A WORKING PAPER ON BRANDS, PRICING POWER AND THE COST OF BEING BELIEVED

THE CANDY STORE THAT ATE WALL STREET

What Warren Buffett Has Actually Said About Marketing – Drawn Entirely From Sixty Years of Berkshire Hathaway Shareholder Letters – And What the Conglomerate Kings of the 1960s Got Catastrophically Wrong

Warren Buffett has written roughly 60 annual letters to Berkshire Hathaway shareholders. He has never once used the phrase “brand equity.” He has never drawn a funnel, never mentioned a persona, and never, as far as anyone can tell, sat through a positioning workshop.

He also turned $1 into $55,023 while the stock market turned it into $391 – and a startling amount of that gap came from businesses whose entire competitive advantage was that customers liked them. A candy company. A car insurer with a talking lizard. A furniture store in Omaha. If you want to know what marketing is worth, you could read a textbook, or you could read the financial statements of a man who paid cash for it and kept the receipts.

This paper does the second thing. Every Buffett claim below is sourced to a specific letter, by year. The numbers are his. The conclusions are the ones the numbers force.

Source: Berkshire Hathaway 2024 Annual Letter, performance table, pp. 14–15. Compounded annual gain 1965–2024: Berkshire 19.9%, S&P 500 with dividends 10.4%. Overall gain 1964–2024: 5,502,284% versus 39,054%.

1. FIRST, THE UNCOMFORTABLE PART: MOST OF THIS MONEY CAME FROM MARKETING

Berkshire is filed away in most people’s minds as an investment company. It isn’t, and hasn’t been for decades. As of the 2024 letter it owns 189 operating businesses outright and posted $47.4 billion in operating earnings. Buffett’s own framing is blunt: Berkshire’s equity activity is ambidextrous – one hand holds whole companies, the other holds slices of a dozen or so household names.

Look at what’s in both hands and a pattern jumps out. Coca-Cola. American Express. Apple. See’s. GEICO. Dairy Queen. Duracell. These are not companies with better factories. They are companies that customers have decided something about. Buffett has spent sixty years buying that decision.

Era	Berkshire, annual	S&P 500, annual	The spread	What Buffett was buying
1965–1974	13.0%	1.2%	+11.8	Cigar butts: cheap, ugly, no brand
1975–1984	42.8%	14.8%	+28.0	See’s, Buffalo News, Nebraska Furniture Mart
1985–1994	27.9%	14.4%	+13.5	Coca-Cola, Gillette, Borsheims, Scott Fetzer
1995–2004	14.9%	12.1%	+2.8	GEICO in full, FlightSafety, NetJets, Dairy Queen
2005–2014	9.9%	7.7%	+2.2	BNSF, Iscar, Lubrizol, Heinz, Duracell
2015–2024	13.7%	12.9%	+0.8	Apple, Japanese trading houses, buybacks

Decade averages computed from the per-share market value series in the 2024 letter. The spread narrows as size bites – Buffett’s own warning: “our advantage, if any, won’t be great.” But note where the two widest decades sit. They sit on top of the brand purchases.

2. RULE ONE – A BRAND IS NOTHING BUT PERMISSION TO RAISE YOUR PRICE

In the 1991 letter Buffett stops discussing stocks entirely and writes what is, if you squint, the best one-paragraph definition of marketing ever put in an annual report. He calls it an economic franchise, and he gives it a three-part test.

The test	What it means in plain English	Passes	Fails
It is needed or desired	Somebody wakes up wanting this. Demand is not something you have to manufacture every quarter.	See’s at Christmas	Most B2B software
Customers believe there is no close substitute	Note the word believe. This is a fact about the customer’s head, not about the product.	Coca-Cola	Regional airlines
It is not subject to price regulation	Nobody in a state capitol gets a vote on what you charge.	Gillette blades	Utilities, hospitals

Source: Berkshire Hathaway 1991 Chairman’s Letter. Buffett’s proof that all three conditions hold is behavioral, not theoretical: the company can price aggressively, year after year, and earn high returns on capital anyway.

Read the second condition again, because it is the whole game. Buffett does not say the product has no close substitute. He says customers think it has no close substitute. There is a perfectly good chocolate in every supermarket in California for a third of the price. It does not matter. Marketing, in Buffett’s accounting, is the business of installing that belief and then never breaking it.

