Data, Declaration, Dance

On the morning of 2 October 2004, thirteen runners gathered in Bushy Park, south-west London. Paul Sinton-Hewitt, sidelined from his running club by injury, borrowed a stopwatch and started the clock. A few friends wrote down the times. Each finisher collected a small metal washer, stamped with a number that recorded a finishing position and nothing more.

That washer holds this entire argument inside one stamped disc. The number on it recorded a fact about one body on one Saturday. Call that data. Sinton-Hewitt ranked the times and published the list, which turned thirteen loose facts into a shape that pointed forward. Call that a declaration. Then he asked everyone to come back the following Saturday at nine o'clock. Call that the dance.

Twenty-one years later, parkrun stages more than 2,300 free five-kilometre events every Saturday morning across more than twenty countries. On 4 April 2026 the charity counted 438,526 finishers worldwide, its highest single day on record. One man with a borrowed stopwatch grew that institution out of three moves and no capital. He counted. He published. He fixed an hour.

I name those three moves Story, Smartbook, and Socializing. In a boardroom I call them Message, Media, and Members: Core Message, Creative Media, Cooperative Memberships, which I shorten to CM3. The vocabulary shifts with the room. The sequence holds. Our decade speaks the dialect of analytics, so I now translate the three into words a founder, a marketer, and a banker already respect. Data. Declaration. Dance.

01Data pays, and someone else holds the invoice

Count the crude first. IDC, working with Seagate, put the world's annual output of created and copied data at 175 zettabytes for 2025, up from 33 zettabytes in 2018. A zettabyte holds a trillion gigabytes. Nobody can picture that pile, so skip the pile and follow the money.

In 2006, at an advertising conference in the United States, the British mathematician Clive Humby, the man who designed the Tesco Clubcard, called data the new oil. Commentators kept his slogan and dropped his point. Crude in the ground pays nobody, Humby argued. Refining pays. A barrel earns its keep after someone cracks it into petrol, plastic, and chemicals, and a database earns its keep after someone breaks it apart and reads it.

Corporations learned the refining trade. Meta reported family average revenue per person of $16.56 for the fourth quarter of 2025, drawn across 3.58 billion daily users of Facebook, Instagram, WhatsApp and Messenger. Reddit went further and sold the raw material outright. Reuters reported in February 2024 that Google agreed to pay roughly $60 million a year to train its models on Reddit content. Reddit booked about $130 million in data licensing revenue during 2024, close to a tenth of its total. Volunteer moderators kept those forums running. Members wrote every sentence. The cheque cleared in San Francisco.

Read those numbers together and the crack opens. Ordinary people generate crude at a scale no oil field ever matched, and almost none of them own a refinery. A life leaks data into other people's graphs all day long. Nobody plots the graph that points back at the life.

02A graph turns crude into shape

On 16 May 2012, Amit Singhal, then head of search engineering at Google, published a short post announcing the Knowledge Graph. He compressed the shift into three words that reorganized an industry: "things, not strings." Search matched text against text for four decades. The new model linked real objects to one another instead. At launch the graph carried more than 500 million objects and more than 3.5 billion facts about the relationships among them.

Google graphed the planet's things. Fourteen years later, almost nobody graphs a person. Your bank keeps a decade of your spending. Your phone keeps your sleep, your steps, your locations, your late-night reading. Each vault plots its own curve for its own purposes, and the curves never meet. The human at the centre of all that measurement walks around with no picture of the shape their own life traces.

A declaration starts where the plotting ends. Gather the points, draw the line, read the direction it travels, up or flat or down. Then write the next stretch of that line in the future tense, as a commitment rather than as a forecast. Steve Zaffron and Dave Logan put the principle into one sentence in 2009: "Future-based language transforms how situations occur to people."

Consultants built that framework out of client practice, not out of controlled trials, so I hold it against harder evidence. Psychology supplies the floor. Paschal Sheeran's review of the research found that intentions predict only about 28 percent of the variance in behaviour. Resolutions carry a light load. Peter Gollwitzer and Sheeran then pooled 94 independent tests covering more than 8,000 participants. They measured what changes when a person converts that intention into an if-then plan naming a cue, an hour, and a place. Goal attainment rose by a medium-to-large margin, d = 0.65.

I resisted that discipline for thirty years. Journalism trains the opposite reflex. A reporter who writes the future writes fiction, so I made my living inside the past tense: what happened, who confirmed it, when the record closed. That reflex protected my copy and shrank my life. I could file a clean account of yesterday, and I carried no sentence about my own next decade. Qualped grew out of the correction.

03The wall stands at nine o'clock on Saturday

Watch what happens between the second move and the third. A team led by Lindsey Reece examined every parkrun registration across the United Kingdom, Australia and Ireland from 2017 to 2019. That set held 680,255 people who typed their details, ticked a box, and printed a barcode. Of that group, 293,542 never attended a single event. Another 147,148 attended once and stopped.

The follow-up survey named the leading obstacle, and the answer humbles the entire motivation industry. Participants cited an inconvenient start time more often than any other barrier. Women more often than men cited a second one: a fear of arriving too unfit for the crowd.

A start time beat the willpower gurus. The wall between a declaration and a dance rarely rises to the height of character. It sits at the level of a Saturday morning: the hour, the distance, the door, the first face at the entrance that greets a newcomer by name.

04Synchrony works even when enthusiasm quits

Scott Wiltermuth and Chip Heath ran three experiments at Stanford and published the results in Psychological Science in 2009. Some participants walked around campus in step with one another. Others held plastic cups, listened to O Canada, and sang the words while moving the cups in time with the music. Control groups walked and sang out of time, each person to a different tempo. Every group then played economic games with real money at stake.

