Podcast monetization: do product buyers also pay for subscriptions or merch?

The main result is 58.2%: the weighted share of podcast-prompted purchasers who explicitly answered No to both ever paying for a podcast subscription and ever buying podcast merchandise. Another 41.4% reported at least one of those two experiences. The remaining 0.4% had unanswered information that prevented classification.^[1][2]

Every percentage in this comparison starts with the same 763 respondents. They were part of the 2,530 respondents who had listened to a podcast in the previous 12 months. We then selected those who said they had bought something because it was promoted or talked about on a podcast. The comparison does not start with all listeners, all customers or all people who encountered an advertisement.

We use “podcast-prompted purchasers” as a short name for that group. The question covers discussion as well as promotion, so it does not identify an advertising-only audience. It also does not identify the item or seller: the purchased item could itself have been podcast merchandise. Nor does the survey establish which podcast prompted the purchase or whether the subscription or merchandise came from that same show.^[2]

For a creator, the result challenges a specific shortcut: treating any evidence of buying through podcasts as proof of demand for a membership or merchandise launch. It does not mean that the 58.2% would refuse a future offer. They reported no past purchase of these two kinds; the survey did not test a new offer, price or message.

Splitting the same purchaser group more finely makes the distinction clearer. On a weighted basis, 12.9% had paid for a subscription but had never bought podcast merchandise; 17.4% had bought merchandise but had never paid for a subscription; and 11.1% reported both. The 58.2% reporting neither remains the largest category. Unresolved answers account for 0.4%.

The two “only” categories have a narrow meaning. Subscription only means Yes to the subscription question and No to the merchandise question. It does not mean the respondent had supported no other creator or spent money in no other way. Donations, live events and other possible revenue sources are outside this comparison.

The merchandise-only share is larger than the subscription-only share in these released records. That does not identify the better business to build. We do not know what offers respondents saw, what they paid, how often they bought or what it cost a creator to fulfill the purchase. Counting people with a history of buying is different from comparing revenue or profit.

Pew’s original report already showed separate percentages for these activities among podcast listeners. Our contribution is to examine the same people within the purchaser group. Separate totals cannot tell a creator how often the experiences belong to one person. The paired answers can show that overlap, while leaving the reasons for it open.^[3]

The practical implication is to keep the evidence for each offer separate. A listener’s story about buying a recommended product belongs in a product-recommendation brief. It can prompt a conversation about a membership or merchandise idea, but it should not be entered as a confirmed buyer of either.

For a subscription, write down what continues after the first payment. Is the offer a defined bonus series, access to a useful archive or another clearly described benefit? Then consider what the creator can deliver consistently and what a listener would need to know before committing. Those questions concern the offer’s value and delivery; this survey does not answer them.

For merchandise, make the item concrete. What would someone receive, which details need explaining, and what are the production and delivery commitments? A preference for a design is useful feedback, but it is not the same as a completed order. Distinguishing those stages helps a small team avoid treating enthusiasm as booked revenue.

The same discipline helps a business working with a podcast creator. Audience evidence should match the proposed activity. A history of buying a mentioned product does not establish that the audience pays for the show; membership experience does not establish that a new advertised product will sell. The brief can state the relevant evidence without converting it into a promise about campaign performance.

If the immediate task is choosing people for audience interviews, our related study on influencer marketing measurement examines how different survey questions select different groups.^[7]

Consider a fictional gardening podcast. Listeners sometimes tell the host that they bought tools discussed on the show. The team is considering a paid seasonal audio series and a branded apron. The following ledger is a planning example, not an observed business or a result from the survey.

Give each offer its own entry with the same five fields: the offer, the evidence already held, the question still open, the smallest useful next step, and the observation that would justify a further commitment. The point is to make the next decision specific. No survey percentage needs to be multiplied by the show’s audience size.

Product recommendation. Evidence held: messages from listeners describing purchases of discussed tools. Question open: whether those messages reflect broader demand or a few memorable experiences.

Next step: keep the examples attached to the specific tools and episodes they concern.