He adds a line that should be printed on the wall of every marketing department: franchises, he wrote in 1991, “can tolerate mis-management.” A weak business needs a genius running it. A real brand survives idiots. That is the actual return on marketing investment – not this quarter’s lift, but the size of the mistake the business can absorb without dying.

3. THE SEE’S CANDIES FILE: 2% VOLUME GROWTH AND $1.35 BILLION OUT THE DOOR

Berkshire bought See’s Candy Shops on January 3, 1972 for $25 million. Buffett almost blew it. The family wanted $30 million, he refused to go above $25 million, Munger told him he was wrong, and the sellers caved. In the 2007 letter Buffett notes that his stubbornness over $5 million nearly cost him $1.35 billion.

Here is what a brand actually looks like on a spreadsheet.

Line item	1972, at purchase	2007	Change
Pounds of candy sold	16 million	31 million	+94%
Revenue	$30 million	$383 million	+1,177%
Implied price per pound	$1.88	$12.35	+558%
Pre-tax earnings	under $5 million	$82 million	roughly 17x
Capital required to run it	$8 million	$40 million	+$32 million
Pre-tax return on capital	60%	205%	–
Cumulative pre-tax earnings, 1972–2007	–	$1.35 billion	–
Cumulative pre-tax earnings, through 2014	–	$1.9 billion	–

Sources: Berkshire Hathaway 2007 Annual Letter, “Businesses – The Great, the Good and the Gruesome”; and the 2014 special letter “Berkshire – Past, Present and Future.” Price per pound is implied by dividing revenue by pounds.

The 1997 figures (31 million pounds, $269 million) come from Berkshire’s segment disclosure, which broke See’s out separately until 1999. Volume grew about 2% a year for 35 years. Nobody at See’s was chasing growth. They were chasing price.

Why this is the single most important table in the letters

1. Growth was almost entirely price, not units. Volume compounded at roughly 2% annually. Revenue compounded at about 7.5%. The gap is pricing power – the cash value of being the chocolate Californians give their mother.
2. The growth was free. Berkshire had to reinvest only $32 million over 35 years. Buffett’s benchmark: a normal company going from $5 million to $82 million of earnings would need around $400 million of investment. See’s needed one-twelfth of that.
3. The product was sold for cash, which killed receivables, and the production cycle was short, which killed inventory. Distribution design is a marketing decision with balance-sheet consequences.
4. See’s earns close to half the entire industry’s profits – in a category Buffett flatly calls unexciting, with low and flat per-capita consumption, where most historic brands are already dead. A dull market with one loved brand beats a hot market with six.
5. Quality was treated as untouchable. In the 1983 letter Buffett makes the point that See’s would not trade product quality for a better margin – the franchise came first and the cost line came second. When management tried to cut 14 of the 100 candies in 1987, customers revolted and the decision was reversed.

Buffett’s Great / Good / Gruesome taxonomy from the 2007 letter, quantified. FlightSafety is a fine business – durable advantage, essential service, 58% of U.S. corporate pilots trained. It just has to buy $12 million simulators to grow. See’s has to buy nothing.

4. RULE TWO – ADVERTISING IS NOT AN EXPENSE. IT IS A CAPITAL PROJECT WITH BAD ACCOUNTING.

If See’s is the argument for brand, GEICO is the argument for spending money on it. And GEICO is where Buffett says the quiet part out loud.

In the 2010 letter he explains that Berkshire enthusiastically spent $900 million on GEICO advertising to acquire policyholders who delivered no immediate profit whatsoever. Then the line that should end most budget arguments forever: “If we could spend twice that amount productively, we would happily do so.” He adds, without apology, that short-term results would be further penalized.

Read that as an accountant and it is madness. Read it as a marketer and it is simply a discounted cash flow. A new GEICO policyholder costs money once and pays premiums for years. The first-year loss is not a loss. It is the purchase price of an annuity, booked in the wrong period because GAAP has no line called “customers acquired.”

