← Articles
Read in another language
Lifestyle

The Run Club Boom: What the Numbers Really Show

Jayden

Maintains the wage calculators and public-data regional information at 생활데이터랩, and analyzes technology, industry, and policy issues.

Published

Key points

  • The London Marathon public ballot set a world record three years running: 840,318 applications for the 2025 race, 1,133,813 for 2026 — the first marathon ballot ever to clear a million — and 1,338,544 for 2027.
  • US registration data confirms the boom is broad-based, not one city's story: across the 2025 Top 100 largest road races the average race grew 6.5% year over year and was up 5.2% against 2019, with more than half now past their pre-pandemic size.
  • The social side of the story — run clubs as a 'third place', even as a dating scene — rests almost entirely on Strava's company-stated survey: 59% club growth in 2024, 58% of surveyed users making new friends, roughly one in five Gen Z users dating someone met through exercise.
  • The health case is an association, not a proof: a meta-analysis pooling 14 studies and 232,149 people linked running to 27% lower all-cause, 30% lower cardiovascular and 23% lower cancer mortality, but the design is observational and self-selection is unresolved.
  • Counterweights belong in the same frame: overuse injury risk for beginners scaling up too fast, and the fact that a group built around a demanding physical activity is not automatically inclusive across pace, ability, disability and schedule.

In May 2026, London Marathon organizers announced that a world-record 1,338,544 people had applied for a place in the 2027 race — the third straight year the ballot broke its own record, and up from 1,133,813 the year before and 840,318 the year before that [source: London Marathon Events, 2026]. That is not a typo. In two years the demand to run one big-city marathon roughly doubled, and the single largest bloc of British applicants was women in their twenties [source: London Marathon Events, 2026]. Somewhere along the way, a solitary sport known for early alarms and sore knees turned into one of the most sought-after tickets in the world.

You have probably felt the shift even if you have never entered a ballot. Parks fill with matching-shirt packs on Saturday mornings. A friend who used to text about brunch now texts about a "long run." Social feeds carry a tidy slogan — run clubs over night clubs — that captures a real change in how a lot of people, especially younger ones, are choosing to spend their time. This piece separates two things that are easy to blur: the part of the running boom you can measure, and the part you can only feel. Both are real. They are not the same, and telling them apart is the whole point.

Why this is suddenly everywhere

Running never disappeared, but a few forces converged to push it back to the center of culture. The pandemic emptied gyms and closed indoor venues, and running — free, outdoors, and impossible to cancel — absorbed a wave of newcomers who never fully left. As normal life resumed, many of them wanted the outdoor habit and the company, and the informal run club filled both needs at once.

At the same time, running became legible on your phone. Fitness platforms turned a private jog into something you could log, map, share, and organize a group around, and the data shows the social side growing fastest of all. Add a low barrier to entry — you need shoes and a sidewalk, not a membership or a court — and a generation looking for cheaper, healthier, less alcohol-centered ways to meet people, and you get a habit that spreads through friend groups rather than advertising. The result is a trend that shows up simultaneously in marathon ballots, race registrations, app statistics, and shoe sales — which is exactly why it is worth measuring rather than just vibing about.

The numbers behind the boom

Start with the hard counts, because they are where the boom is least arguable. The London ballot is the loudest signal: 840,318 applications for 2025, 1,133,813 for 2026 — the first marathon ballot ever to clear a million — and 1,338,544 for 2027, each a world record in turn [source: London Marathon Events, 2025][source: London Marathon Events, 2026]. Ballot entries measure demand, not finishers, and one race is not the world. But the direction is unambiguous, and it is not just a London story.

In the United States, the industry association Running USA reported that participation across the nation's top races rose in 2025 versus the prior year, with growth across every major distance — 5K, 10K, half marathon, and marathon [source: Running USA, 2026]. Registration data backs that up: across the 2025 Top 100 largest US road races, the average race grew 6.5% year over year and was up 5.2% compared with 2019, and more than half of those races have now surpassed their pre-pandemic size — led by the New York City and Chicago marathons [source: RunSignup, 2026]. After years of describing running as "still recovering," the big-race economy has finally climbed back above where it was before 2020.

