What no one wants to admit about Zwift: the platform that rescued indoor training from being the most miserable hour of your week might also be the reason your FTP hasn't moved since last winter.
That's not a knock on Zwift. I mean it. Before Zwift, turbo sessions were you, a back wheel clamped into a dumb trainer, staring at a garage wall while your sweat pooled on a towel below. Nobody misses that. Zwift turned the indoor trainer from an instrument of suffering into something people actually looked forward to, and that alone makes it one of the most important things to happen to amateur cycling in the last decade. But something else happened along the way. The same features that make Zwift brilliant at keeping you on the bike — the races, the segments, the group rides where everyone surges on the climbs — started replacing training itself. And that's where the wheels come off.
You know the rider. Maybe you are the rider. Zwift race on Tuesday evening. Another on Thursday. A group ride on Saturday that turns into a race within the first five minutes. Three hard efforts a week, every week, from October through March, all of them threshold-to-VO2max, none of them planned, none of them periodised. And come April, the outdoor numbers are exactly where they were a year ago. Same effort. Same fatigue. Same plateau.
The intensity trap, amplified
Here's where the science comes in, and let me be really clear about this: the problem with Zwift racing as your primary training isn't that it's bad exercise. It's that every Zwift race is essentially the same physiological stimulus. The gun goes, you settle at threshold, the climbs pull you into VO2max territory, and you finish somewhere between wrecked and very wrecked. There's no version of a Zwift race that's easy. There's no version that targets a specific energy system in isolation. There's no recovery built into the format. It's a hard effort, every single time.
Professor Stephen Seiler has spent decades studying intensity distribution in endurance athletes, and the finding that keeps emerging from his research is that the best performers across cycling, rowing, cross-country skiing, and running converge on roughly the same pattern: about 80 percent of their training at low intensity, about 20 percent at high intensity, and very little in between. When I had Seiler on the podcast, he talked about how this distribution isn't a coaching philosophy — it's an observation about what actually produces long-term aerobic development. The athletes who get the best results, year after year, protect their easy days ferociously and make their hard days count.
Now picture the Zwift racer doing three races a week. That's three hard sessions. For a rider training six or seven sessions a week, that's already 40 to 50 percent of their week at high intensity — more than double what Seiler's research suggests. And the rider training four or five times a week? They're pushing 60 to 75 percent intensity. The distribution is completely inverted. The gamification that keeps them coming back is the very thing pulling them away from what works.
This is the same trap Christian Schrot described when he talked about why time-crunched amateurs train too hard. Limited time creates a psychological pressure that makes easy riding feel wasteful, so riders default to hard. Zwift takes that existing pressure and turbocharges it with leaderboards, race results, and the social proof of other riders going full gas around you. The trap doesn't just close — it locks.
When Zwift actually helps
This isn't a case against Zwift. It's a case for using Zwift deliberately, the way you'd use any training tool — with a purpose and a plan.
Zwift is remarkably good at a few things that no structured training platform replicates. The first is social accountability. When your mate is expecting you on a group ride at seven in the morning, you clip in. When it's just you and a workout file, the snooze button wins more often than it should. That social pull is real and it matters, especially during the dark months when motivation is at its lowest.
The second is Zone 2 group rides, and this one surprises people. A well-organised Zwift endurance ride, where the group commits to staying at a steady pace, can be brilliant for low-intensity work. You've got company, you've got the visual stimulus of the virtual road, and you've got a pace that's set by the group rather than by your own tendency to push. The temptation to surge is lower when everyone around you is holding steady. It's not perfect — there's always someone who attacks on the KOM — but it's more engaging than staring at a flat power line on a blank screen for ninety minutes.
The third is workout mode itself. Zwift's ERG mode does exactly what any structured platform does: it holds the resistance at the prescribed power regardless of cadence, so you hit the target whether the virtual road goes up or down. If you load a structured workout into Zwift and ride it in ERG mode, you're getting the same physiological stimulus as you would on TrainerRoad or Wahoo SYSTM or any other platform. The workout doesn't care about the scenery behind it.
Where Zwift helps is anywhere the social element or the visual engagement keeps you doing the work you were supposed to do. Where it hurts is anywhere the gamification pulls you into work you were not supposed to do.
When Zwift wrecks your plan
Here's where it gets really interesting, because the damage isn't always obvious. The rider who races on Zwift three nights a week knows those are hard sessions. What they often miss is that the "easy" rides on Zwift are also too hard.
Free-riding on Zwift is almost never genuine Zone 2. You pass another rider on a climb and your power spikes. You see someone ahead on a flat and you close the gap. The orange jersey for the fastest time on a segment is right there, and it only takes thirty seconds of effort. Each of those surges is small on its own, but they add up to a ride that sits in the grey zone — too hard to be recovery, too broken to be a real interval session. Dan Lorang, who coached Jan Frodeno and has spoken extensively about session purpose, would call this a session without intent. It burned matches but didn't buy anything specific.
The gamification is the culprit. Every design choice in Zwift — the XP system, the segment timers, the draft mechanics, the power-ups — exists to create engagement. And engagement, in Zwift's world, means effort. The platform is not trying to make you train well. It's trying to make you train more, and more often, and more intensely, because that's what keeps you subscribed. That's not cynical; it's how the business model works. But it means the incentives of the platform and the incentives of your training plan are pointing in opposite directions.