Year	GEICO marketing spend	U.S. auto share	Context from the letters
1993	–	2.0%	Tony Nicely promoted to CEO. Profitable but not growing; 7th largest auto insurer.
1995	$33 million	2.5%	652 telephone counselors. Berkshire buys the other half for $2.3 billion.
1997	about $67 million	3.0%	Industry does $115 billion a year. Buffett: “We’re planning to step on the gas.”
1998	$143 million	–	Counselor count up to 2,162 – capacity built alongside demand.
1999	$242 million	–	Buffett states there is no limit to what Berkshire will invest in new business.
2007	$751 million	7.2%	Ad spend up 23x since 1995; share up nearly 3x. Motorcycle share 2.1% to 6%.
2010	$900 million	8.8%	Third-largest U.S. auto insurer. Buffett wants to double the spend.
2023	$838 million	–	Spend cut hard under Todd Combs; policies in force fall 9.8%.

Marketing figures 1995–1999 from the 1997, 1998 and 1999 Chairman’s Letters; 2007 and 2010 from those years’ letters; 2023 from S&P Global Market Intelligence statutory filings analysis. Share figures as reported by Buffett in the 2007 and 2010 letters.

Twenty-seven times the budget bought roughly four times the share. That is not a great conversion rate – and Buffett kept spending anyway, because he was buying a lifetime of premiums, not a quarter of them.

The 2023 footnote nobody wants to talk about

The last row is the control experiment. When GEICO cut advertising roughly 35% in 2023 – down $443 million from 2022 – private auto policies in force fell 9.8%, on top of an 8.9% decline the year before; the company’s own 10-K names the ad reduction as a cause. Margins improved, and Buffett called the 2024 result spectacular. Both things are true, and that is the point: advertising at GEICO behaves like maintenance capital expenditure. Skip it and the numbers look better right up until the asset shrinks.

5. RULE THREE – SHARE OF MIND COMES FIRST. SHARE OF MARKET IS THE RECEIPT.

The 1993 letter contains Buffett’s clearest statement on what a brand is worth. Comparing Coca-Cola and Gillette to the technology companies of the day, he argues their business risk is lower, and credits “the might of their brand names” alongside product attributes and distribution strength.

Business	Share of its global market	Berkshire’s read
Coca-Cola	about 44% of all soft drinks	Bought for roughly $1.3 billion, 1988–94. Dividend received: $75 million in 1994, $704 million by 2022.
Gillette	over 60% of blade market by value	A razor is a commodity. A Gillette is not. That is a marketing achievement, not an engineering one.
Wrigley	dominant in chewing gum	Buffett’s aside: he knows of no other significant businesses with such durable global power.
See’s	close to half of industry profits	Regional, not global – but the same mechanism at smaller scale.
NetJets	near 90% of the large-cabin market	Fleet more than twice the size of its three main competitors combined (2007).
Clayton Homes	record 31% share (2007)	Won share while industry volume fell from 131,000 units to 96,000.

Sources: 1993 Chairman’s Letter (Coca-Cola, Gillette, Wrigley); 2007 Annual Letter (See’s, NetJets, Clayton); 2022 letter (Coca-Cola dividend history).

Notice what these have in common. Not one is the technically superior product in its category. Every one is the default answer to a question. In the 2007 letter Buffett gives that the name everyone now uses – the moat – and lists exactly two reliable ways to dig one: be the low-cost producer, like GEICO and Costco, or own a powerful worldwide brand, like Coca-Cola, Gillette and American Express. Two strategies. That is the whole menu. Everything else is a Roman candle, his term for companies whose moats proved illusory and were crossed almost immediately.

6. RULE FOUR – “SELL CHEAP AND TELL THE TRUTH”

The other moat is price, and Berkshire’s teacher on that subject was a Belarusian immigrant named Rose Blumkin who could not read English, started with $500 borrowed from her brother in 1937, and ran Nebraska Furniture Mart out of a pawnshop basement.

Her entire marketing strategy was four words: sell cheap and tell the truth. Buffett bought 90% of the business in 1983 for $60 million on a handshake. He did not commission an audit. He did not ask for one.