The money follows the feet. The market-research firm Circana reported that performance running shoes led US footwear-industry growth in 2025, standing out as one of the strongest categories even as the overall footwear market stayed roughly flat [source: Circana, 2026]. None of these numbers, on its own, proves a cultural movement. Together — demand, participation, and spending all rising at once — they make a strong measured case that more people are actually running, not just talking about it.

More than a workout: the run club as a "third place"

Here the story shifts from measurement to meaning, and the evidence changes character. The most-quoted numbers about run clubs as a social scene come from Strava, the fitness platform, whose 2024 Year in Sport report found that running clubs on its service grew 59% in a single year, making running the fastest-growing social sport on the platform [source: Strava, 2024]. Strava also reported that group activities ran about 40% longer than solo ones, and — more strikingly — that 58% of surveyed users said they had made new friends through a run club or fitness group [source: Strava, 2024].

The relationship angle is what turned run clubs into a media story. In Strava's survey, roughly one in five Gen Z users said they had gone on a date with someone they met through exercise, and Gen Z respondents were four times more likely to say they would rather meet people through a workout than at a bar [source: Strava, 2024]. That is the measured kernel inside the "run clubs are the new dating apps" headline: not proof that running has replaced bars, but real self-reported evidence that a meaningful slice of young people now treats the group run as a place to meet others.

Two cautions keep this honest. First, this is company-stated data: Strava's users skew toward the engaged, urban, and already-active, and survey answers are self-reported, so the figures describe Strava's community, not the whole population [source: Strava, 2024]. Second, "third place" — the sociologist's term for a social anchor that is neither home nor work — is a framing, not a measurement. The feeling that run clubs are replacing bars is widespread and worth taking seriously; it is simply a different kind of claim than "59% growth," and the two should not be quoted as if they were equally hard.

Why Gen Z and millennials are lacing up

Beneath the numbers is a fairly intuitive set of motivations, and they help explain why the boom skews young. The first is connection. After years in which remote work and screens thinned out casual, in-person contact, a standing weekly run offers something scarce: a recurring, low-pressure way to see the same faces without scheduling a formal hangout. You show up, you run, you talk — and the activity does the social work that small talk usually has to.

The second is structure. A run club supplies what solo motivation often cannot — a set time, a set place, and other people who will notice if you skip. The third is cost. In an expensive decade, running is close to free, and a club adds companionship without a bar tab or a membership fee. The fourth is wellness, broadly defined: many younger runners describe drinking less and seeking daytime, sweat-based socializing over late nights out, a shift Strava framed with its own tidy line about workouts no longer being about burning out [source: Strava, 2024]. None of these motivations is new. What is new is how neatly running satisfies all four at once, which is why it spreads through a friend group the moment one person starts.

The health and connection dividend — with caveats

It would be easy to end on an unqualified cheer for running, and easy to overstate it. The honest version is that the health benefits are real but should be described as associations, not guarantees. A large systematic review pooling 14 studies and more than 232,000 people found that running was associated with a 27% lower risk of death from any cause, a 30% lower risk of cardiovascular death, and a 23% lower risk of cancer death — and that even modest amounts, as little as once a week, were linked to benefit [source: British Journal of Sports Medicine, 2020].

Read that finding precisely. It is observational, which means it shows a correlation, not proof of cause: people who run may differ from people who do not in ways — diet, income, baseline health — that also affect how long they live. Running very likely helps, and the biological case for regular aerobic activity is strong, but "runners live longer" is a weaker claim than "running makes you live longer," and the study supports the first more firmly than the second. The social benefits carry the same asterisk. It is plausible, and consistent with the self-reported data, that the companionship of a run club is good for people; it is not the same as a controlled demonstration that joining one improves your health or your loneliness.

