The episode features Michael Lewis, author of Moneyball, Liars Poker, The Blind Side, and host of Against The Rules, joining Ben Gilbert and David Rosenthal to analyze Acquired's 10-year journey.

Recorded in the original Google garage where Larry Page and Sergey Brin had their first office, the conversation explores why Acquired succeeded when 99% of podcasts fail.

Lewis discovered Acquired in July 2024 through a prominent CEO's recommendation at Google Campus, immediately recognizing the show's ability to create the immersive environment he seeks with books.

The discussion covers Acquired's evolution from 40-minute acquisition analysis to four-hour company histories, examining the partnership dynamics, creative process, business model, and lessons learned from studying the world's greatest companies.

The Partnership Foundation and Early Chemistry

Ben and David met at a Passover Seder while working at a venture capital firm, bonding over shared interests in Apple rumors, strategy analysis, and Ben Thompson's writing

David felt like a fraud as a VC - Princeton French literature major who worked on Wall Street before venture capital, never having built products himself

Ben was a software engineer hired for incubation work who didn't understand the real job of venture capital, viewing David as a peer who could explain what partners were discussing

"We each have two spouses. We have our actual romantic spouses that we have families with, and then we have each other" - describing their partnership intensity

David and Ben shared a bank account before David's wife and David shared one, demonstrating the financial trust in their partnership from early days

The partnership has had only two or three real fights or tensions over 10 years, with no status difference or competitive dynamic between them

Scarcity as Strategy: The NFL Lesson

Acquired started making 26 episodes per year at 40-80 minutes each, following conventional podcasting wisdom about consistent output and building habits

"The product is scarce" - NFL's 162 baseball games versus limited football games taught them that scarcity creates event-driven engagement rather than passive consumption

Conventional podcasting advice demands constant presence, but Acquired now releases only 6-8 episodes annually, treating each as a major event

"Every minute is a churn opportunity. If we don't live up to the expectations that our listeners have, we will burn them and they will lose us forever" - David on their constant fear

The scarcity model allows all outputs (audience, revenue, importance) to scale independently of inputs, as Ben and David do essentially the same work regardless of audience size

The 2022 Reset: Burning Cigarettes on Arms

End of 2022 brought FTX collapse, rising interest rates, and podcast advertising market crash - Acquired's revenue dropped 40% overnight

Doug Leone told them about Sequoia's response to dot-com crash: "We looked at each other and you could burn cigarettes on our arms and we wouldn't flinch" - they refused to lose money for investors

Ben and David stopped doing "specials" (random interviews and undifferentiated topics) and focused exclusively on work only they could do

They cut most sponsors, keeping only favorite partners, and focused on building the most durable stories and brand associations

2023 became the year of LVMH, NFL, and Nvidia - episodes that established Acquired as a brand about durability and compounding over decades

"All that matters is the late years of compounding. In any given year, you look at a Mag 7 company and their profits from the last year are greater than the first 20 years combined" - realizing long-term reputation matters most

Research Process: From Internet to Phone Calls

Research always starts with canonical work - reading everything ever written about the subject, beginning with who has done great work previously

The spider web of research mentions Something Else, leading to old YouTube videos and obscure sources, but always starting with pre-existing canonical material

Phone calls to sources didn't begin until 2023, now comprising about 50% of research after getting arms around publicly available information

"What is most misunderstood and what incorrect stories are out there where we can set the record straight?" - their key question to sources

For Google, they conducted 40 research calls across both hosts, representing a dramatic increase from earlier episodes with zero primary source interviews

Microsoft series marked the first time reaching CEO-level access when they talked to Steve Ballmer, who said "I don't listen to podcasts but I talked to people I trust"

NVIDIA's Jensen Huang listened to their two-part series made entirely from public information, then reached out wanting to meet and do an interview

The Recording Process: Scripted Improvisation

Early episodes used a single shared Google Doc where both hosts dumped research, eliminating all surprise and disagreement by recording day

Now they prepare separately with different documents, trying to do separate research calls to preserve the ability to surprise each other during recording

David writes 10,000-20,000 word scripts in sentence form but doesn't read them verbatim - has three screens during recording to reference while maintaining conversational flow