1. The Fair Trade lawsuit is the best PR case study in the letters. Omaha retailers pressured manufacturers not to sell to her; she got merchandise anyway and cut prices; she was hauled into court for violating Fair Trade laws. She won every case, received what Buffett calls invaluable publicity, and at the end of one trial sold the judge $1,400 worth of carpet.
2. One store. Over $100 million a year. In the 1983 letter Buffett reports NFM doing more than $100 million in annual sales from a single 200,000 square-foot store – more than any home furnishings store in the country. By 2007 the Omaha and Kansas City stores were doing roughly $400 million each.
3. Low cost is a marketing position, not a finance one. GEICO’s advantage is identical in structure: rock-bottom operating costs let it offer Americans the cheapest way to buy something they are legally required to own. Buffett calls that an enduring moat competitors cannot cross.
4. Buffett’s own verdict on competing with her: “I’d rather wrestle grizzlies than compete with Mrs. B.”

And then there is the annual meeting, which is a trade show wearing a disguise

Berkshire’s annual meeting – Buffett’s Woodstock for Capitalists – is the most under-analyzed marketing asset in American business. It looks like a shareholder obligation. It is a retail event, a recruiting event, an owner-loyalty program and a media buy, all financed by people who pay their own airfare to attend.

Metric	1997	2007	Recent
NFM “Berkshire Weekend” sales	$5.3 million	$30.9 million	–
Attendance	–	about 27,000	tens of thousands
Exhibition hall devoted to selling	–	194,300 sq ft	–
Borsheims sales gain, shareholder weekend	–	+27%	–
Dinners served at Gorat’s, Shareholder Sunday	–	915 in one day	–
Poor Charlie’s Almanack sold at the meeting	–	~50,000 lifetime	5,000 in one day (2024)

Source: Berkshire Hathaway 2007 Annual Letter, “The Annual Meeting,” and the 2024 letter. Note the 1997–2007 figure: weekend sales at one furniture store grew 483% – more volume in four days than most furniture stores do in a year.

Buffett states the mechanism without embarrassment. He tells shareholders the doors open at 7 a.m., invites them to shop, jokes that he will lock the doors if they don’t, and notes that manufacturers with ironclad no-discounting policies make an exception for the weekend. He is running a promotion. He is just honest about it, which is the point.

7. RULE FIVE – A MOAT YOU HAVE TO KEEP REBUILDING WAS NEVER A MOAT

From the 2007 letter, and it is the most quotable sentence Buffett ever wrote about durability: “A moat that must be continuously rebuilt will eventually be no moat at all.”

This is the rule that kills most modern marketing. Performance channels that stop working the day you stop paying are not moats. They are rent. Buffett’s criterion of enduring also rules out businesses that depend on a superstar: a partnership built around the best brain surgeon in town has outsized earnings and no moat, because the moat leaves when the surgeon does. You cannot name the CEO of the Mayo Clinic. That is the whole argument.

What it costs when you get durability wrong: the Dexter file

The Dexter Shoe transaction	Figure
Purchase price, 1993, paid in Berkshire stock	$433 million
Shares handed over	25,203 Class A
Value of the business a few years later	zero
Value of those shares as of the 2007 letter	$3.5 billion
Value of those shares as of the 2014 letter	$5.7 billion
Share of Berkshire given away for nothing	1.6%

Sources: 2007 Annual Letter (“confession time”) and the 2014 special letter, where Buffett says the error deserves a spot in the Guinness Book of World Records. His diagnosis: what he assessed as durable competitive advantage vanished within a few years because of foreign competition.

A brand that a cheaper foreign factory can erase in five years was never a brand. It was a distribution arrangement with a logo on it. The bill for confusing the two was $5.7 billion.

8. THE CONGLOMERATE GRAVEYARD – WHERE MARKETING GOES WHEN IT’S POINTED AT THE WRONG AUDIENCE

In 2014, for Berkshire’s fiftieth anniversary, Buffett wrote a long history of the company. Halfway through he admits that Berkshire is a sprawling conglomerate and that conglomerates richly deserve their terrible reputation – then explains, with obvious enjoyment, exactly why. The 1960s conglomerate boom was the greatest marketing campaign in American corporate history. Its product was the stock. Its customer was Wall Street. Its mechanism was arithmetic dressed as genius.

