Podcast sponsor data and ad reads
AdReads indexes public podcast ad reads so teams can search sponsor mentions by brand, promo code, placement, offer, URL, show, and transcript language. It is designed for media buyers, operators, and competitive research teams who need the exact read, not a dashboard full of guesswork.
If you're looking for podcast summaries, audio clips, and general podcast data, check out our other site Quicklets.ai.
What the library tracks
Ad-read transcripts
Clean snippets, raw source context, offer language, codes, URLs, and timestamps.
Sponsor footprint
Brand profiles with categories, recent reads, show coverage, and detected landing pages.
Podcast context
Show pages with sponsor mix, Apple rank context, and episode-level reads.
Review workflow
Internal queues for corrections, visibility changes, takedown requests, and import-ready records.
Estimated placement value methodology
V1AdReads estimates what a comparable single-episode placement on the show might cost today. The model is a directional planning proxy—not reported advertiser spend, a historical invoice, verified audience delivery, or campaign total.
1 · Audience range
A stored show-popularity rank—not measured downloads or verified audience—is converted into broad AdReads bands: ranks 1–10 use 100K–5M; 11–50 use 50K–1M; 51–200 use 15K–300K; and 201–2,000 use 3K–75K.
2 · CPM range
Host reads use $18–$30 CPM below 45 seconds and $24–$40 from 45–120 seconds. Announcer, mixed, and unknown reads use $12–$22.
3 · Placement factor
Mid-roll is 1×, pre-roll and unknown placement are 0.9×, and post-roll is 0.5×. Each low, midpoint, and high input is multiplied independently.
No estimate appears when the popularity input or duration is missing, the popularity rank is outside the modeled range, or a read runs longer than two minutes. Historical reads show a current equivalent value.
Sponsor-directory spend and podcast-directory gross-earnings estimates add qualifying reads from episodes published during the displayed trailing 30 UTC calendar dates. Reads without an episode publication date are excluded from that window. Detail pages lead with the same 30-day estimate. Daily and weekly reports instead group reads detected in the report’s UTC window to rank sponsors by estimated spend and podcasts by estimated gross earnings. Those lists are two non-additive views of the same modeled placements and use current popularity inputs without extrapolating missing reads. Episode publication or detection is not dynamic-ad delivery or billing timing, and no estimate is an invoice, complete campaign budget, booked-period revenue, total market spend, or verified creator payout.
Public benchmark inputs are informed by current publisher and marketplace guidance from Libsyn Ads and Acast. The popularity rank is only a proxy: Apple explicitly notes that podcast charts combine multiple engagement signals and are not a measure of the largest audience in absolute terms. See Apple's chart methodology.
Data scope and freshness
The public library includes quality-qualified podcast ad reads with usable sponsor, podcast, episode, and source evidence. Hidden records, records below publication thresholds, placeholder sponsors, and excluded source families are not counted on public discovery pages.
The latest included detection is Aug 13. Counts describe activity detected in the AdReads dataset. They are not estimates of advertiser spend, audience reach, total campaign volume, or market share, and they do not represent a complete advertiser roster for every podcast.
For search and AI agents
AdReads publishes a bounded preview API and machine-readable documentation for finding source-backed podcast ad reads. Results link back to the canonical transcript, audio, sponsor, and podcast pages.