Acquired — Coca-Cola: The Complete History & Strategy

Why this is in the vault

Coca-Cola is the cleanest case of brand power as a moat, and arguably the longest-running case of “the system, not the company” as the unit of analysis. A 4-hour treatment of a sugar-water company that is now worth ~$300B because of how the system of bottlers, retailers, restaurants, fountain operators, billboard owners, and marketing partners is incentivized to sell more Coke. In the vault for three reasons:

It is the canonical Charlie Munger thought experiment in operating form. Munger’s framework: starting in 1885 with $2M, what business do you build to reach $2T? Constraints: non-alcoholic beverage, throws off billions in dividends, becomes globally distributed. The answer is essentially “build Coca-Cola from first principles” — global trademark, taste universal across cultures, calorie-rich and habit-forming (sugar + caffeine), distribution outsourced to capital-light franchisee bottlers, marketing locked to “happiness” Pavlovian associations. The episode treats the Munger frame as the design spec and walks through how Candler/Woodruff/Goizueta executed it.
It documents “system, not company” as a strategic primitive. David’s quintessence: Coca-Cola is not a single company; it is an incentive-aligned partnership of bottlers, retailers, restaurants, billboard owners, fountain operators — every one of whom makes good money selling Coke. Robert Woodruff’s mantra: “everyone who has anything to do with Coca-Cola should make money.” This is the most under-theorized form of moat in startup discourse — most modern companies want to capture all the value, while Coca-Cola deliberately gave huge slabs of value to partners to ensure the system was self-reinforcing.
The New Coke disaster (1985) is the most expensive empirical proof in business history that brand emotional attachment is real and unmeasurable by taste preference data. 200,000 blind taste tests said New Coke won. They never asked “how would you feel if New Coke replaced original Coke?” because Goizueta later admitted “you can’t ask emotional questions like that — people don’t know.” The disaster cost ~3 months and a near-mutiny. The reversal generated more brand goodwill than any campaign in history. The lesson is not “don’t change products” — it’s “you cannot get real data on emotional questions from instrument-based research.”

