosures: a defined episode window, a dated data snapshot, distinct inventory types, explicit limits, and a real contact step. None of these public examples proves that its media-kit design caused a sponsorship sale or campaign result.
Choose one honest episode baseline
For a small show, “average downloads per episode” is often the most useful starting number—but only after defining the cohort and allowing every episode the same amount of time to mature.
The IAB Tech Lab Podcast Measurement Technical Guidelines v2.2 treat a download, an estimated listener, and an ad delivered as different metrics. A download is a filtered file request, not confirmation that one human heard an episode. A listener estimate depends on a stated time frame and technical identifiers. “Ad delivered” has its own conditions. The public v2.3 page checked on September 8, 2026 still linked to a document labeled for public comment, so this article uses the final v2.2 definitions rather than presenting v2.3 as final.
Consider this completed fictional dataset. All names, dates, and figures in the worked example are invented.
The Independent Money Desk publishes weekly. Its analytics cutoff is September 1, 2026 at 00:00 UTC. The most recent eight standard episodes with complete 30-day windows are F-E14 through F-E21.
| Episode | Release date | First-30-day downloads |
|---|---|---|
| F-E14 | June 8 | 680 |
| F-E15 | June 15 | 720 |
| F-E16 | June 22 | 700 |
| F-E17 | June 29 | 760 |
| F-E18 | July 6 | 740 |
| F-E19 | July 13 | 800 |
| F-E20 | July 20 | 780 |
| F-E21 | July 27 | 820 |
| Total / mean | 8 mature episodes | 6,000 / 750 |
The median is also 750, and the observed range is 680–820. That is a defensible statement:
The show averaged 750 downloads in the first 30 days across the most recent eight standard episodes with complete 30-day windows at the September 1 cutoff.
It is not the same as “750 listeners,” “750 ad impressions,” or “every future episode will deliver 750.”
The mixed-maturity trap
The five August releases had only 28.5, 21.5, 14.5, 7.5, and 0.5 days to accumulate downloads at the cutoff. Suppose someone selects the newest eight episodes, uses complete D30 figures for the first three, and uses partial-to-date totals for the other five:
(800 + 780 + 820 + 1,020 + 930 + 820 + 680 + 150) / 8 = 750
The result is also 750. The arithmetic is correct; the sample is not. Five rows did not receive a full 30-day window. A plausible aggregate can therefore survive a spreadsheet check while failing the measurement definition.
Preserve catalog context separately. In the same fictional fixture, all 26 episodes generated 6,400 downloads during August and 21,800 downloads since launch. Those figures describe the catalog over different periods. They are not extra people to add to the 750 baseline, and they do not create back-catalog ad inventory if the show cannot insert ads into old episodes.
Describe audience specificity in layers
A narrow editorial premise is valuable because it gives a sponsor a reason to investigate fit. It is not proof that every listener has the identity, job title, budget, or authority suggested by the topic.
Use three layers and label each one:
- Editorial positioning: what the show is designed to cover and whom it intends to serve.
- Observed distribution: provider- or platform-labeled activity such as download geography for a named period.
- Audience research: what respondents said, with the recruitment method, exact question, denominator, missing answers, and limits.
For the fictional finance show, the positioning is “practical money administration for US independent service professionals.” That sentence comes from the editorial plan. It should not be restyled as “an audience of verified business owners.”
Its August hosting export attributes 4,480 of 6,400 downloads to the United States, or 70%; 128 downloads, or 2%, have unknown geography. This is August RSS download geography, not verified residence, episode-cohort geography, or US-only delivery capability. Multiplying 750 × 70% would produce 525, but not 525 verified US listeners: the unit and population do not match.
