Two stories this period are really the same story told from different angles: who gets to decide what a podcast number means, and whether the people running those numbers can be trusted to count against their own interest. Amazon just admitted its ad math was generous. October’s charts quietly showed where attention is actually moving. And one of the few independent voices who built his career on asking what audiences really do, rather than what platforms claim, stepped back from the company he founded.
Amazon Admits It’s Been Overcounting Its Own Ads
Amazon Ads announced it is changing how it counts a display ad impression. Right now, an impression counts the moment an ad starts to download to a shopper’s device. Starting January 1, 2027, it will only count once the ad actually appears on the screen. Amazon’s own warning to advertisers: expect impression counts to drop by an average of 63% under the new method.
Read that number again. Amazon is telling its advertisers that roughly two-thirds of what it has been billing or reporting as “impressions” never actually reached a screen. That’s not a rounding error — it’s an admission that the metric was structurally inflated, and Amazon is only fixing it now, with over a year’s notice before the change takes effect.
Podcasters have spent years arguing over a much smaller discrepancy — the gap between a download and an actual play, which runs closer to 13%. Amazon’s fix is a useful reality check on scale: if one of the biggest ad platforms in the world was overcounting by 63% on its own display inventory, the download-versus-play argument in podcasting looks almost modest by comparison.
What this means for your show: any ad deal priced on “impressions” or “views” deserves one follow-up question — impression of what, counted when? The definition is doing real work in the price.
This change is specific to Amazon’s display ad product, not to podcast downloads directly, and it doesn’t take effect until 2027. But the admission itself is the news: a major ad platform conceding that its own counting method was flattering its results.
Read the full story at Podnews.
October’s Charts: The Same Names on Top, One New Name Breaking In
Podscribe’s October 2026 rankers had the usual leaders — Crime Junkie reached the most people, The MeidasTouch Podcast logged the most plays — plus two notable shifts. iHeartMedia’s monthly plays jumped 63%, and two shows on the video platform Rumble made it into the top 10.
The iHeart number is a network story: a single company pulling a big jump in plays across its portfolio, which says more about consolidation and cross-promotion muscle than about any one show getting better. The Rumble shows are the more interesting data point for an independent operator — a video-first platform that isn’t a traditional podcast app is now producing top-10 audience numbers.
Online retailer Quince was named the single biggest podcast ad spender for the month, a reminder that direct-response advertisers, not legacy categories, are increasingly setting the pace on ad budgets.
None of this changes what a small show should do tomorrow — these are the biggest shows in the business, not a blueprint for a local or niche feed. But the direction is worth tracking: video-native platforms are no longer a side bet, and the advertisers writing the biggest checks are the ones selling directly to consumers.
Read the full story at Podnews.
Fred Jacobs Steps Back From the Company He Founded in 1983
Fred Jacobs has resigned as CEO of Jacobs Media, the research and consulting firm he founded in 1983, moving into a Chairman role with what he calls “a much more limited basis” of involvement. He disclosed a Parkinson’s diagnosis from 2023 as part of the decision. He’ll continue helping produce Techsurvey, the firm’s long-running annual research study of radio listener behavior and technology adoption, and plans to keep writing his blog — the same one he’s kept for 21 years — once a week.
Jacobs is also the person credited with starting Classic Rock as a radio format, launching the first station in 1985 to serve an age group mainstream radio was ignoring. It became one of the most successful formats in radio history by betting that an underserved, specific audience was worth building a whole product around — the same logic a lot of niche independent podcasts run on today, whether they’d name it that way or not.
No successor CEO was named in the announcement, and “limited basis” leaves real ambiguity about how present Jacobs will actually be in the firm’s day-to-day research. Techsurvey’s credibility has rested heavily on his name for decades; whether that holds without him running the company is an open question, not a settled one.
Read the full story at RAIN News.
None of this is dramatic on its own. But put together, it’s a period where the industry’s measurement layer got a little more honest in one place and a little less certain in another. If you sell ads on your show, both lessons point the same direction: ask harder questions about what a number actually counted, and don’t assume the people who’ve always counted it will keep doing so the same way next year.