Two more counterweights belong here. Running carries genuine injury risk — overuse injuries are common when beginners scale up too fast — so the boom will produce sore knees as well as new friendships. And a group built around a demanding physical activity is not automatically inclusive: pace, ability, disability, and schedule all shape who can actually take part. The dividend is real; it is also uneven, and worth claiming carefully rather than loudly.

Reading the boom honestly

So is this a durable shift or a passing fad? The measured evidence points to something more than a fad: you do not get three consecutive marathon-ballot world records, a broad-based return above pre-pandemic race participation, and running leading footwear growth all at once from a fleeting trend [source: London Marathon Events, 2026][source: RunSignup, 2026][source: Circana, 2026]. Those are lagging, hard-to-fake indicators of behavior, not just sentiment.

At the same time, the flashiest part of the narrative — run clubs as a dating and nightlife replacement — rests mostly on one platform's self-reported survey, and platform data can run ahead of the wider population [source: Strava, 2024]. It is also worth being precise about what this boom is not: it is a specific story about running community and participation, distinct from broader debates about social isolation, from other social-sport surges like racket sports, and from the "movement snacks" idea of squeezing small bits of activity into a sedentary day. Lumping them together makes each blurrier. The running boom deserves to be judged on its own, unusually solid, evidence.

What to watch

The smart posture is neither hype nor dismissal. Running has, by several independent measures, genuinely surged, and a large number of people — younger women especially — are choosing the group run as a place to exercise, belong, and meet others. The measurement is strong; the meaning is still being written.

Three things will show whether the boom sets or fades. First, the registration numbers: do 2027–2028 marathon ballots and race counts keep climbing, plateau, or slip once the novelty cools? Second, whether the social scene shows up beyond a single app's survey — in independent research, in club membership that outlasts the initial buzz, in whether today's new runners are still lacing up in three years. Third, whether the infrastructure keeps pace: more races, safer routes, and clubs that stay welcoming as they scale. Watch the counts, hold the vibe lightly, and the run club boom will tell you what it really is.

Charts

London Marathon public ballot applications, three consecutive world records

London Marathon public ballot applications, three consecutive world records2025 race 840,318, 2026 race 1,133,813, 2027 race 1,338,544840,3182025 race1,133,8132026 race1,338,5442027 race
Ballot applications — a measure of demand, not of finishers. All three counts come from London Marathon Events, in one unit and on one basis; the 840,318 and 1,133,813 figures were published in the May 2025 release and the 1,338,544 figure in the May 2026 release, so no single link would cite all three correctly. This is the only set of figures in this article that shares one producer, one unit and one measurement basis.

Timeline

  1. A systematic review and meta-analysis in the British Journal of Sports Medicine pools 14 studies and 232,149 participants and reports running participation associated with 27% lower all-cause, 30% lower cardiovascular and 23% lower cancer mortality — observational, so association rather than proof of cause.

    British Journal of Sports Medicine (Pedisic et al.) (opens in a new tab)
  2. Strava's Year in Sport report states that running clubs on its platform grew 59% in 2024, that group activities ran about 40% longer than solo ones, and that 58% of surveyed users made new friends through a run club or fitness group (company-stated, survey self-report).

    Strava (opens in a new tab)
  3. London Marathon Events announces 1,133,813 applications for the 2026 race — the first marathon ballot ever to exceed one million, breaking the previous record of 840,318 set for the 2025 race.

    London Marathon Events (opens in a new tab)
  4. RunSignup's 2025 Top 100 largest US road races data shows the average race up 6.5% year over year and up 5.2% against 2019, with more than half of those races now above their pre-pandemic participation.

    RunSignup / Running USA (opens in a new tab)
  5. Running USA reports that participation across the nation's top races rose in 2025 versus the prior year, with growth across every major distance — 5K, 10K, half marathon and marathon (association-stated summary; the full numeric report is paid).

    Running USA (opens in a new tab)
  6. Circana reports that performance running shoes led US footwear-industry growth in 2025, standing out as one of the strongest categories while the overall footwear market stayed roughly flat (market-research-firm data, cited directionally).