The trick, in Buffett’s own reconstruction

1. By personality, promotion or dubious accounting – and often all three – a CEO drives his young conglomerate’s stock to roughly 20 times earnings.
2. He then issues shares as fast as possible to buy businesses trading at ten-or-so times earnings.
3. Pooling accounting is applied immediately, which raises per-share earnings without a dime of change in the underlying businesses.
4. The rise is presented as proof of managerial genius, which justifies keeping – or even raising – the acquirer’s multiple.
5. He promises to repeat this forever. The press applauds. Bankers collect fees. Auditors, as Buffett puts it, sprinkle their holy water.
6. The fatal side effect: since the game requires buying low-multiple businesses, the CEO must bottom-fish. The collection gets junkier every year. Nobody cares, because investors are buying deal velocity, not businesses.

Conglomerate	The pitch	The numbers at the peak	How it ended
ITT (Harold Geneen)	Management science: 250+ profit centers under one financial control system	Sales $765M (1961) to about $17 billion; profits $29M to $550M; 58 consecutive quarters of earnings growth; ~350 acquisitions in 80 countries	Broken up in the 1990s. Buffett’s verdict on the whole class: now long gone
LTV (Jimmy Ling)	“Project redeployment” – buy a company, spin the pieces out at higher multiples	Sales of $36 million in 1965 to No. 14 on the Fortune 500 two years later	Ling himself was, in Buffett’s phrasing, spun off – that is, fired
Litton (Tex Thornton)	Technology-flavored diversification, from typewriters to shipbuilding	Roughly $3 million in 1954 to nearly $2 billion by the late 1960s	Share price fell 86% into May 1970
Gulf+Western (Charles Bluhdorn)	Auto parts to Paramount Pictures to Simon & Schuster	Borrowed $85 million from Chase against $5 million of cash – three times the combined annual worth of its companies	Dismantled through the 1980s; renamed Paramount Communications, 1989
Teledyne (Henry Singleton)	No pitch at all. Singleton barely spoke to Wall Street.	20.4% annual returns 1963–1990 vs ~11.6% for the index; retired ~90% of shares for $2.5 billion; EPS up 40-fold 1971–84	The one that worked

Buffett names ITT, Litton, Gulf & Western and LTV directly in the 2014 letter. Supporting figures: Encyclopedia.com and Harvard Business School leadership profile (ITT); D Magazine, October 1982, on Ling, cited by Buffett himself; Wikipedia and Forbes (Litton, Gulf+Western); The Outsiders and Teledyne shareholder records (Singleton).

Peak-to-trough share price declines, late 1968 to May 1970. The ten largest conglomerates of the era lost an average of 86%. The Dow lost 13% over the same quarter. Source: contemporaneous market data as compiled by Nasdaq/Dun’s Business Review.

Ling’s spinoffs, and the sentence Buffett wrote about him

In 1967 Ling bought Wilson & Co., a meatpacker that also sold golf equipment and pharmaceuticals. He promptly split it into three partially-spun-off companies, which Wall Street immediately nicknamed Meatball, Golf Ball and Goof Ball. LTV’s 1966 annual report had explained the underlying magic: acquisitions must satisfy the formula that two plus two equals five, or six. Buffett’s assessment of Ling is the best line in the letter: Munger taught him “never underestimate the man who overestimates himself.”

ITT grew sales faster than anyone. It no longer exists in the form Geneen built. Teledyne and Berkshire, which did the least promotion, produced the returns. Growth in revenue and growth in shareholder value are different products with different customers.

9. RULE SIX – 2 + 2 = 4, AND ANYONE SELLING YOU OTHERWISE IS SELLING

Buffett’s closing argument on the conglomerate era is not really about conglomerates. It is about persuasion, and it applies wherever a story is doing work the numbers cannot support. Business models built on serially issuing overpriced shares, he writes, work exactly like chain letters: they redistribute wealth spectacularly and create none. The ending never varies – money flows from the gullible to the fraudster.

His prediction: more Jimmy Lings will appear. They will look and sound authoritative, the press will hang on their words, bankers will fight for their business, and whatever they are saying will recently have worked. His advice – remember that two plus two will always equal four, and when someone tells you that math is old-fashioned, zip up your wallet.

