Culture is a tool.

The argument in three lines. Rows are the claims; the second column is the example that carries each one.

Claim	Example
Same goal, same values.	Startup values pages. Integrity tops the list in three separate counts of large companies.
Different goal, different values.	A Benedictine vows to stay in one house for life. A Jesuit vows to leave at once for wherever he is sent.
The context does the rest, and sometimes it comes first.	Google. Its success made its culture possible, not the other way round.

All culture sessions at software startup company retreats are the same.

The room is asked what makes the company unique. The wall ends up with speed, ownership, the customer and some word for excellence.

I have three observations from the companies I worked in. The values are always the same.

Nobody beyond their first two jobs believes in them. Everybody still does the session.

I have nothing against culture. It matters a lot.

Culture is a tool. Organizations keep trying to make it an absolute. It isn't one.

By the end of this piece you should be able to read an organization's goal and predict its culture before anyone picks up a marker. It works on a seed-stage startup and on a religious order founded in 1540.

Researchers have counted this. Donald Sull's team at MIT read the official values of nearly 700 large companies and found that more than 80% publish a list.

The same few words top every count.

Integrity is on the wall in all three studies

Enron's four values, printed in its 2000 annual report, were Communication, Respect, Integrity and Excellence.

Then there is the belief. The people who write the values and the people who live under them report different things.

Almost everyone publishes values. About a quarter believe them.

Sull's team then compared each company's official values with how its own employees described the culture in more than a million Glassdoor reviews. They found no correlation.

Edgar Schein, who more or less founded the field, defined culture as what a group learned while solving its problems of survival. It worked, so the group kept it and taught it to newcomers.

In that definition the problem comes first.

It starts with the goal and it grows with the context. Sometimes it runs the other way, and the context comes first.

I worked inside two of the more different cultures around: Google and Rocket Internet.

Google was a monopoly, as a US court ruled in 2024. It could afford a culture built around its people.

It's not the culture that made Google. It's Google's position and success that made the culture possible.

A bunch of engineers found themselves wildly successful, and that turned into a culture nobody else had.

Try to apply that amazing culture to launching ecommerce copycats in the developing world. A disaster would ensue.

That was Rocket's job. In October 2011 Oliver Samwer emailed his teams a plan to be number one in online furniture within a year. He called it a "blitzkrieg", a word he later apologised for.

I ran ZALORA markets for Rocket.

My hypothesis is simple. Tell me who funds the company and what kind of market it is in, and I can guess the words.

My hypothesis, with examples. Rows are the situations a company can be in; the columns give the values I expect and words that real companies in that situation publish.

Situation	What I expect	Words on the page
VC-backed	Speed. The investor's return is the goal, and a fund has about ten years to produce it.	Meta: "Move Fast". Stripe: "Move with urgency and focus". Uber until 2017: "Always be hustlin'".
A competitive market	Speed and excellence	Amazon: "Bias for Action", "Insist on the Highest Standards". Netflix: "Great and Always Better". Shopify: "Chase excellence".
An exploratory field	Innovation, creativity and their variants	OpenAI: "Creativity over control". IDEO: "empathy, humility, and open-mindedness". Netflix: "Curiosity", "Creativity".
A regulated environment	Integrity, safety, control	JPMorgan Chase: "fortress balance sheet", "good compliance is good business". Goldman Sachs: "Integrity". Boeing: "Safety & Quality", listed first.
Bootstrapped	Resourcefulness and independence. That is what the company and the founder are about.	Zoho: "Leaving money on the table". 37signals: "Bootstrapped, profitable, & proud".

These are minor fluctuations. Most companies operate in competitive markets, need to get things done and need to grow revenue or profit.

All companies' cultures are more or less the same because the goals are the same.

A real test needs organizations with other goals.

What if we don't call it culture, and we look at political parties and religious organizations? Their goals are different, and here the values finally change.

They change so much that they are not even called cultural values anymore. But they are there.

Every organization has a culture, written or not, shouted in a company retreat with 20-somethings or practiced in the corridors of Santa Marta.

scroll the table sideways →

Each row is compressed from encyclopedias and standard histories, and the goals are my reading. I am not a historian.

The cleanest case sits inside one church. A Benedictine vows to stay in one house for life. A Jesuit vows to leave at once for wherever he is sent.

They share a God, a pope and a set of scriptures. The job is different, so the value is the opposite.

Rome is the same organization with two goals. The Republic ran on aristocrats competing for office and military glory. Its armies were citizens levied for one war at a time.

Augustus needed the opposite. After decades of civil war he had to convince Romans that prosperity without war beat the wealth and honour a war could bring.

So the legions became a permanent paid profession, he became the ultimate source of patronage, and peace got its own altar. He kept the Republic's name and its letters, SPQR.

The Nazi party shows the change at its most brutal. Before 1933 the goal was to win power, and the culture was struggle: permanent campaigning and the SA fighting in the streets.

After 1933 the goal was to hold the state, and Hitler needed the army. The SA's chief wanted a "second revolution" and wanted his men to replace that army.

The culture that had won power now threatened the new goal. In three days in 1934 the regime killed the SA's leaders, at least 85 people in all.

Santa Marta is the quiet version. Casa Santa Marta, in Latin Domus Sanctae Marthae, is the Vatican guesthouse where Pope Francis lived from 2013 until his death in 2025.

