Podmerse's Anuj Agarwal: A missing podcast sponsor isn't proof the ad failed
A podcast sponsor disappearing is not proof its ads failed. Podmerse co-founder Anuj Agarwal explains what repeat bookings show, and what they cannot.
An advertiser disappears from a podcast's sponsor list. It is tempting to assume the campaign didn't work. But the brand might simply have changed its media plan, leaving the podcast with a misleading signal.
That distinction matters as Podmerse, a database of podcast sponsorship activity, makes advertising placements easier to trace. Its records show where brands ran ads, not how many customers the ads brought in. For agencies building podcast media plans, those are very different questions.
Table of contents
- A database that tracks the buyers, not the results
- Two advertisers left shows for very different reasons
- The small investing show that kept drawing mining sponsors
- Why copying a competitor's podcast list can mislead
- What the database cannot tell a media buyer
- How marketers should use the trail
A database that tracks the buyers, not the results
Launched publicly on October 28, 2026, the platform lets podcasters look up brands sponsoring similar shows and advertisers inspect where other companies are buying podcast ads. Its database is updated daily and includes more than 2 million sponsored episodes, nearly 44,000 podcasts with recorded sponsorships, and more than 368,000 brands, according to its launch announcement.
The distinction between monitored and recorded is important. Podmerse says it monitors new episodes from more than 2 million podcasts, but it has detected sponsorships on only a subset of those shows. Its search tools let buyers filter sponsorship activity by category, industry and how recently an advertiser appeared.
The launch also came with a striking number. Of 125,429 brand-and-show pairings observed between the first and second quarters of 2026, only 34% appeared on the same podcast in both quarters. In calculating that figure, the company excluded shows that had stopped running ads. It would be easy to read the other 66% as failed relationships. That would be an overreach.
Two advertisers left shows for very different reasons
ContentGrip asked Anuj Agarwal, Podmerse's co-founder, what a missing sponsor can actually reveal. His answer started with a caution:
A brand not appearing again on the same show is not, on its own, evidence that the campaign underperformed.
Anuj gave two anonymized examples from the company's records. In the first, a VPN advertiser sponsored 234 shows in Q1 2026 and appeared again on 171 in Q2. It left 63 shows but added 63 others. The abandoned shows continued to carry ads from other companies. To him, that pattern looks more like changing the mix than abandoning podcast advertising.
The second pattern was a broad pullback. A meal-kit advertiser sponsored 257 shows in Q1 and returned to 139 in Q2, while adding only 14 new shows. Its presence across the medium also fell over time: 236 shows in February, 102 in June and 72 in September. The shows it left continued to attract other sponsors. A single podcaster losing this buyer could mistake a company-wide budget shift for a judgment about its own audience.
Across all pairings that did not repeat in Q2, Anuj said 31% involved brands still advertising on other shows in the company's records. For the remaining 69%, Podmerse recorded no podcast ads anywhere in its coverage during that quarter. Neither figure reveals whether individual ads converted customers.
The small investing show that kept drawing mining sponsors
Another example cuts against the instinct to buy the largest audience first. Anuj described a daily investing podcast that interviews analysts and small-cap company leaders. A modelled estimate from MillionPodcasts puts its audience at only a few thousand listeners a month, not a verified download count.
Even so, 110 junior mining companies sponsored the show in Q1 2026, and 58 appeared again in Q2. That is roughly a 53% same-show repeat rate, compared with 34% across all the tracked pairings. Several companies had appeared in Podmerse's records since spring 2025 and were still buying placements between August and October 2026.
For a company trying to reach investors in smaller public businesses, such a concentrated audience might be more useful than a much bigger general-interest podcast. Repeat buying does not establish return on ad spend, but it points buyers toward niche shows that a media kit dominated by audience size could overlook.
Why copying a competitor's podcast list can mislead
Imagine an agency discovers that a competitor sponsored 15 podcasts last quarter. Anuj argues that booking the same 15 would be a poor starting point: some placements may have been one-off tests, and the competitor's intended buyers may be different.
The company's data also varies sharply by industry. Among the classifications brands list on LinkedIn, Technology, Information and Internet advertisers returned to the same show 51% of the time, compared with 46% for Software Development, 27% for Marketing Services, 26% for Professional Training and Coaching, and 25% for Business Consulting and Services. The overall figure was 34%.
Anuj's advice is to start with recurring sponsors in the right category, then inspect how recently they appeared and whether competing brands have crowded the same show. A podcast can be popular with competitors and still be a poor place for another near-identical ad.
What the database cannot tell a media buyer
Podmerse processes the show notes of active podcasts and converts audio to text for popular shows. It uses language models to help classify sponsors, remove noise and consolidate variants of the same name. For instance, "Athletic Brewing Co." and "Athletic Brewing" are grouped under one brand.
There are limits. The records do not label an ad as host-read or dynamically inserted, and smaller shows are primarily detected through show notes rather than full audio transcription. The company acknowledges occasional false positives in its automated detection.
Crucially, it does not observe advertisers' actual campaign results. Those require first-party measurements such as promo-code redemptions, campaign-specific URLs, post-purchase surveys and attribution reports. Sponsorship histories can help identify an advertiser's movement across shows or its return in a later quarter; they cannot prove a campaign paid off.
The product is also new. Asked for a concrete example of buyer feedback changing its search interface, Anuj said the team had only launched on October 28 and expected to learn more as podcasters and media planners began using it. He did not claim to have an established customer-feedback case yet.
How marketers should use the trail
For an agency, sponsorship history works best as a shortlisting signal, followed by a proper audience-fit assessment and a measurement plan agreed with the advertiser. Before ruling out a podcast because one sponsor vanished, check whether that advertiser continued buying other shows, whether its spending contracted across the category, and whether it returned later. A seller's media kit and an advertiser's own outcome data still matter.
Anuj's personal answer to ContentGrip's closing question also reflected his appetite for solving hard-to-see problems. Outside this launch, he keeps coming back to building products, from FeedSpot to MillionPodcasts and now Podmerse, because he enjoys making difficult information easier for people to find.