Tim Kerrison, who built the training system at Team Sky and later Ineos Grenadiers, was meticulous about session design. Every session had a purpose, every interval had a target, and the spaces between hard efforts were protected as carefully as the efforts themselves. The recovery rides were properly easy, because Kerrison understood that the adaptation from Tuesday's hard session happens during Wednesday's easy spin, not during Wednesday's accidental race. Zwift makes Wednesday's accidental race incredibly easy to stumble into.
The hybrid approach that actually works
The fix isn't complicated, and the good news is you don't have to choose. The riders I see getting the best results from indoor training are the ones who treat Zwift as one tool among several, not as the whole toolbox.
The framework looks something like this. You keep Zwift for one race per week — maybe two during a build phase — and treat that as a genuine hard session. It counts. It slots into your week alongside your other hard effort, whether that's a sweet spot session or a VO2max block. You might also keep a Zwift group endurance ride as your long steady session, provided the group actually rides steady and you have the discipline to sit in rather than chase surges. That gives you the social element, the visual stimulus, and the accountability without wrecking your intensity distribution.
For your key structured sessions — the ones where hitting specific power targets for specific durations matters — you move to a platform or a method that removes the temptation entirely. TrainerRoad is purpose-built for this. Its adaptive training system adjusts your workouts based on how you respond to previous efforts, and the interface is designed around execution rather than entertainment. There's nothing to chase, no one to race, no segment timers pulling you off target. You see the interval, you see the power, you do the work. For riders who want structured training and nothing else, it's the cleanest option.
Wahoo SYSTM takes a slightly different approach, wrapping structured intervals in video footage so you've got something to watch while you suffer. The key sessions are well-designed and the targets are precise. It sits somewhere between TrainerRoad's austerity and Zwift's full virtual world. For some riders that middle ground is exactly right.
And then there's the simplest option of all: load your workout onto your head unit and ride. A Garmin or Wahoo computer will walk you through the intervals with power targets on screen, and the trainer holds ERG regardless. No platform, no subscription, no leaderboard. Just the session and the numbers. For the rider who finds Zwift's distractions actively counterproductive, this is the cleanest path. You're not paying for a virtual world you're deliberately ignoring.
Why amateurs get this wrong and pros don't
The pattern Seiler's research identifies — 80 percent easy, 20 percent hard — isn't just for professionals, and it's not a suggestion. It's what consistently produces the best aerobic development across ability levels. But amateurs struggle with it far more than professionals, because professionals have a coach standing over their intensity distribution and amateurs have Zwift offering another race.
Dan Lorang talks about periodisation as a conversation between stress and recovery, where every session has a role and the week is designed as a whole rather than as a collection of individual efforts. The Zwift racer isn't periodising. They're reacting — to the race calendar, to the social pressure, to the dopamine hit of a good result. Each race feels productive in the moment, the same way each hard Tuesday ride feels productive for the time-crunched rider who never goes easy. The problem only becomes visible over months, when the numbers flatline and the fatigue becomes background noise.
Here's where the comparison gets sharp. A rider doing three Zwift races a week for twenty weeks accumulates sixty hard efforts with no planned progression, no taper, no recovery weeks, and no variation in stimulus. A rider following a structured plan for twenty weeks accumulates maybe forty hard efforts, but they're distributed across different energy systems, they progress in duration and intensity, they include deliberate recovery periods, and they build toward a peak. The structured rider does less total hard work and gets more out of it, because the work was designed to produce adaptation rather than just fatigue.
Making the switch without losing what Zwift gives you
The riders who resist moving away from Zwift racing usually aren't confused about the science. They know. They've read about polarised training, they understand intensity distribution, and they can see that their week is too hard. What holds them is that Zwift is fun and structured training on a blank screen is not. That's a real objection and it deserves a real answer.
The answer is the hybrid. You don't lose Zwift. You stop letting it run your training week. One race stays. The group ride stays if the group rides steady. The structured sessions move to a format where you can't accidentally race, whether that's TrainerRoad, Wahoo SYSTM, Zwift's own workout mode with the HUD stripped back, or a head unit with the intervals loaded. The total time on the trainer might not change at all. What changes is the distribution — more of that time spent properly easy, less of it spent in the grey zone, and the hard sessions concentrated where they'll produce the most adaptation.
The shift feels wrong at first, the same way easy riding always feels wrong for the rider who's used to going hard. You'll finish an easy Zwift group ride thinking you wasted an hour. You'll watch your average power drop on those sessions and wonder if you're losing fitness. You won't be. You'll be building the aerobic base that your Zwift racing was burning through, and when the next race comes — the real race, outdoors, the one you've been training for — the legs will be there in a way they haven't been before.
The platform is a tool, not a plan
Let me be really clear about this: the problem was never Zwift. The problem is treating a platform as a training plan. Zwift is a tool for delivering training stimulus in an engaging format. It does that brilliantly. But a hammer is a brilliant tool too, and you wouldn't use it for every job in the house.
The science has finally caught up with what the best coaches have known for years — that intensity must be rationed, that sessions need purpose, that recovery is where adaptation happens, and that consistency over months matters more than heroics on any given Tuesday. Zwift can exist comfortably inside that framework. It just can't be the framework.
If you want the structure behind this — the actual session design, the periodisation, the coaching input, and a community of riders who are serious about getting faster without burning themselves into the ground — the Not Done Yet community on Skool is where we put it all together. No pressure. But if the Zwift-race-every-night approach has run its course and you're ready for something that moves the numbers, that's where to start.