The corollary for people who market things for a living

1. Never let the marketing outrun the product. Every conglomerate above eventually had to deliver businesses as good as the story. None could.
2. Watch what the incentive structure is actually buying. Buffett notes that fees too often lead to transactions rather than transactions leading to fees. Read “fees” as “campaign budgets” and the sentence still lands.
3. Post-mortems are the missing discipline. Buffett, a director of 19 public companies, says he has never once heard “dis-synergies” discussed after a deal closes, and that honest comparisons of reality to projection are rare in American boardrooms. They should be standard practice.
4. The ABCs of decay are arrogance, bureaucracy and complacency. Buffett names General Motors, IBM, Sears Roebuck and U.S. Steel – companies that sat atop their industries and whose strengths seemed unassailable. Historical earning power proved no defense.
5. Fooling the customer starts with fooling yourself. From the 2024 letter, on why Berkshire reports its own mistakes: “if you start fooling your shareholders, you will soon believe your own baloney.” He notes he used the words mistake or error 16 times in five years of letters, and that many huge companies have never used either word once.

10. THE SCORECARD: BUFFETT’S MARKETING RULES, AND WHAT EACH ONE COST HIM TO LEARN

The rule	Where it comes from	The evidence	The price of ignoring it
A brand is permission to price aggressively	1991 letter	See’s: price per pound up 558%, volume up 94%	You compete on features forever
Advertising is capital spending misfiled as expense	2010 letter	$33M to $900M at GEICO; share 2.5% to 8.8%	GEICO 2022–23: spend cut 35%, policies down 9.8%
Belief beats substitutes	1991, 1993 letters	Coke at 44% of global soft drinks; Gillette above 60% of blades by value	Commodity margins
Sell cheap and tell the truth	1983 letter	$500 in 1937 to $100M+ from one store	Somebody else becomes the cheap honest option
Quality is not negotiable against margin	1983 letter	See’s near half of all industry profits	The 1987 candy cut, reversed after customer revolt
A moat you rebuild yearly isn’t one	2007 letter	Dexter: $433M purchase, $5.7B in stock, zero value	Roman candles
Own the event, don’t rent the audience	2007 letter	NFM weekend: $5.3M to $30.9M in a decade	You pay for attention every single time
2 + 2 = 4	2014 letter	ITT, LTV, Litton, G+W: all lionized, all gone	Average 86% loss, 1968–1970

The one-sentence version

Buffett’s entire theory of marketing is that a brand is a financial instrument: you buy it once with quality, honesty and patience, it pays out in pricing power for decades, it requires almost no maintenance capital, and the only way to destroy it is to break the promise that created it. Everything else – the campaigns, the channels, the funnels – is either building that instrument or renting attention until you can afford to.

He paid $25 million to learn it, and has said more than once that watching See’s operate taught him the value of powerful brands and opened his eyes to every profitable investment that followed. That education is free. It runs about 1,500 pages and sits on a website that still looks like 1997 – which is, when you think about it, its own kind of brand statement.

SOURCES AND NOTES

Primary – Berkshire Hathaway Chairman’s Letters, berkshirehathaway.com/letters/letters.html:

1983 (See’s quality; Mrs. B and the Fair Trade suit) · 1991 (the economic franchise test) · 1993 (Coca-Cola and Gillette global share) · 1997 (GEICO budget, 3% share, $115bn industry) · 1998–1999 (spend, counselor headcount) · 2007 (The Great, the Good and the Gruesome; the moat; See’s accounting; GEICO share; NetJets; FlightSafety; Clayton; meeting figures; the Dexter confession) · 2010 (the $900 million decision) · 2014 special letter, “Berkshire – Past, Present and Future” (conglomerates, Jimmy Ling, the ABCs of decay, Dexter at $5.7bn) · 2022 (Coke dividends) · 2024 (operating earnings, 189 businesses, the sixty-year table).

Secondary, for the conglomerate history and market data:

Encyclopedia.com and Harvard Business School’s 20th Century Leaders profile (ITT financials) · D Magazine, October 1982, on Jimmy Ling – the article Buffett tells readers to look up · Nasdaq/Dun’s Business Review on the 1970 crash · Wikipedia (Litton, Gulf and Western, Nebraska Furniture Mart) · S&P Global Market Intelligence, April 2024, on 2023 insurer ad spend · The Outsiders and Distant Force on Teledyne · Berkshire segment disclosures through 1999 for See’s. Where a figure is implied rather than stated – price per pound, decade averages, capital ratios – it is derived arithmetically from numbers Buffett reports directly.