The Phantom Podcast: Questions About a Show Nobody Seems to Listen To
At Genuina Media, we buy programmatic podcast advertising on behalf of campaigns targeting Spanish-language audiences. Earlier this year, something caught our attention.
A Mexico City local news podcast kept appearing in our programmatic inventory at a scale that made no sense. In May 2026, it ranked among the top 65 most-served shows in the US programmatic podcast market, delivering nearly 7 million ad impressions to American buyers, more than TED Talks Daily, roughly equal to The Glenn Beck Program, and within reach of The New York Times' flagship audio product, The Daily.
In April, it ranked #31 in the US, ahead of Lore, The Tim Dillon Show, The Rubin Report, The Ezra Klein Show, and Radiolab. That month, its US programmatic inventory was roughly double that of La Cotorrisa, one of the most-streamed Spanish-language podcasts in the world.
So we went looking for the audience.
The show has only 52 ratings on Spotify. It has 1,000 followers. On Apple Podcasts, four people have reviewed it.
Our Genuina Media colleague Jorge González Barragán had flagged this publicly about a year ago, comparing it to This American Life and noting that nearly 90% of its downloads appeared to come from a "Generic Android App." The industry did not respond. The show is still in the marketplace and advertisers keep buying from it.
We are not in a position to make definitive claims about what is happening here. What we can do is share the data, the patterns we found, and ask the industry to take a closer look.
What the Data Shows
Nearly all inventory comes from Android Browser
In podcast advertising, "player" refers to the technology through which a listener accesses a show. Legitimate audiences use Spotify, Apple Podcasts, Amazon Music, and dedicated podcast apps. These are the platforms where real listeners live.
Over three months of programmatic data from AdsWizz's AudioMax platform, here is how this show's inventory broke down by player:
- March 2026: 56.9 million of 57.4 million impressions came from Android Browser, 99.0%
- April 2026: 67.2 million of 67.7 million total impressions, 99.3%
- May 2026: 51.8 million of 52.2 million total impressions, 99.2%
Android Browser is not a podcast app. It is a mobile web browser, probably the hosting site page. For context, here is the same metric for other shows in the same dataset during May 2026:
- La Cotorrisa: 0% Android Browser (76% Spotify, 19% Apple Podcasts)
- Leyendas Legendarias: 0% Android Browser
- The show in question: 99.2% Android Browser
The pattern appears on a second show from the same network. In April, that show showed 94.6% Android Browser across 4.85 million impressions.
A quarter of global inventory has no identifiable origin
In May 2026, 13.1 million of the show's 52.4 million total impressions, roughly 25%, had no identifiable geographic origin. In April, 14.4 million of 67.8 million, about 21%, were similarly unresolvable.
Real listeners have internet service providers that resolve to geographic locations. A significant share of impressions with no resolvable geography is unusual and, in our view, worth scrutiny from the platforms carrying this inventory.
The top cities do not match the show's home market
This is a Mexico City local news podcast. Its publicly visible content covers municipal politics, infrastructure projects, and local government announcements. It is about as hyperlocal as content gets.
Yet when we broke down its April 2026 inventory by city worldwide, the top identified cities were Santiago (Chile), Buenos Aires (Argentina), Bogotá (Colombia), and Lima (Peru), in that order, with Mexico City absent from the top five entirely. The same four cities appear in the same order in May. Real audience geography tends to shift over time. These rankings did not move at all across two months.
The hourly delivery pattern has an unusually high floor
Aggregating hourly data across three months, the show's delivery does follow a curve that peaks during Mexico City afternoon hours and drops overnight, broadly consistent with a Mexican audience on the surface.
However, at its lowest point, around 1:00am Mexico City time, the show still generates over 10,000 impressions per hour. That floor is higher than the total peak hourly delivery of most legitimate Spanish-language podcasts. Additionally, its weekend delivery dropped only 14.5% compared to weekdays over a sample window we examined. Industry research suggests real podcast audiences typically drop 25 to 40% on weekends, when commute listening disappears. Especially in a news podcast.
The second show on the same network shows a different anomaly: its delivery peaks at around 5:00am Mexico City time. That is also an unusual peak hour for a sports podcast.
All programmatic inventory flows through Spreaker
Every impression this show generated across every country and every month in our dataset flowed exclusively through Spreaker, a podcast hosting and monetization platform owned by iHeartMedia that offers open self-serve monetization to any show.
The Questions This Raises
We want to be clear: we are not accusing any company of fraud. We do not have access to internal systems, and there may be explanations for these patterns we are not aware of.
What we are saying is that this data raises serious questions that the industry should be asking, and that as programmatic buyers, we are directly affected by the answers.
For Spreaker and iHeartMedia: The anomalous player data, specifically the near-total Android Browser dominance, has been visible in programmatic data for over a year. What mechanisms exist to detect and investigate unusual traffic patterns in shows monetizing through Spreaker? Has this inventory been reviewed?
For AdsWizz and programmatic marketplaces: How does inventory with these characteristics enter and remain in open programmatic exchanges available to buyers? What traffic quality standards apply, and how are they enforced?
For Podtrac: What explains a show appearing in top-ten US rankings with no corresponding Spotify or Apple Podcasts footprint? Are measurement methodologies designed to catch this kind of discrepancy?
For the industry broadly: As programmatic buying in Spanish-language audio grows, buyers are increasingly making automated purchasing decisions based on inventory signals they cannot independently verify. If inventory like this can circulate for over a year after being publicly flagged, how widespread might the problem be?
What This Costs Buyers
We raise these questions in part because the financial stakes are real. We are spending money on this inventory. So, very likely, are other buyers of Spanish-language programmatic audio who have no idea this is happening.
In April alone, the show generated 9.7 million impressions in the US market. At a conservative programmatic CPM of $5, well below what premium Spanish-language audio typically commands, that represents approximately $48,500 in potential ad spend in a single country in a single month, from a show whose audience signals are, to say the least, difficult to reconcile with that volume.
Across three months and multiple countries, the gap between inventory purchased and real human listeners reached is potentially significant. The programmatic buying process is automated by design. Brands and agencies purchasing Spanish-language audio at scale do not manually vet every show in their buys. A podcast that appears in a legitimate exchange, with plausible-looking geographic distribution, passes most automated filters. That is precisely what makes patterns like this so difficult to catch, and so costly when they go unaddressed.
A Note on Methodology
The inventory data in this piece comes from AdsWizz's AudioMax platform, which we access as a programmatic buyer. Inventory figures represent available ad impressions, meaning the number of times an ad slot was generated and made available in the marketplace, not necessarily impressions purchased or delivered to a paying advertiser.
The Android Browser player classification is a standard AdsWizz dimension reflecting the technology through which the ad request was generated. Comparative show data comes from the same platform and the same time periods.
Spanish-language audio is one of the fastest-growing segments in programmatic advertising. If inventory like this can circulate for over a year after being publicly flagged, the industry owes buyers a clearer answer. If you have data, context, or experience with patterns like this, we want to hear from you. Reach out at hello@genuinamedia.com