    Circana (opens in a new tab)
  7. London Marathon Events announces a world-record 1,338,544 applications for the 2027 race — the third straight annual ballot record — with women in their twenties the single largest bloc of British applicants.

    London Marathon Events (opens in a new tab)

Analysis

The measured boom and the felt boom are two different claims

Ballot records, registration counts and shoe sales are hard, lagging indicators of behaviour. 'Run clubs over night clubs' is a cultural reading of that behaviour. Both can be true at once, but only the first has been counted, and quoting them at the same confidence is the single most common error in coverage of this trend.

A ballot counts wanting, not running

1,338,544 applications is a demand figure. It does not say how many people ran, trained, or even own shoes — the London field is a fraction of that number. Demand rising is genuinely significant, but it is a different measurement from participation, which is why the US registration data matters as a separate check rather than a repetition.

The strongest evidence is the boring evidence

The most persuasive part of this story is not the 59% or the dating statistic; it is that more than half of the 2025 Top 100 US road races are now larger than they were in 2019. That is a registrar's count, hard to inflate and slow to move, and it is what turns 'recovering' into 'past pre-pandemic levels'.

Company-stated is not the same as wrong — it is differently scoped

Strava's numbers describe Strava's community: engaged, urban, already active, answering a voluntary survey. That is a real population and a real signal about it. It is simply not a population-level measurement, and the article treats it as evidence about run-club culture on one platform rather than about society.

Two percentages from the same source can answer different questions

RunSignup's 6.5% and 5.2% are both average per-race growth, but one is measured against 2024 and the other against 2019. They describe momentum and recovery respectively — related, not interchangeable — so they are reported side by side in a table here rather than plotted as if they sat on one axis.

Association is the honest word for the health finding

The meta-analysis is observational: runners may differ from non-runners in diet, income and baseline health, all of which also affect mortality. 'Runners live longer' is supported; 'running makes you live longer' is a stronger claim the design cannot carry. The same asterisk applies to the social benefits, which are self-reported.

The dividend is real and uneven at the same time

Overuse injuries are common when beginners scale up too fast, and a group organised around a demanding physical activity is not automatically open — pace, ability, disability and schedule all decide who can actually show up. A boom story that omits this is describing only the people it worked for.

Why this article has only one chart

Nearly every other figure fails at least one honesty test. The mortality reductions of 27%, 30% and 23% are three different endpoints, so they do not share an axis. RunSignup's 6.5% and 5.2% use different baseline years. Strava's 59%, 58%, 40% and the one-in-five dating figure measure four different things, and the 52% walking-club figure was left out of the article's body altogether. 'More than half' of the Top 100 is a floor, not a value. Only the three London ballot counts share one producer, one unit and one basis — so they are the only series plotted.