Core argument

The Munger thought experiment is the right design lens for the entire company. The episode opens with Munger’s frame (“$2M in 1885 → $2T”) and uses it as the spec. The conclusion: Coca-Cola is what you would design from scratch if you were trying to build a 1-million-x return constrained to non-alcoholic beverages, with capital-light distribution, global brand, dividend yield, and a moat that survives 140 years. Most of business history is post-hoc rationalization; Coca-Cola is unusually well-fit to the Munger framework because Munger built the framework partly from observing Coca-Cola.
Three CEO eras define the company. Asa Candler (1888-1923): invented the franchise-bottler model after a near-accident — sold the bottling rights for $1 to two lawyers in Chattanooga who couldn’t afford to back out. The contract effectively gave bottlers perpetual, exclusive territory rights at fixed syrup prices. This was the original “system, not company” architecture, accidentally created. Robert Woodruff (1923-1981): institutionalized the “everyone who touches Coca-Cola should make money” doctrine, drove global expansion (especially via WW2 — every U.S. soldier could buy a Coke for 5 cents anywhere in the world, government-subsidized via wartime sugar exemptions), and built the “always delicious, always refreshing” brand discipline through 50+ years of repetition. Roberto Goizueta (1980-1997): restructured the bottler economics (the originally locked-in syrup prices were destroying Coca-Cola’s margins by the 1980s), launched Diet Coke (the first product-line extension in the company’s history, debated for years), launched and reversed New Coke, and grew market cap from ~$4B to ~$150B.
The bottler structure is the moat — and almost killed the company. Candler’s original franchise contract gave bottlers fixed-price syrup forever. By the 1970s this was a death spiral — Coca-Cola made nothing on syrup sales while bottlers captured all the marginal economics. Goizueta’s first major project was unwinding the perpetual contracts (took years and required threats of bankruptcy at some bottlers). The lesson: a system-not-company moat can ossify into a system-not-company anchor; the structure that creates the moat must be renegotiated periodically.
Pepsi’s only real punch landed in 1934 via counter-positioning. Pepsi was bankrupt 2-3 times before this. In 1934, they started selling 12oz Pepsi in recycled beer bottles for the same nickel price as a 6oz Coke. Twice the cola for the same price. Coca-Cola couldn’t respond — they and their bottlers had invested all their IP and capital in the contour 6.5oz bottle. Pure counter-positioning: incumbent’s prior investment is the constraint. The hosts call this “textbook counter-positioning.” (Pepsi’s eventual 70+ year challenge to Coke is real, but every era of Pepsi advance has been similar — find a structural choice Coke can’t match without abandoning prior investment.)
The Pepsi Challenge (1975, John Sculley) was the second big punch. Blind taste tests showed Pepsi winning. Sculley turned this into national TV advertising. Coke’s response: New Coke (1985) — a sweeter formulation reverse-engineered from the Pepsi Challenge data. They tested taste; they didn’t test emotional attachment. The 200,000-person research was the most extensive consumer research ever conducted at the time and it missed the entire question that mattered. Goizueta later: “you can’t get real data on emotional questions like that.” This is the canonical example of “what gets measured is not what matters” in consumer products.
The reversal of New Coke (78 days later) generated more brand goodwill than any campaign in history. Letters described Coca-Cola in the language of betrayal, lost love, and stolen childhood. One letter: “I tasted betrayal on your lips.” A woman in Marietta, Georgia attacked a Coke delivery man with an umbrella for stocking New Coke. The reversal — explicitly framed as “we listened to you, the consumer, and we’re bringing back Classic Coke” — moved hundreds of millions of dollars of free PR through every news outlet for months. The episode’s wry line: “the most expensive A/B test in history.”
The “secret formula” is a brand asset, not an actual moat. The hosts have an extended argument here. Mark Pendergrast’s research found the original 1886 Pemberton formula in archives. He took it to Coca-Cola executives and asked what they thought. Their response: “Publish it. What’s anyone going to do with it? Make a drink? Distribute it how? Brand it as what? We’ll sue you for using the name. How are you going to invest in marketing? How are you going to get the scale economies?” The formula matters because the meaning imbued in the formula is part of the brand power, not because the chemical composition is irreplicable. Pepsi has a better formula by blind taste — and it doesn’t matter.
Cornered resource: bottlers, not formula. David’s correction in the seven-powers analysis. The bottlers have great distribution, exclusive territory rights, and they’re not bottling for anyone else. Switching teams as a bottler is technically legal (probably) and operationally impossible. The 1980 federal antitrust exemption for soft drink bottlers (mentioned in trivia) makes the territory exclusivity legal. This is the actual durable moat — not the formula, not the brand alone, but the bottler-territory system.
Repetition works. 150 years of “always delicious, always refreshing.” Ben’s quintessence. Most luxury brands chase reinvention (Hermès whimsy, Pierre Alexis annual themes); Coca-Cola has chased the same core message for a century and a half with cosmetic variation. The hosts argue this is the right strategy when the underlying customer need is universal and unchanging — refreshment, sugar, caffeine. Repetition compounds when the message is true and the need is invariant.
Could Coca-Cola be run by a ham sandwich? (Buffett/Munger frame.) David: yes, in most periods of company history. Ben: no, in 1985 (15 years of Pepsi share loss demanded action) and in the 2000s (obesity crisis demanded category re-strategy). Both agree the system is the durable thing; CEO quality matters at inflection points and not much in between.