The fictional show also ran a voluntary English web survey through show notes and its newsletter from August 1–21. It received 48 usable submissions, was unweighted, and had no unique invitation denominator. Listening eligibility, unique-human identity, and purchase authority were not verified.
| Exact survey item | Responses | Correctly scoped result |
|---|---|---|
| “Which best describes your current work?” | 36 independent service professionals; 8 employees; 4 other; n=48 | 75% of submissions answering this item selected independent service professional. |
| “Do you make or share decisions about purchasing bookkeeping or invoicing software for your work?” | 12 sole; 10 shared; 18 no role; n=40, 8 missing | 22/40, or 55% of item respondents, self-reported some role. They are not verified buyers. |
| “Which topics would be useful to you? Select all that apply.” | 32 uneven income; 24 tax-prep records; 18 retirement; n=48 | Multi-select percentages may overlap and need not total 100%. |
| “How many of the episodes released July 6, 13, 20, and 27 have you listened to?” | 30 selected three or four; 18 selected zero to two; n=48 | 62.5% reported that behavior; this is not platform-measured retention. |
The AAPOR Disclosure Standards are written for research disclosure, not as a legal checklist for podcast kits, but their logic is useful here: preserve question wording, recruitment, collection mode and dates, sample sizes, weighting, processing, and limitations. Do not calculate a conventional margin of sampling error for a self-selected nonprobability survey without an appropriate model. Do not multiply 75% by 750 and call the result “independent listeners.” You do not know that the survey respondents represent the episode-download population.
Platform analytics can add a separate layer, but they should stay platform-specific. Apple defines Engaged Listeners as people who listened to or watched at least 20 minutes or 40% of an episode, while Average Consumption is an average per device and can exceed 100% after repeat listening. Apple also requires at least five unique users for the selected metric and time frame to appear. “Not Enough Data” is not the same as zero, and an Apple subset should not be added to RSS downloads as though the counts were deduplicated.
When a needed audience field is unavailable, write not available or remove the claim. More importantly, change the offer: without verified US unique reach or geographic delivery controls, do not sell a US-only reach guarantee.
Define inventory as dates, placement, rights, and reporting
“We offer pre-rolls and mid-rolls” describes possibilities, not available inventory. A buyer needs to know which release, how many paid messages can run, whether the slot is open or held, what the host will read, where the message remains, and what report will arrive.
Here is the fictional show’s availability as of September 8, 2026. It permits one paid host-read mid-roll of up to 60 seconds per standard episode.
| Release | Status | Offer now? | D30 window closes |
|---|---|---|---|
| October 5 | Open | Yes | November 4 |
| October 12 | Hold H-001 until September 11, 17:00 UTC | No | November 11 |
| October 19 | Open | Yes | November 18 |
| October 26 | Open | Yes | November 25 |
| November 2 | Open alternative | Only if separately agreed | December 2 |
October capacity is four episodes, but current October availability is three. The show cannot confirm four consecutive October placements while the October 12 hold exists. It can offer October 5, 19, and 26 plus November 2 as a nonconsecutive alternative.
The kit should also state:
- Creative: host-read and identified as a sponsor message; sponsor supplies accurate product details, substantiation, CTA, URL, qualifications, and an approval contact.
- Editorial control: the host may correct or decline unsupported claims and will not invent personal use or enthusiasm.
- Included scope: named new-episode audio placements and show-notes links.
- Excluded scope: dynamic catalog insertion, geographic targeting, newsletter, social, video, category-wide exclusivity, and external use of the host’s voice or likeness unless separately agreed.
- Reporting: publication proof plus provider-labeled D7 and D30 episode downloads, naming the episode, source, window, cutoff, and timezone.
For US-facing advertising, the FTC’s Endorsement Guides Q&A is a useful boundary: paid connections should be disclosed clearly, endorsements should not say something the endorser does not believe or claim experience they do not have, and objective product claims need support. A media kit cannot settle every campaign-specific legal question, but it can make the production and approval responsibilities explicit.
Notice the reporting dates. The October 26 episode’s D30 window closes on November 25. “October campaign” does not mean the final report is complete on October 31.
Completed fictional podcast media kit: 17 fields
The following is a compact sponsor-facing version of the worked example. It is fictional and must not be presented as real analytics.