Comparison

The signals behind the boom — what each one actually measures, and what kind of evidence it is
SignalWhat it measuresEvidence tier
London Marathon public ballot: 840,318 → 1,133,813 → 1,338,544Applications for a place — demand for one raceOrganizer-stated hard count
2025 Top 100 largest US road races: average race +6.5% YoY, +5.2% vs 2019Registrations actually taken at the largest US racesRegistration-platform data (hard counts)
Running USA: growth across 5K, 10K, half marathon and marathon in 2025Direction of participation across distancesIndustry-association summary (qualitative; full report paid)
Circana: performance running led US footwear growth in 2025Consumer spending on running shoesMarket-research-firm data, cited directionally
Strava: running clubs +59% in 2024, 58% made new friends, ~1 in 5 Gen Z dated someone met exercisingActivity and attitudes among Strava usersCompany-stated platform data and survey self-report
British Journal of Sports Medicine: 27% / 30% / 23% lower mortalityAssociation between running and death ratesPeer-reviewed but observational — correlation, not causation
Same trend, four kinds of speaker — who says what, and what type of statement it is
ClaimWho says itType of statement
A world-record 1,338,544 people applied for the 2027 raceLondon Marathon Events (organizer)Counted registration figure
More than half of the 2025 Top 100 races are now above their 2019 sizeRunSignup, via Running USAPlatform-measured count, stated as a floor
Running clubs on the platform grew 59% in a single yearStrava (company)Company-stated platform metric
58% made new friends through a run club; ~1 in 5 Gen Z dated someone met exercisingStrava survey of active peopleSelf-reported survey response
Performance running shoes led US footwear growth in 2025Circana (market-research firm)Firm's market reading, used directionally
Running is associated with 27% lower all-cause mortalityPedisic et al., British Journal of Sports MedicinePeer-reviewed observational association
Run clubs are becoming a 'third place' and replacing barsWidely held cultural readingPerception / framing — not a measurement
The other figures in this article, exactly as published — and why none of them is on a chart axis
FigureAs publishedWhy it is not on an axis
Mortality reductions27% all-cause, 30% cardiovascular, 23% cancerThree different endpoints from one meta-analysis — a shared axis would imply a shared basis
US race growth+6.5% year over year, +5.2% versus 2019Same producer and unit, but two different baseline years
Strava club and social figuresClubs +59%, group runs ~40% longer, 58% made friends, ~1 in 5 Gen Z datedFour different measurements, and the one comparable pair (walking clubs +52%) was left out of the article body
Share of Top 100 races above 2019More than halfA floor, not a value
Meta-analysis scale14 studies, 232,149 participantsStudy scope, not a comparable series
Strava's own reach135M+ people, 190+ countries, survey of ~5,000Floors and an approximation describing the sample, not the trend

Process

  1. Ask what unit is being counted

    Applications, registrations, finishers and app users are four different things. The London ballot counts the first; the Top 100 data counts the second. Conflating them is how a demand story becomes an unsupported participation story.

  2. Identify the producer and what they can see

    An organizer sees its own ballot. A registrar sees its own races. A platform sees its own users. A market-research firm reads retail. None of them sees the population, and each figure inherits the limits of its vantage point.

  3. Check the baseline year

    'Up 6.5%' and 'up 5.2%' sound like the same claim until you notice one is measured against last year and the other against 2019. Growth figures without a stated baseline are unreadable.

  4. Separate logged activity from survey answers

    Club counts are recorded by the platform; 'made new friends' and 'went on a date' are answers people gave about themselves. Both can appear in one press release and still belong in different evidence tiers.

  5. Keep association language intact

    The mortality study says 'associated with'. Any retelling that upgrades it to 'running cuts your risk by 27%' has added a causal claim the observational design does not support.

  6. Look for the counterweight before concluding

    Injury risk, uneven inclusion and survivorship all cut against the cheerful reading. A boom account that cannot name its own downside is a marketing summary, not an assessment.

Sources

  1. London Marathon Events — Record 1.33 million people apply for 2027 TCS London Marathon (2026-05).View source (opens in a new tab)
  2. London Marathon Events — New ballot world record for 2026 TCS London Marathon (2025-05).View source (opens in a new tab)
  3. Strava — Strava Releases Annual Year in Sport Trend Report (2024-12-04).View source (opens in a new tab)
  4. Running USA — Running USA Releases "2025's Top Races: A Look at the Year in Review" Industry Insights Report (2026-01-20).View source (opens in a new tab)
  5. RunSignup / Running USA — RunSignup Releases 2025 Top 100 Largest Road Races (2026-01-12).View source (opens in a new tab)
  6. Circana — US Footwear Industry: performance running leads 2025 growth (2026).View source (opens in a new tab)
  7. British Journal of Sports Medicine — Is running associated with a lower risk of all-cause, cardiovascular and cancer mortality? A systematic review and meta-analysis (Pedisic et al.) (2020).View source (opens in a new tab)

Tags

  • #run-club
  • #running-boom
  • #marathon
  • #gen-z
  • #third-place
  • #community