Mapping against RDCO

“System, not company” is the under-theorized moat in startup discourse and the right frame for any platform/marketplace/network play. RDCO writing about platform businesses, network effects, and ecosystem strategy should reach for this frame. The discipline: design the system so every participant makes money in proportion to system success, then guard against the system structure ossifying. Worth a vault concept doc cross-linking Coca-Cola, Costco (member-supplier-employee triangle), TSMC (foundry-fabless ecosystem), and modern platform examples (Stripe, Shopify, Substack).
Counter-positioning that lands: incumbent’s prior investment is the constraint. Pepsi’s 12oz bottle for the same nickel was counter-positioning textbook because Coca-Cola’s bottling capital was the moat that became the prison. This is the strongest test for whether a structural attack will land: not “will the incumbent want to respond” but “what specifically can they not afford to abandon to respond?” RDCO writing about incumbent disruption should always pose the question this sharply.
“You cannot get real data on emotional questions” is the right frame for any RDCO writing about user research, PMF measurement, brand decisions. New Coke is the canonical empirical proof. The 200,000-person taste test was correct on its own question and irrelevant to the actual decision. RDCO advisory work should always ask: what is the real question we’re testing? Most product research instruments measure properties, not preferences-against-substitution. Worth a vault concept doc on “the New Coke principle.”
“Repetition works” is a contrarian frame for content strategy. Most content advice is “find new angles, fresh hooks, novel framings.” Coca-Cola’s 150-year discipline of “always delicious, always refreshing” is the counter-example. When the underlying truth is durable, repetition compounds the brand. The Sanity Check editorial discipline should reach for this when there’s pressure to chase the next AI news cycle — the underlying RDCO message about anti-silver-bullets, vault-as-knowledge-graph, and harness-as-leverage is the durable truth that repetition will compound.
The “ham sandwich” frame for institutional resilience. Buffett and Munger argue Coca-Cola could be run by a ham sandwich because the system is so well-designed it doesn’t require active management. The opposite extreme is most early-stage businesses where the founder is the entire moat. RDCO advisory work should always ask: is this business closer to the ham-sandwich end (system-designed-for-resilience) or the founder-mode end (everything depends on one person’s daily judgment)? The transition path between the two is the hard part — and most businesses never make it.
The 1980 federal antitrust exemption is a useful reminder. Industry-specific carveouts in regulation are real moats. Most analysts ignore them because they look like background noise. Worth scanning the data and analytics regulatory landscape for similar carveouts (HIPAA exemptions, CCPA carveouts, sectoral data-use exemptions). Curiosity question candidate.
Goizueta unwinding the perpetual bottler contracts is a case study in moat-as-anchor. The structure that built the moat eventually became the constraint that nearly killed the company. Worth tagging in any RDCO advisory work where a client’s competitive structure has hardened into something that prevents adaptation. The advice: periodically renegotiate the moat structure, even when it’s painful and creates internal conflict.

Open follow-ups

“System, not company” concept page. Coca-Cola, Costco (employee-member-supplier triangle), Visa/Mastercard, Stripe, Shopify, Substack. Pattern: every participant in the value chain makes money in proportion to system success; capturing all value would kill the system. Cross-link to “structural counter-positioning” frame from Trader Joe’s.
“The New Coke principle” concept page. Document the 1985 disaster as the canonical example of right-instrument-wrong-question research. Generalize: when do you reach for taste-test instruments and when do you reach for emotional-attachment instruments? RDCO advisory frame.
Bottler-as-cornered-resource is worth distinguishing from formula-as-cornered-resource. Most casual analysis of Coca-Cola says “the formula is the moat.” Acquired’s analysis says no, the formula is a brand asset; the bottlers are the actual cornered resource. Worth a brief reference doc on the distinction — useful when discussing other “secret IP” businesses where the IP is brand-narrative, not literal IP.
“Repetition works” as a content-strategy concept page. Cross-link to RDCO editorial principles, Sanity Check voice consistency, the founder’s commitment to anti-silver-bullet messaging. Concept article candidate.
Goizueta-style moat renegotiation is a vault-concept candidate. When the moat structure ossifies, you have to renegotiate it before competitors can exploit it. Pattern: Coca-Cola bottlers, big-tech employment contracts (the AI-talent compensation race), franchise restaurant economics. Concept doc.
Curiosity question: are there sectoral antitrust exemptions in data/analytics regulation that nobody talks about? The 1980 soft-drink-bottler exemption is the example. Worth a research-backlog item.

Sponsorship

This episode included sponsor reads from:

JP Morgan Payments — Featured as season sponsor. Standard intro/mid-roll; same long-running Acquired sponsor as the TSMC, Trader Joe’s, and Live at Radio City episodes. Treat as paid placement; not a subject of analysis.
Work OS — Mid-roll. Auth/SCIM/SSO infrastructure. Same sponsor as Trader Joe’s episode. Treat as paid placement.
Shopify — Mid-roll. Standard fall 2025 season sponsor. Treat as paid placement; not editorial assessment of the company despite being the subject of the parallel ACQ2 episode this week (Tobi Lütke).
Sentry — End-roll. Software error monitoring. Standard sponsor read.

Note: the Coca-Cola business analysis is editorial and unpaid. The 4-hour episode is part of the fall 2025 season opener with rotating sponsors throughout. None of the four sponsors are subjects of analysis in this episode.