Is Zwift Worth the Money? Run the Numbers on Your Own Hours

Every "is Zwift worth it?" article makes the same mistake: it quotes a price. Prices change, differ by country, and differ again with a hardware bundle. So this article contains none. It gives you the arithmetic instead, and you plug in whatever your own bank statement says.

The short version: the fee is almost never the deciding factor. Cost per hour is, and it is dominated by how many hours you ride. Below roughly eight hours a month the number looks bad whatever you pay. Above it, the real questions become hardware, adherence, and whether the platform buys you something the alternatives cannot.

Cost per hour, not cost per session

Dividing by sessions hides what you are buying: a 30-minute recovery spin and a three-hour endurance ride are one session each and are not the same purchase. Hours are the honest unit, because the subscription sells unlimited access to time and time is what you consume.

Use a climb as an anchor so hours are easy to count. One ascent of Alpe du Zwift - 12.2 km, 1,036 m of gain, roughly 8.5% average, 21 hairpins - takes about an hour at 3.2 W/kg. This site's fit, time_seconds = 148.60·(W/kg)² - 1954.08·(W/kg) + 8329.87, gives 59:58 at 3.2 W/kg, about 1:12:53 at 2.5 W/kg and about 48:11 at 4.0 W/kg. So "one Alpe a week" is roughly four hours a month, while Ven-Top, Zwift's Mont Ventoux at about 20.9 km and 1,534 m, is closer to 90 minutes an attempt. Check your own with the Alpe du Zwift calculator or the Ven-Top calculator.

Here is the shape of P ÷ H, as a fraction of one month's fee so it works in any currency.

Every row adds the same four hours and every row helps less than the one before. That is not an accident: the derivative of P ÷ H is -P ÷ H², so an extra hour is worth less in proportion to the square of the hours you already ride. Four to eight hours halves your cost per hour; twenty to twenty-four barely moves it. The financial decision lives in the first ten or twelve hours of the month. Past that you are optimising training, not money.

The same logic settles monthly versus annual without prices. Call R the annual price divided by one month's price. Annual wins if you use the platform for more than R months in the year: if R is ten, you need eleven months of riding to be ahead, and a rider who stops in May is not that person.

The rest of the cost stack, which is usually bigger

Amortise the trainer the same way. If it costs T and lasts L hours, the hardware line is T ÷ L per hour. The trainer then costs more per hour than the subscription whenever T > P × (L ÷ H), and L ÷ H is simply the trainer's lifetime in months. So the two lines are equal when the trainer's price equals its lifetime in months times the monthly fee: a trainer you keep for five years crosses over at sixty months of subscription. Price your trainer in "months of Zwift" and you know at once which line dominates your setup.

The small lines rarely matter. A trainer, a fan and a screen draw on the order of 100-250 W from the wall, so an hour is roughly 0.1-0.25 kWh - my rough estimate, not a measurement, but at any European rate it sits well below the per-hour fee. Consumables are sneakier: sweat eats bolts, cables and bearings faster than road riding does, as in towels during Zwifting.

The honest comparison set

A subscription is only expensive relative to something. The four real alternatives are a gym, a structured-training app, a coach, and riding outdoors. Each buys something the others do not.

The pair that overlaps least is Zwift and a coach: one is a place to train, the other is a decision about what to train. So the ranking depends on whether your problem is showing up or knowing what to do. Zwift is very good at the first and only adequate at the second; a coach is the reverse. If it is neither, and you simply want to be outside, no subscription fixes that - Zwift vs outdoor cycling makes that case.

Break-even, as a formula rather than a number

The first is usually decisive. If one drop-in studio class costs a third of a month's subscription and lasts an hour, then C = P ÷ 3 and H* = 3. Three hours a month - less than one ride a week - and the subscription wins on price alone. That is why "Zwift is expensive" rarely survives a paid alternative.

The second decides things for adults. What Zwift sells against outdoors is overhead: kit, lights, the ride to the edge of town, cleaning a winter bike. Call it 20 minutes a ride, conservative in a Danish February, and twelve outdoor rides a month is four hours reclaimed. Whether four hours is worth one monthly fee is your call, but for most people with a job it is not close. The same arithmetic drives having time to Zwift during university.

Where the value actually comes from

The winter-only rider gets the cleanest return, because the alternative is nothing and the difference shows up in the spring. The only mistake available is a pricing one: compare R against five months, not twelve. The year-round racer is buying a field of humans at 20:10 on a Tuesday, because a race extracts an effort a workout file cannot - but racing is easy to over-consume, so pair it with a training plan and an honest look at the racing category calculator.

