Acquired's Ben Gilbert and David Rosenthal on 10 years of business storytelling, going to 8 episodes a year, and the Google series

Summary

Ben Gilbert and David Rosenthal, the hosts of Acquired, are ten years into what started as a side project recorded after hours at Madrona's Seattle office. The first episode on Pixar was researched in roughly an hour — both hosts scrolling the internet separately, then hitting record. By the end of year one, they had about 400 listeners.

The show is cutting its episode count from 12 this year to 8 in 2026, with Gilbert and Rosenthal betting the episodes will be better for it. The move is a deliberate rejection of the interview-podcast-as-news model. As Gilbert puts it, nobody has ever wanted to know what a cable news panel said four days ago — so building a show around timely guest interviews was always the wrong template.

Research at scale

For their recently completed Google series, they spoke to more than 30 people across hour-long background research calls — not recorded, not for the show, treated like book research. The opening question on every call is the same: what does traditional press most commonly get wrong about your company? The instinct is to correct the record before layering on independent research.

Books about companies, in their view, tend to get the core story right and then over-index on whatever was controversial at the time of writing — Cambridge Analytica getting 80 pages in a Facebook history being the obvious example. The framing that sticks at the moment of publication rarely survives as the canonical story.

The Google series

The central argument of the Google episodes is that the history of Google is the history of AI. Nearly every significant leader in foundation model companies traces a lineage back to Google, most of them there around 2015–2016 — Ilya Sutskever, Dario Amodei, Geoffrey Hinton, Sebastian Thrun, Andrew Ng, Andrej Karpathy, Jeff Dean. Yann LeCun at Meta is the notable exception.

Google trained the people who built its biggest competitive threats, then published the transformer paper that made it possible. The irony is that the antitrust judge in Google's search case cited the intensity of AI competition — much of it from former Googlers — as a reason not to break the company up.

The other thread running through the series is Microsoft. Google's entire product expansion — Gmail, Docs, Maps, Chrome, Android — was driven by the fear that Microsoft could destabilize its search business at any moment. Internet Explorer held roughly 70% browser market share on Windows' ~90% OS market share, making Google a tenant on Microsoft's property for its most profitable years. When OpenAI, after Elon Musk pulled his funding, took Microsoft as its capital partner, Google's read was that the company it had spent two decades escaping had come back through a different door. At the moment ChatGPT launched, Microsoft held 49% of OpenAI.

What makes an episode work

Gilbert describes three ingredients for Acquired's best episodes: a hero protagonist with a genuine arc, a secret hiding in plain sight (Costco's low SKU count and the inventory turnover dynamics it unlocks being the example he reaches for), and a third element he admits he blanked on mid-conversation.

The luxury company episodes — LVMH and others — were originally conceived as a way to teach a tech audience something it didn't know: that brand is a real and durable competitive asset. The framing worked because the hosts were genuinely learning during the research and sharing that discovery in real time.

The production model

Acquired sits at one extreme of the podcast format barbell. The show they appeared on does roughly 1,000 interviews a year, going live and publishing within an hour or two of recording, making one edit per episode (cutting the five-minute pre-show countdown). Acquired does the opposite — months of research per episode, no live format, and a research process where neither host knows what the other is going to say going into the recording.

The scale advantage that has compounded over ten years is that people who would headline most podcasts now take background research calls with Gilbert and Rosenthal because the show's reputation makes it worthwhile. That wasn't true when they started.