| Field | Completed entry |
|---|---|
| K01 — Identity | The Independent Money Desk, hosted by Example Host. Fictional show and host; no professional credential claimed. |
| K02 — Version | Kit: September 8, 2026. Analytics cutoff: September 1, 00:00 UTC. Inventory as of September 8. Source: synthetic hosting-export fixture; no provider certification claimed. |
| K03 — Editorial fit | Uneven income, invoicing routines, cash-flow organization, and records to discuss with an accountant, intended for US independent service professionals. Bookkeeping/invoicing tools may be reviewed; investment products and high-cost lending are excluded. Positioning is not verified audience identity. |
| K04 — Format | Weekly English audio, Mondays at 12:00 UTC, 25–35 minutes, 26 standard episodes. Public RSS distribution; no video placement in this offer. |
| K05 — Episode baseline | 750 average D30 downloads across F-E14–F-E21, the latest eight mature standard episodes. Median 750; range 680–820; total 6,000. Five incomplete August windows excluded. Not unique people, ad delivery, or a forecast. |
| K06 — Catalog context | 6,400 downloads in August 2026; 21,800 since launch through the cutoff. Separate context scopes, not additive audience or purchased catalog inventory. |
| K07 — Geography | August RSS downloads: US 4,480 (70%), Canada 640, UK 512, Australia 256, other known 384, unknown 128; total 6,400. Provider attribution, not residence or targeting. |
| K08 — Survey | Voluntary English web survey, August 1–21, recruited in show notes and newsletter; 48 usable unweighted submissions; no verified eligibility or invitation denominator. Report every question with its own n and missing count. |
| K09 — Platform evidence | Optional Apple-only appendix for F-E21: 160 Listeners, 112 Engaged Listeners, 63% Average Consumption for July 27–August 26, 2026 inclusive, UTC calendar dates. Not the same unit/window as RSS D30; Spotify and ad-play data are not available. |
| K10 — Other channels | 420 active newsletter subscription records on August 31. Opens, clicks, and cross-channel deduplication are not available. Newsletter inventory is not included. |
| K11 — Availability | One paid host-read mid-roll per episode. October 5, 19, and 26 open; October 12 on hold; November 2 available as an alternative. All dates require written confirmation. |
| K12 — Creative process | Sponsor supplies product facts, substantiation, CTA, URL, qualifications, and approval contact. One factual review is included in this example; host retains editorial control and discloses sponsorship. The illustrative 10-business-day brief deadline is not an industry rule. |
| K13 — Included and excluded | Included: three named baked-in new-episode mid-rolls and three show-notes links. Excluded unless separately agreed: dynamic insertion, US-only delivery, catalog, newsletter/social/video, external reuse, sponsor editorial control, and show-wide exclusivity. No “forever” hosting promise. |
| K14 — Reporting | Placement proof and provider-labeled D7/D30 episode downloads, delivered in this example within three business days after each window closes. No invented clicks, conversions, unique reach, ad listens, or ROI. |
| K15 — Proof | No paid-sponsorship outcomes, advertiser testimonials, or campaign ROI are available. Historical downloads do not guarantee future delivery or sponsor demand. |
| K16 — Next step | Placeholder only: partners@example.com and https://example.com/fictional-podcast. A real inquiry must identify product, customer, markets, objective, dates, format, evidence requirements, reporting, exclusivity, and reuse. Pricing is outside this kit example. |
| K17 — Release gate | Replace every identity, number, date, link, survey result, and availability record. Confirm insertion technology, ad load, reporting, approvals, rights, and category limits. Unsupported fields stay not available, or the offer changes. |
Five sponsor requests the kit should qualify
A media kit is useful when it changes the answer—not merely when it produces a nicer PDF.
| Sponsor request | Decision | Example response |
|---|---|---|
| US invoicing product; three October audio placements | Proceed to fit review, without a buyer-audience guarantee. | “The editorial focus is relevant. We can discuss the October 5, 19, and 26 placements with episode-download reporting. We cannot promise every listener is a software buyer. Please send the product claims, eligible markets, and measurement requirements.” |
| Four consecutive Monday placements in October | Rescope. One date is held. | “Only three October dates are open. October 12 is on hold, so four consecutive October placements cannot be confirmed. The available alternative is October 5, 19, and 26 plus November 2, subject to written confirmation.” |
| At least 5,000 unique US listeners and US-only delivery | Decline the stated guarantee and targeting requirement. | “We cannot supply verified 5,000-person US reach or US-only delivery with this inventory. We can offer the named untargeted audio placements and their download reports only. That is a different scope, not an equivalent substitute.” |
| Verified CFOs controlling enterprise software purchases | Decline the audience claim. | “We do not have evidence of verified CFO reach or enterprise purchasing authority. The show should not be selected on that basis from this kit.” |
| Proven prior sponsorship ROI plus paid-social rights to the host’s voice and likeness | Disclose missing proof and exclude the rights. | “There are no prior paid-sponsorship outcome data in this example. External paid-ad use of the host’s voice or likeness is not included. Neither can be represented as part of the current offer.” |
Even 4 × 750 = 3,000 would be only a multiplication of a historical download baseline. It would not answer a request for 5,000 unique US people, guarantee future delivery, or prove who heard an ad.