For the time-crunched parent the overhead formula is the entire product: 45 minutes plus five of setup exists on a Tuesday night, while 45 minutes outdoors plus 20 of overhead plus daylight does not. That rider usually has the highest hours and the lowest cost per hour while paying what everyone else pays. The rider coming back from injury is buying control - ERG holds the ceiling a physio gave you and no hill pushes you through it, worth a lot for six weeks and much less afterwards. That is a good argument for switching the subscription on for a purpose rather than treating it as a permanent utility bill.

The behavioural economics of a flat fee

The scheduled group ride is a commitment device

A group ride starts at a fixed minute whether or not you are on the bike. That turns a vague intention into an appointment with a deadline, and deadlines produce behaviour far more reliably than intentions. With a stated pace and riders around you, dropping out also feels like leaving rather than stopping. None of it requires you to be competitive - only that the clock belongs to someone else.

The sunk-cost nudge

A sunk cost should not influence a decision, and behaviourally it always does. Having paid, you ride to avoid the feeling of waste - the same effect that makes people use a gym they joined in January. It is irrational and it works. It is also most likely to turn on you, because the feeling that gets you riding on a good day gets you riding on a day you should rest.

The badge treadmill

Levels, route badges, drops and unlockable equipment drip completable goals unrelated to fitness. Some are short, some absurdly long - the Zwift Concept Z1, the Tron bike, needs 50,000 m of climbing in the Everest Challenge, which is why I own one and why it took so long. The long ones work because they are long: a goal you are 60% through is a reason to ride tonight. The achievements guide lays out the ladder, and the streak mechanic gets the same treatment in keeping your streak while travelling.

The counterpoint: engineered adherence can collapse

Adherence built on external rewards lasts exactly as long as the reward supply. Unlock what you cared about, ride the badges, hit the level you were chasing, and the drip stops. If that was the reason you rode, the riding stops with it. When someone trains hard for two winters and does nothing in the third, the diagnosis is usually not laziness - the game finished and no other reason had grown underneath it.

There is a test. Cancel for one month, keep the trainer, and see whether you still ride four times a week with free software or none. If you do, the subscription buys convenience, which is fine. If you ride nothing, it is buying the behaviour itself - worth more, but know that is what you pay for, and be alert to the day the novelty runs out.

Alternatives worth knowing about

I have no affiliate deals with anyone, and this site's editorial policy says why. So here is the landscape in categories, because features and prices here go stale faster than any article can track.

• Other virtual worlds. Several platforms render a world and put riders in it, and they differ mainly in how many people are online at your hour, which matters more than graphics. At least one has run without a subscription fee - check whether that still holds.
• Real-video platforms. Footage of a real road with resistance matched to the real gradient. Better for riding somewhere specific, weaker on live humans.
• Structured-training apps. No world, no avatars, just a plan and a workout player. If your problem is not knowing what to do, this category solves it directly and Zwift only indirectly.
• The app your trainer came with. Most manufacturers ship a companion app that runs ERG workouts. For a rider whose need is "hold 220 W for eight minutes, six times", it is often sufficient and already paid for.
• No app at all. A trainer, a head unit showing power, and a workout on paper. Unglamorous, effective, and the correct control experiment for the cancel-for-a-month test.

None of them change the hardware cost. Buy the trainer for its own sake, then choose software annually.

When it is genuinely not worth it

• You ride under four hours a month. At H = 4 you pay a quarter of a month's fee per hour, and almost any pay-per-use option beats that. Commit to more hours or pay per session.
• You have a real riding season all year. If January is genuinely rideable outdoors, the overhead formula collapses and the main argument goes with it.
• Your goal needs handling, not watts. Criterium cornering, descending, gravel, pack positioning - none of it is trainable indoors. Zwift builds the engine, not the driver.
• You have no smart trainer and do not want one. Without controllable resistance and reliable power, much of what you pay for is unavailable and the money belongs in hardware first. The setup guide covers what is actually required.

How I decide, personally

I live in Aalborg. From November to March the choice is not "Zwift or a nice ride outside", it is "Zwift or nothing", and while I build toward Ironman Copenhagen the bike volume has to land somewhere. At the hours I ride, the fee is a small fraction of the per-hour cost of the setup, so it stopped being a decision several winters ago. In summer it is much closer, and I would not defend an annual rate for months I spend outdoors.

The rest of the numbers-first material lives in the data analysis category, the FAQ answers the short questions, and you can read who is writing this.