Ben maintains a giant text edit document with mechanical points and story beats, focusing on explaining how things work rather than narrative flow

"Every episode now going into recording day feels like a high wire act because we haven't fully scripted it out" - the risk of not knowing if it will come together

They record 8-9 hours of raw audio with hundreds of retakes, constantly producing each other in real-time to maintain energy and clarity

Production meetings were added six months ago - one week before recording, they agree on episode structure but deliberately don't share actual content

Post-Production: The Stephen System

Stephen, their audio engineer and only contractor, receives 8-9 hours of raw recording with dozens or hundreds of retakes marked throughout

Stephen cuts the raw audio roughly in half to about 5 hours, creating what they call "release candidate one" in software terminology

Ben and David then make 500-800 additional cuts using Descript, watching and listening at different speeds (Ben at 1X, David at 2.5X)

"You have to get so sick of the material where you're just like cutting to the bone" - Ben on the necessity of listening at normal speed to feel the pacing

The editing process takes three days, going through two full cycles of cuts before final release, with Ben typically doing the first pass

"You can cut the beginning of almost everything. You're always throat clearing, winding up. You don't need it. You just go right to it" - Michael Lewis on editing discipline

Episodes are never truly "done" - they run out of time with deadlines, always wanting one more edit pass before release

Business Model: Sponsors as Partners

Acquired's B2B sponsor model focuses on companies doing significant multi-million dollar annual deals, aiming to deliver a couple incremental customers per sponsor

Many sponsors have been ROI positive from signing one large customer who heard about them on Acquired, often through events rather than just podcast mentions

They write custom reads for every sponsor in every episode, treating it as a mini two-minute business analysis rather than generic ad copy

"We don't work with agencies. If an agency reaches out, we write them a very nice note saying we don't work with agencies" - maintaining direct relationships only

Sponsor selection requires "Switzerland" positioning - avoiding companies where endorsement would mean taking sides in contentious industry debates or picking winners

Events with sponsors include customer dinners, speaking at annual conferences, sporting events with fireside chats, typically committing a couple days of time per sponsor

They added an investment fund in 2024, investing in five private company sponsors in the first year, aligning financial incentives with content partnerships

The fund operates as "podcast with venture capital firm" not "venture capital firm with podcast" - only investing in companies they'd want as sponsors anyway

Episode Selection: Passion Over Performance

Three criteria for great Acquired episodes: compelling hero's journey from obscurity to ubiquity, secret hiding in plain sight, and importance in the world

Nintendo underperformed by 20% versus benchmark despite being an incredible 100+ year story of a Yakuza-connected playing card company becoming gaming giant

They followed Nintendo with a part two, compounding the performance problem since "people really don't love part twos" and won't tune in without part one

Indian Premier League cricket also underperformed but was Michael Lewis's first Acquired episode, leading to his discovery of the show

"If one of us feels passionate about something, even if the episode is a relative dud, it's still worth doing because someone else latches onto that" - lesson from underperformers

Nintendo episode was specifically listened to by one Meta executive who sent it to the entire executive team, eventually leading to Mark Zuckerberg interview at Chase Center

"If you don't feel anything, it's a chance nobody's going to feel anything. But if you feel a lot, someone's going to feel something" - Michael Lewis on passion magnitude

They killed the Federal Reserve episode after significant research because they couldn't find the fun in it, applying their passion test

Bell Labs was killed despite fascinating material because there were too many different stories and characters to create coherent narrative structure

Costco Deep Dive: Low SKU Count Philosophy

After trying too hard on Nike with nine books of research, they deliberately played loose on Costco using just Saul Price's autobiography and CFO whiteboard session

Costco carries only 4,000 SKUs versus Walmart's 100,000-200,000, creating the tagline "you get what you get, you don't pitch a fit"

Low SKU count means any given product quickly becomes high volume, making Costco a meaningful seller to every vendor and aligning incentives

Merchandisers manage only seven vendors instead of hundreds, allowing deep expertise - chocolate buyers monitor cocoa commodity markets directly

Inventory turns every 27 days on average, with some SKUs turning 10 times per month, meaning vendors finance inventory on net-30 terms with three days grace

"There's no working capital in this business other than building more Costcos" - the cash conversion cycle creates self-funding expansion

