The complete narration of this episode, published from the approved script rather than from machine transcription, so the names are spelled the way they are spelled. Timestamps mark the 11 chapter boundaries; the lines between them run in order. Dramatized dialogue scenes are not included.
This is the most successful thing Starbucks ever built.
And this is the man who built it, on tape, two years ago, describing it.
The biggest Achilles heel. And it's not even a close second.
He is talking about a Chicago store at eight in the morning. He says:
Everyone shows up, and all of a sudden we got a mosh pit. And that's not Starbucks.
Seventeen years before he said that, the same man sat down and wrote a memo warning that exactly this would happen. He used a word for it. He was right. And then his company spent seventeen years building the thing anyway.
This is the story of how the company that invented the third place automated it away, and what it is now paying, by hand, to get it back.
On the fourteenth of February, two thousand seven, the founder of Starbucks, at that point no longer running it, wrote a memo to his chief executive. It was not meant to be public. It became public within a week.
the watering down of the Starbucks experience, and, what some might call the commoditization of our brand
Commoditization. That is the word. It is the word a founder uses when he is frightened that the thing he sells has stopped being special and started being a product.
He was writing about espresso machines that were too tall to see over. About stores that smelled of sandwiches instead of coffee. About the romance going out of it.
He was not writing about a phone. Nobody was writing about a phone. The iPhone had been announced five weeks earlier and would not go on sale for another four months.
In two thousand seven the man who built this company put it in writing that the danger was becoming a commodity. Everything that happens for the rest of this film is his own company walking into the word he wrote down.
Here is the machine, and here is how fast it arrived.
December two thousand fourteen: mobile order and pay is a pilot. One city. Portland.
September two thousand fifteen: it goes national. Seven thousand four hundred stores in a single step.
Share of United States transactions placed on the app: three percent. Then seven. Then about twenty-five. Then thirty-one percent. A record, in the quarter that ended in December two thousand twenty three.
Nobody has published a fresher number. Which is its own kind of answer.
A third of the business, ordered by people who never spoke to anybody, walking into a room built entirely around the idea that they would.
In January two thousand seventeen, on an earnings call, the man about to become chief executive stood up and described the problem in a sentence so corporate it is almost beautiful.
The tremendous success of mobile order and pay has also created a new operational challenge in our highest volume stores that has been building for several quarters, significant congestion at the hand-off plane.
The hand-off plane. That is the counter. That is the place where, for thirty years, the entire promise of the company was supposed to happen, where somebody said your name.
He said the word congestion eight times on that call.
And then he said the part that costs money. Customers, he said:
entered the store or considered visiting a Starbucks store and then did not complete a transaction.
They walked in. They looked at it. And they left.
At that point the app was seven percent of transactions. Seven. The crowding that made people turn around and walk out was being caused by seven percent.
It would go to thirty-one.
The thirteenth of August, two thousand twenty-four. Before the market opened.
Starbucks put out a press release. It said its chief executive was leaving, effective the previous day. And in the same release it named the man replacing him.
Not a search. Not an interim stretch announced and then filled some months later. One release, one morning, both halves of it.
Laxman Narasimhan had run the company for about seventeen months. The stock fell roughly twenty percent while he did. He was treated as terminated without cause. From the terms filed in the eight-K, the press computed his cash severance at about ten point six million dollars, with equity treatment reported up to about twenty-one million.
His replacement had a title before he had a start date. Chairman and chief executive, both, in the headline of the release.
Then the market answered, and it answered inside one session.
Starbucks closed up twenty-four and a half percent. That is its largest single-day gain since the company went public in nineteen ninety-two, and it put about twenty-one billion dollars of market value on the company in a day.
Chipotle, the company the incoming man was leaving, fell about seven and a half percent the same afternoon.
Two public companies re-priced, in one session, on the movement of one person. Nobody had made a cup of coffee any differently yet.
There is a question sitting underneath that morning. Who wanted it?
In July, three weeks earlier, the Wall Street Journal reported that Elliott Management had built a stake. The stock rose six or seven percent that day. The size everyone quotes, as much as two billion dollars, is a press estimate.
A Schedule thirteen-D is the filing that makes an activist show its hand. None was ever filed here. Elliott's own statement that August called it one of the largest investors in Starbucks. That sentence is the whole disclosure.
The push that preceded two chief executives' worth of strategy never had to show a share.
Now look at who they chose.
Brian Niccol ran Taco Bell from twenty fifteen and drove its mobile ordering rollout. He took over Chipotle in March twenty eighteen. Over his tenure the stock rose somewhere between seven and eight hundred percent, on a business he made one of the most digital in American fast food. By twenty twenty, about forty-six percent of everything Chipotle sold was ordered digitally.