Further commitment: evaluate a proposed partnership against its own audience fit and agreed business measures. The messages alone do not establish attributable sales for a future partner.

Paid seasonal audio series. Evidence held: no purchase evidence for this proposed series yet. Question open: which part of the series listeners value enough to pay for, and whether the team can deliver it.

Next step: present a clear outline and ask listeners what they would use.

Further commitment: consider a clearly described small paid offer only after checking the content promise and delivery workload. Interest expressed in a conversation should remain labeled as interest.

Branded apron. Evidence held: a design concept, but no orders. Question open: whether listeners want this particular item at a feasible price.

Next step: review the design and practical requirements, including sizing where relevant, production and delivery.

Further commitment: make an inventory decision from evidence specific to the item and the business’s capacity, rather than from the historical survey’s merchandise percentage.

At the next planning meeting, the team can read the entries side by side. The product example contains reported buying behavior; the two proposed offers still contain questions. That difference makes it easier to decide which task comes next without pretending the research has chosen a winner.

Keep the ledger short enough to use. Update the evidence when something changes and keep proposed actions distinct from completed purchases. Its purpose is to make the next commitment easier to assess, not to produce a revenue forecast from the survey.

Four of the 763 purchasers could not be classified as either having supported a podcast in one of the two measured ways or having done neither. They represent 0.4% of the purchaser group after weighting. We kept them in the denominator instead of silently dropping them.

If every unresolved case belonged in the support group, the weighted share with at least one support history would rise from 41.4% to 41.8%. If every unresolved case belonged in the neither group, its share would rise from 58.2% to 58.6%. Those ranges describe the possible effect of these unanswered items only. They do not measure all uncertainty in the survey.

Even the lower end for neither is above half. That is why we can describe a majority of the weighted purchaser group as reporting neither experience without deciding how the unanswered cases should be assigned. We cannot use the small missing-answer range as evidence that the result is precise for every podcast audience.

These are not confidence intervals. The evidence describes released, weighted US survey responses from December 2022. It does not establish today’s willingness to pay, and translating the article does not turn the result into a local-market estimate.

We analyzed Pew Research Center’s American Trends Panel, Wave 118, fielded December 5–11, 2022. The release contains 5,132 US adult respondents. Of these, 2,530 reported listening to a podcast in the previous 12 months and were eligible for the action questions. The present analysis uses the 763 who reported buying something because it was promoted or discussed on a podcast.^[1][2]

The subscription and merchandise questions ask about ever having done each activity. The product-purchase question does not give a separate action window. The previous-12-month condition establishes who received the questions; it should not be applied to all their purchase histories. These answers do not identify current subscriptions or the sequence in which purchases happened.^[2]

We used the final released survey weight for every percentage. Within the 763-person purchaser group, we summed the weights for a category, divided by the total weight of the whole group and multiplied by 100. Unweighted respondent counts describe the number of answers; they are not the numerator for the displayed weighted percentages.

A Yes to either support question establishes at least one of the two experiences, even if the other answer is unanswered. Neither requires an explicit No to both. All other combinations remain unresolved. The five-category table separately distinguishes subscription only, merchandise only, both, neither and incomplete answers. No unanswered response is treated as No.

Before calculating the new overlap, we checked the eligible counts and reproduced the published answer percentages within their reported rounding intervals. Calculations used the stored weights before display rounding. The article and downloads show percentages to one decimal; small differences from sums of rounded components should not be interpreted as extra respondents or categories.^[4]

Pew’s methods describe probability sampling and adjustments for selection and nonresponse. The material available for this analysis did not establish a reproducible survey-design variance method for these new conditional shares. We therefore report descriptive weighted shares and missing-answer ranges, with no confidence intervals, significance tests or claim that one monetization model outperforms another.^[5]

The release documentation also explains that some rare values may be swapped to protect respondents. The analysis concerns those released records. Self-reported histories, unanswered questions and privacy protection all matter when interpreting the result; none can be resolved by treating a weighted percentage as a sales prediction.^[6]

Source attribution: Pew Research Center’s American Trends Panel. The opinions expressed herein, including any implications for policy, are those of the author and not of Pew Research Center.