Research shows selling three kinds of jam outsells 30 kinds because excessive choice paralyzes consumers - Costco leverages this with curated selection

The strategy is high-leverage: carrying very few things puts enormous onus on making exceptional choices, but creates Charlie Munger's favorite company

Seven Powers Applied to Acquired

Scale economies: 1.5 million subscribers allow amortizing enormous input costs across large audience, enabling unreasonable effort per episode that competitors can't match

Counterposition everywhere: not volume-driven like CPM-based podcasts, no shareholders to answer to, refuse to work with agencies, structurally different business model

Network economies (weak): water cooler effect where more people listening makes each listener's experience more valuable through shared conversation topics

No switching costs: "Another podcast is one click away. Listeners can switch. There's no cost of switching out of acquired into whatever might come along"

Branding: "People just trust Acquired more" - growing brand equity that compounds over time as reputation for quality builds

Cornered resource: Ben Gilbert and David Rosenthal themselves, plus potentially Stephen the editor whose name they won't fully divulge

Process power in spades: the multi-stage research, separate preparation, scripted improvisation, hundreds of retakes, and iterative editing creates unreplicable output

"Explaining a process is a lossy compression of the actual process. Language is a lossy compression of thought" - Ben on why process can't be copied

Creating Spectacle: From Habit to Event

Acquired abandoned the podcast industry goal of creating listener habits, instead focusing on making each episode "the current thing" for water cooler conversation

Climate Pledge Arena in Seattle was first arena show with one section, allowing them to claim experience when pitching Chase Center

Chase Center show with Mark Zuckerberg in September 2024 was first major spectacle, but guests weren't announced beforehand to test if audience came for Acquired itself

Radio City Music Hall in December 2024 with 6,000 people featured Jamie Dimon, Meredith Kopit Levien, and Barry Diller as surprise guests

"Did all those people show up because it said Mark's name? Now we know" - deliberately not announcing guests proved audience loyalty to Acquired brand

Only 0.4% of Acquired's audience attended Radio City, but the "heat and light created from the idea that you did that" generates disproportionate impact

NFL Innovation Summit at SF MOMI on Friday before Super Bowl 2025 will feature Roger Goodell and NFL partners, with Acquired MCing the event

Michael Lewis on Creative Process and Trust

Lewis's process: handwritten notes filter what's interesting enough for the page, then typed notes filter again, creating 500+ page files before framing the story

"When I don't know what the end is, I don't know what the beginning is. It's that simple" - Lewis on narrative structure necessity

Lewis was a year into Sam Bankman-Fried research without knowing what to do with it until FTX collapsed, then suddenly had the story

"The queen died and then the king died is not a story. The queen died and then the king died of heartbreak is" - defining what makes narrative versus chronology

Lewis's ideal subject: "When I'm really interested in it and nobody else is. If I'm at a dinner party and after 60 seconds I can see their eyes glaze over"

"I assume when I write a book that what goes into people is something different than came out of me. I have to construct it so there's a hole for the reader to go in"

Lewis tried to get Ex Officio underwear to sponsor his podcast by giving them a list of products he actually uses, but they never returned his calls

Warren Buffett tracked Lewis's Berkshire Hathaway share purchases and sales through the financial crisis, asking through a publicist why Lewis sold some shares

Why Acquired Works: The Compounding Answer

AirPods launched a year after Acquired started, making it societally acceptable to listen while moving through the world doing other tasks

Spotify entered podcasting in 2018 and now represents over half the market, bringing hundreds of millions of new listeners to the medium

Apple Podcasts not becoming YouTube was beneficial - it's a place where clicking subscribe creates durable, valuable listener accumulation without algorithmic intermediation

"Corporate America becomes ever more important. What is going on in the economy is mysterious to people" - Michael Lewis on why business storytelling matters now

"Acquired helps you understand why the world is arranged the way it is" - a promise they couldn't have made in 2015 but absolutely deliver on now

"The biggest reason Acquired works is our partnership. The magic exists between us. If just one of us were making Acquired, it would be nothing" - David

Path dependence was critical: they always had the right product for the current value it created, allowing sustainable growth without premature scaling

"I am terrified of betraying that trust. Anytime we make an episode, I think of it as a churn opportunity" - maintaining quality as existential imperative