The company whose founder called mobile ordering its biggest Achilles heel hired the industry's most successful builder of it, to take it back out.
On his second day he published a letter called Back to Starbucks. The diagnosis inside it is the memo's word again, in plainer English.
It can feel transactional. Menus can feel overwhelming. Product is inconsistent. The wait too long or the handoff too hectic.
Transactional. Commoditization. Two men, seventeen years apart, naming the same illness. The first wrote it in private and watched it leak. The second published it on purpose, on day two.
He was paid for the trouble. A ten million dollar cash signing bonus, and seventy-five million in replacement equity for what he forfeited at Chipotle.
All of that is the machine. This is the room the machine replaced.
Nineteen seventy-one. Pike Place, Seattle. Three men opened a shop that sold beans and equipment. Jerry Baldwin, Gordon Bowker, Zev Siegl. No drinks. You carried the beans home.
There was no counter to wait at, because there was nothing to wait for.
The man this film has been about does not arrive for another eleven years. When he does, he is a salesman, and what he is selling them is plastic cone filters.
Nineteen eighty-three. Starbucks sent him to a housewares show in Milan. He walked the city's espresso bars, and what he says he found there was not a product.
During my first trip to Milan in nineteen eighty-three, I was captivated by the sense of community I found in the city's espresso bars.
Community. That is the entire thesis of the company, and he is describing a room where people stand around and talk to a man who knows their names.
He comes home and tells the owners they should sell the drink, not the bean. They say no.
They say no for two years. That refusal is the most repeated scene in the story, and one of the men in it has been on the record about it since twenty sixteen.
Apocryphal. And: It's like Howard body-slammed Gordon and me... but that's not the case.
Baldwin disputes the blocked-visionary version of his own life. He also had reasons that never travel with the story. He and Bowker had just bought Peet's, and were carrying six dollars of debt for every dollar they owned.
So when the young man quit, in nineteen eighty-five, to build it himself, what happened next was the opposite of a rejection.
Starbucks put a hundred and fifty thousand dollars into the company that would one day buy it, and Baldwin took a seat on its board. By Schultz's own book, the men who said no were his first investors.
He raised the rest at ninety-two cents a share. Four hundred thousand dollars of seed money, and then a year of asking. A physician named Ron Margolis, by Schultz's telling, never asked for a business plan. He asked how much was needed and wrote a hundred thousand dollars on the spot. That hundred thousand was later worth more than ten million.
Two years after that, the owners sold him the company he could not talk them into changing. He signed on a Friday afternoon in August, nineteen eighty-seven. By his own account he raised about three point eight million dollars to do it. No primary document for that figure has ever been produced, in any telling of this story.
He was thirty-four years old, and he owned the room.
There is one number everybody knows about that year of asking.
Two hundred and forty-two people. Two hundred and seventeen of them said no. It is in his book, on the page, in his own words. We read it there.
Two things never travel with it. The first is that the year of rejection was for Il Giornale, the espresso company he started after he left. It was not the founding of Starbucks.
The second is that eleven months before the book came out, he told the Seattle Times a different version. Two hundred and forty-two solicited, twenty-three put up money. That is two hundred and nineteen refusals.
Same man, same year, two counts, and no list of names has ever existed to settle it.
The podcast usually cited as the independent check is not one. The host says the numbers back to him and he agrees. That is the book being read aloud.
Forty years on, the company is still arguing about whose count is real. Only now the people doing the counting work there.
There is a second story running underneath all of this, and it belongs to the people who make the drink.
Since twenty twenty-one, workers have voted to unionize at seven hundred Starbucks stores, covering more than fifteen thousand people. That is elections won, from Labor Board data, tallied on the twenty-fifth of June this year.
Starbucks frames the same facts differently, and its framing is also accurate. Fewer than one percent of its ten thousand-plus company-operated United States stores. Under four percent of its baristas.
Both are true. They answer different questions.
The record in between has been adjudicated, and it does not run one way. In twenty twenty-three a labor board judge found, across twenty-one Buffalo-area stores, that the company had broken federal labor law hundreds of times. His phrase was egregious and widespread misconduct.
Then in June twenty twenty-four the Supreme Court sided with Starbucks. In Starbucks against McKinney the court held that labor board injunctions require the ordinary four-factor test, not the looser one used in the Memphis case. Eight to one, and no justice voted to keep the lower courts' standard.
Here is the part that is not in dispute. Four and a half years after the first store voted, no collective bargaining agreement has been ratified at any Starbucks store in the United States.
About thirty-three tentative agreements were reached by April twenty twenty-five. Union delegates rejected the economic package, eighty-one percent against, by the union's count.