Download the completed fictional kit, blank template, episode worksheet, and five sponsor decision records.
Copyable blank podcast media kit template
Use not available for an unsupported number. Use zero only when the measured result is genuinely zero. Capabilities such as dynamic insertion, geo-targeting, and external content rights require confirmation; they cannot be inherited from a template.
[SHOW NAME] — SPONSOR MEDIA KIT
K01 Identity: [show] / [host] / [verified credentials or none]
K02 Version: [kit date] | Analytics cutoff: [timestamp + timezone] |
Inventory as of: [date] | Source: [provider, metric label, method link]
K03 Editorial fit: [subject] for [intended audience need].
Evidence-backed audience description: [claim + source or not available].
Considered/excluded categories: [policy].
K04 Format: [language, audio/video, cadence, length, live distribution links].
K05 Episode baseline: [value or not available] [unit] in [window].
Cohort: [episode IDs, n, release range]; cutoff: [timestamp];
exclusions: [rule]; mean/median/range: [ ]; worksheet: [location].
K06 Catalog: [calendar-period value + dates] / [lifetime value + dates].
K07 Geography: [unit, period, source, counts, denominator, unknown].
Targeting capability, separately confirmed: [yes/no + scope].
K08 Survey: [method, eligibility, dates, recruitment, usable records,
weighting, limits]. For each result: [exact question/options, counts,
valid n, missing, multi-select status].
K09 Platform evidence: [platform, episode/scope, dates, exact metric,
value or not available, platform definition].
K10 Other channels: [channel, date, unit/count, included?, overlap known?].
K11 Inventory: [release, placement, duration, max slots, open/hold/booked,
hold expiry, written confirmation needed].
K12 Creative: [read type, sponsor inputs, factual review, editorial control,
disclosure, brief and approval deadlines].
K13 Scope and rights: [included, excluded, insertion, term/removal,
exclusivity boundary, external voice/likeness/content rights].
K14 Reporting: [placement proof, metric, window, episode, source, timezone,
window end, report due, outcomes not measurable].
K15 Proof: [verifiable campaign/content evidence, what it proves, period,
permission] or [not available].
K16 Contact: [live email and show URL]. Inquiry must state [product, customer,
markets, objective, dates, format, evidence, reporting, rights].
K17 Release check: [data owner, last checked, refresh triggers,
unverified fields, approval to send].
A small show’s advantage is not that a template can transform 750 downloads into 5,000 buyers. It is that a specific editorial environment, a reproducible baseline, and tightly bounded inventory can let the right sponsor make a faster, more honest fit decision. The kit earns trust by making both the evidence and the “no” visible.
Sources
- Podcast Measurement Technical Guidelines v2.2, IAB Tech Lab, final release May 2024.
- Podcast Technical Measurement Guidelines v2.3 status page, IAB Tech Lab, last updated July 21, 2026; public-comment document linked when checked September 8, 2026.
- Analytics and Data requirements and troubleshooting, Apple Podcasts for Creators, checked September 8, 2026.
- Disclosure Standards, American Association for Public Opinion Research; standards page approved April 2021, checked September 8, 2026.
- FTC’s Endorsement Guides: What People Are Asking, Federal Trade Commission; references the Guides revised in 2023, checked September 8, 2026.
- Change Happened, Then What? Media Kit, show publisher, publication date not stated.
- Footnoting History Media Kit — June 2024, Footnoting History.
- The Art of Online Business Podcast 2025 Media Kit, The Art of Online Business; download figures dated December 19, 2024.
- Trail Correspondents Season 4 Media Kit, The Trek; statistics dated February 20, 2023; public copy hosted by AvantLink.
- Sponsor AONE, The Art of Network Engineering, live page checked September 8, 2026.