In December 2014 he listed the Curia's habits to its face: gossip, rivalry, courting superiors, closed circles. None of that was written on a wall. It was the culture anyway.

The nation state is the most important organization of our time. It is the same tool at the largest scale, installed through schools and armies.

Ernest Gellner argued that an industrial economy needs a mobile, literate workforce, so the state spreads one standard culture through schooling. Each state still wrote its own version, because each had its own problem.

scroll the table sideways →

France and Italy needed a people. Prussia needed soldiers, Japan needed to catch up, Turkey needed a break with its past and Singapore needed to survive.

Singapore is the one that did it in the open. A government wrote five values and a parliament voted on them, which makes it the largest culture session I know of.

Italy has the best line. "Italy is made, now we must make Italians" is credited to Massimo d'Azeglio, who probably never said it in those words.

The fourth column of the long table is where the culture lives. None of these organizations relied on a workshop. The values were installed through who got admitted, promoted, punished and paid.

Culture is a tool to achieve the goal of the organization. It starts with the goal and it ends with the culture, even if this is rarely specified or admitted.

Startup people are quick. They understand what it takes to be promoted long before the retreat.

And when the startup is a bit less ruthless, they manage to slip some compassion in. They find a way to be nice somewhere inside the organization.

That is the goal of the employees. It is not necessarily the goal of the founder.

So one group can hold different goals, and the different goals become different values. The founder's list goes on the wall. The others live in the teams.

This may be why only about a quarter of employees say they believe the official values. The list answers to one goal in a room that has several. That is my guess, not a finding.

Two objections are strong enough to threaten the whole idea.

The first: same goal, different cultures. James Baron and Michael Hannan at Stanford followed nearly 200 Silicon Valley technology startups.

Same valley, same years, same goal. The founders still built five different kinds of company.

Founders' blueprints in the Stanford study. Rows are four of the five models; the columns give what holds people in the company and what the study found.

Blueprint	What holds people	What happened
Engineering	Challenging work, hired for skill, controlled by peers	The most common choice, nearly a third of founders
Commitment	Attachment to the company itself, hired for fit	Least likely to fail and fastest to go public
Star	Exceptional people given autonomy	Fastest growth in market value after going public
Autocracy	Money, and a boss who watches	Most likely to fail

Firms that later changed blueprint were about 2.3 times as likely to fail.

So the goal did not pick the culture. The founder did, from what he already believed before the company existed.

The choice still changed the result, which keeps culture a tool. But some founders picked a better tool than others for the same job, and none of them could swap it cheaply.

This is the Google point seen from the other side. The people and the moment write the culture as much as the goal does.

The second: the tool becomes the goal. Robert Merton saw it in bureaucracies in 1940. A rule starts as a means, people are trained to follow it, and it ends up valued for itself.

In his words, "an instrumental value becomes a terminal value".

Robert Michels saw it in the German Social Democrats in 1911. The party was built to change society. Its own growth overshadowed the programme, and the leaders put its survival first.

The Janissaries were recruited to depend on the sultan alone. They ended as a faction that made coups, and the sultan destroyed the corps in 1826.

Philip Selznick gave it a name in 1957. To institutionalize, he wrote, is to "infuse with value beyond the technical requirements of the task".

This one cuts deeper. I said organizations keep trying to make culture an absolute. Merton would say it is what normally happens.

If he is right, culture is a tool only at the start. Later it has its own weight, and it starts choosing the goals.

I don't have a clean answer. The organizations in my table that changed goal had to break something first: Rome a republic, Germany its own stormtroopers, Uber its founder.

The format has authors. In 1996 Jim Collins and Jerry Porras told companies to pick five to seven people and have them name three to five core values.

Sociologists can explain why it spread. Organizations facing uncertainty copy each other, and some practices are adopted as ceremony, to look legitimate.

I have one more theory. People need to believe in something, and they need to belong.

Psychologists treat the second need as basic. Roy Baumeister and Mark Leary argued in 1995 that the need to belong is a fundamental human motivation.

Maybe some of the people in the room are looking for an absolute, and they pour that longing into corporate buzzwords.

Or it is simpler. People need a sense of belonging, and a culture is the answer a company can offer to that need.

Either way the session would serve the people in the room as much as the goal of the company. It would also explain the second objection: a tool that answers a need like that does not stay a tool for long.

I wonder if in Italy they do it less because they saw Fantozzi, and they are very careful not to look like Fantozzi.

Ugo Fantozzi is an accountant invented by the Italian comedian Paolo Villaggio. The first book came out in 1971 and sold over a million copies. Ten films followed between 1975 and 1999.

Fantozzi works at the Megaditta, the mega-company. His colleagues drag him into company outings: bicycle races, football on muddy fields, tennis at six on a Sunday morning.

Nobody wants to be there. Everybody has to fit in somehow.

In the second film, from 1976, a boss who loves art cinema makes the whole staff sit through Battleship Potemkin at the company film club. Everyone pretends to love it.

Then Fantozzi stands up and says what he thinks of the film. The room gives him a standing ovation that lasts 92 minutes.

I have always seen Fantozzi as the fake side of these retreats. People have to fit somehow, and they end up doing the format Collins suggested.

Would Collins have suggested that format if he had seen Fantozzi? And would these retreats work if people had seen him?