On Red Cup Day, the thirteenth of November twenty twenty-five, the union began the longest strike in the company's history. It ran a hundred and thirty-one days, into March.
The union counted more than a thousand baristas at sixty-five stores at the start, growing to roughly three thousand eight hundred. Starbucks said less than one percent of its coffeehouses were experiencing any level of disruption, and that the vast majority of its two hundred and forty thousand partners came to work.
A company whose entire product is a welcoming room has spent four and a half years not signing a contract with the people standing in it.
So what does undoing a machine actually look like? Mostly it looks like furniture and labour.
Roughly thirty percent of the menu, cut. Handwritten names back on the cups. Ceramic mugs and free refills if you stay. A four-minute service standard, backed by more than five hundred million dollars of added labour hours. People on the floor instead of another screen.
Six hundred and twenty-seven underperforming stores closed. And the purest piece of app-era architecture, the pickup-only store, was discontinued outright. The strategy before this one had wanted that format for fourteen percent of every new store.
And the company committed to bolting more than twenty-five thousand cafe seats back into its United States stores by the end of this fiscal year. Chairs. Into rooms redesigned to get you out of them.
Then the strangest repair of all. In May twenty twenty-six, scheduled mobile ordering launched across North America. You pick an exact pickup time, up to an hour ahead.
The app that made the mosh pit learned to make an appointment. Nineteen years after the memo. Eleven after the national rollout.
And it is working, which is the part that complicates the story.
In the quarter that ended in December, United States comparable sales rose four percent, and transactions grew for the first time in eight quarters. In the quarter that ended in March, comparable sales rose seven point one percent, transactions four point three, and the company raised its guidance for the year.
The warmth came with discipline attached. By May this year the notes on the cups were mandatory, with write-ups possible for a barista who did not write one, and the union filed a complaint with the labor board over it. A filing is an allegation, not a finding.
And the retreat that helped pay for it: in March the company closed the sale of sixty percent of its China business. The company that once opened a store there every fifteen hours now owns forty percent of it.
The man hired to undo the machine says the plan is working. The receipts, so far, agree with him.
On the twenty-ninth of July, two thousand twenty-six, after the market closed, the report card landed.
United States comparable sales rose seven point nine percent. But the number this film has been waiting on is the one underneath it. Transactions rose four point two percent. People, not prices. That is the third quarter in a row.
A year ago this same business was shrinking. North America transactions fell three point three percent in the third quarter of last year. This year they rose four point five.
And the company raised its guidance for the year again.
The top line says the opposite happened. Consolidated revenue fell one percent, to nine point three billion dollars. That is not customers leaving. In April the company handed its China stores to a joint venture and stopped counting their sales as its own. In North America, where the repair is actually happening, revenue rose seven percent.
Adjusted earnings came in at eighty-five cents a share, up seventy percent. Most of that lift is not coffee. The company's adjusted tax rate fell nine hundred and sixty basis points. The operating figure is the honest one, and it is strong. Adjusted operating margin widened four hundred and thirty basis points.
The chief executive's own words on it: Our third quarter results are proof they do. And, in the same statement, We have more work to do.
And the stores kept opening. A hundred and seventy-five net new stores in the quarter, ending at forty-one thousand three hundred and four.
One number is not in here. The company did not say what share of its business now moves through the app. The last time it disclosed that figure was the first quarter of fiscal two thousand twenty-four, when it was thirty-one percent. The machine at the centre of this story is, on the public record, still unmeasured.
Nineteen years ago the man who built this company wrote down what it was costing him. The company built it anyway. It is now paying, seat by seat, to take it back out. On this quarter's evidence, the customers are coming back. Whether that holds is next quarter's print.
Anchor
Origin
Act I
THE MYTH
THE INVERSION
THE FUNDRAISE
THE FIND
Act II
OPEN THREAD
Act III
LANDMINE
KILLER CARD
HONESTY BEAT
CORRECTION
RIGHTS
OPEN GAP
THE DRIFT
CODA
THE AUDIT
THE LEGEND
Act IV (2008-10)
MODERN LANDMINE
Act V (digital)
Act VI (interregnum)
Act VII (union)
Act VIII (Niccol)
Act IX (present)
Compounded is a documentary series produced by Agate Street Studios. Dramatizations use generated imagery; figures and quotes are sourced from public filings and reporting. This is not investment advice. Do your own research before making any financial decision.
NOW · E01 Netflix
NOW · E02 Trader Joe’s
NOW · E03 Crocs
NOW · E04 Patagonia Notes from Andrew
Key takeaways, the research stills behind this episode, and the occasional director’s-cut note.