The groups that moved in time cooperated more in the games that followed, "even in situations requiring personal sacrifice." The authors then tested the comfortable explanation and knocked it down. Happy feelings did not drive the effect. Bodies moving on a shared beat produced cooperation whether or not the people enjoyed themselves.

Set that finding beside the parkrun numbers, and the third move stops looking like a reward for finishing the first two. Synchrony does work of its own. A group that moves at the same hour manufactures the cooperation that carries a plan, and the mood of the members hardly enters the calculation. This overturns the order most people assume. We treat community as the prize we collect after we sort ourselves out. Community does the sorting.

05One country ran the whole sequence for twenty years

Iceland supplies the largest documented run of the three moves. Through the 1990s, Icelandic teenagers drank harder than almost any cohort in Europe. Researchers in Reykjavik, led by Inga Dóra Sigfúsdóttir alongside colleagues including Álfgeir Logi Kristjánsson and John P. Allegrante, designed a prevention model around population data rather than around lectures and slogans.

The data move: every February, schools survey the country's ten-to-sixteen-year-olds about drinking, smoking, sport, sleep, hours with parents, and friendship. The declaration move: analysts hand the results back to each municipality within weeks, and each community writes a plan from its own numbers instead of from a national average. The dance move: parents sign shared agreements on curfews and unsupervised parties, councils fund sport and music for every child, and a whole town keeps step for a decade.

The numbers moved with them. Among fifteen- and sixteen-year-olds, the share who reported drunkenness during the previous thirty days fell from 42 percent in 1998 to 5 percent in 2016. Daily smoking dropped from 23 percent to 3 percent. Cannabis use fell from 17 percent to 7 percent.

Read one caution beside those figures. The European Union Drugs Agency, reviewing the model for its prevention registry, records that evaluators cannot yet separate the effect of each component. Iceland also enforced a strict alcohol policy across the same years. A country changes many things at once, as countries do. The shape of the sequence still stands out: a survey every year, a local plan written in the future tense, and twenty years of neighbours moving together.

06Size your tribe by attendance, never by folklore

Founders love to quote a number for the size of a tribe. Robin Dunbar extrapolated 150 from a regression line linking neocortex size to group size among primates, and the figure travelled the world, through business books and office redesigns. In 2021, Patrik Lindenfors, Andreas Wartel and Johan Lind rebuilt that calculation in Biology Letters using modern statistical methods. One approach produced average group sizes between 69 and 109, another between 16 and 42, with 95 percent confidence intervals running from 4 to 520 and from 2 to 336. No single number survives a spread that wide.

Drop the folklore and count attendance instead. A core community reports its true size every week, in a room, at an hour, through a door. Twelve people who walk back in each Saturday outrank twelve thousand followers who scroll past a post.

The stakes reach well past marketing. In May 2023 the Office of the U.S. Surgeon General published an advisory on loneliness and isolation. About half of American adults describe loneliness in their own lives, the advisory reported. Weak social connection raises the risk of heart disease by 29 percent and of stroke by 32 percent. The mortality burden of chronic disconnection rivals smoking up to fifteen cigarettes a day. A dedicated core community performs medical work.

The upside shows on the other side of the ledger. When Helen Quirk and colleagues surveyed UK parkrun participants for BMC Public Health in 2021, 60,000 people answered. Among them, 90.7 percent credited the event with a stronger sense of personal achievement, 78.8 percent with more happiness, and 69.3 percent with better mental health. Self-selected surveys flatter their subject, and only 2.7 percent of the people contacted replied, so weigh those percentages as testimony rather than as proof. Testimony from 60,000 people still carries weight.

07The Saturday you can author

Picture a Saturday two years from now, and read the scene as invention rather than as report. A woman in Scarborough sits at her kitchen table at seven in the morning, coffee cooling beside the laptop. She exports a year of her own numbers: hours worked, dollars earned, kilometres walked, books finished, nights of broken sleep, names of people she met more than twice. QwaiAI draws the line in ninety seconds and shows her the direction: flat income, falling sleep, three friendships thinning toward nothing. She writes four hundred words in the future tense and names the wall inside the same page. The chapter lands in her simsbook. At ten o'clock she unlocks a rented room above a bakery on Lawrence Avenue, and eleven people walk through the door. They meet at that same hour every week. By March the room holds forty, and two members open businesses of their own.

I authored this essay with QwaiAI running as cognitive executive. The machine gathered the studies, held the sources, checked my claims, and drew the argument into shape. I chose the thesis, carried the stake, and signed the work. Keep that division clean when you build your own. A machine can graph your data, draft your declaration, and remind you of the hour. No machine walks into the room for you. Attendance stays human, and attendance decides the outcome.

So run the sequence yourself, this month, at the smallest workable scale:

1. Collect thirty days of your own data. Hours, dollars, words written, kilometres, sleep, and the name of every person you met more than once.
2. Plot it. Draw the line on paper or let QwaiAI draw it, then read the direction out loud to one other person.
3. Write four hundred words about the next stretch of that line in the future tense, and name the wall on the same page.
4. Convert each promise into an if-then sentence carrying a cue, an hour, and a place.
5. Send the page to twelve people, fix a weekly hour, open a door, and count who arrives. Repeat the following week.

Sinton-Hewitt started with a borrowed stopwatch, a stamped washer, and one fixed hour. He counted thirteen people. He published the list. He asked them to come back, and he promised to stand at that start line every week for the rest of his life. Close to four hundred thousand people now walk to a start line every weekend because one man completed all three moves instead of two.

Gather your numbers this week. Draw the line. Write the next stretch in the future tense, name the hour, and open the door. Then stand at the start line and count the people who came.

Shaun Michael Samaroo ◆ Ontario, Canada