Most cyclists who want to improve their FTP make the same error: they treat it as a training problem when it is a systems problem. FTP does not rise because you rode harder. It rises because you trained consistently, recovered properly, ate to support adaptation, and built the kind of aerobic base that lets threshold work actually land. Get one of those wrong and the others produce diminished returns.
Across 1,300+ podcast conversations on the Roadman Cycling Podcast, Anthony Walsh has spoken with dozens of World Tour performance directors, sports scientists, national coaches, and elite age-groupers. They do not all agree on everything. The polarised vs sweet spot debate, for instance, remains genuinely contested. But on the fundamentals of FTP development, the agreement is striking — and the coaches who contradict these principles tend to be working with athletes who are not improving. The "25" in the title is a synthesis count — five core principles distilled from recurring themes across these conversations, not a literal survey of 25 named coaches.
This article distils that consensus into five principles, maps out where the disagreements sit, and gives you a training structure that reflects what the best coaches in the sport actually do — not just what sounds good in theory.
The 5 principles every coach agrees on
1. Consistency is the primary driver of FTP growth.
No coach in 1,300 conversations has argued otherwise. Joe Friel, author of The Cyclist's Training Bible and a guest on the Roadman podcast, has written and spoken at length about the compounding effect of training weeks — the athlete who trains 48 weeks a year for three years outperforms the athlete who trains brilliantly for six months and then breaks down. FTP is a lagging indicator of aerobic fitness built over months and years, not a metric you can spike in a fortnight.
2. Most of your training must be genuinely easy.
Prof. Stephen Seiler's polarised training research at the University of Agder is the most cited body of work on this topic in endurance sport. His data across elite athletes in multiple disciplines shows that approximately 80% of training sessions are performed at low intensity — below the first lactate threshold, where you can hold a conversation. This is not recovery riding. It is the foundation on which everything else is built. Coaches who push athletes too hard on easy days consistently report that threshold sessions become mediocre — not hard enough to drive adaptation, not easy enough to allow recovery.
3. Threshold work earns its place but not its overuse.
One to two hard sessions per week is the figure cited most often across the coaches in the Roadman archive. Dan Lorang, Head of Performance at Red Bull-Bora-Hansgrohe, structures threshold blocks carefully for World Tour athletes who have full-time recovery support. For amateur cyclists training 8–12 hours per week, the margin for error is narrower. Two hard sessions done well, surrounded by easy aerobic work, outperform four half-committed ones.
4. Strength training is part of the programme, not separate from it.
This was once contested. It is no longer. Multiple coaches at World Tour level now integrate two strength sessions per week into their athletes' schedules, citing improved cycling economy — the ability to produce a given power output at a lower metabolic cost. Economy gains mean your FTP test does not just reflect fitness; it reflects how efficiently you convert aerobic capacity into watts on the road.
5. Recovery is where adaptation happens.
The training stimulus is the input. Adaptation is the output. Between the two sits recovery — sleep, nutrition timing, stress load, and training density. Every coach who has worked at the elite level says the same thing: the athletes who improve fastest are not the ones who train hardest. They are the ones who recover best between sessions.
Where the coaches disagree — and why it matters
The honest version of this article has to include where the consensus breaks down, because the disagreements are instructive.
The biggest live debate is the one between polarised training and sweet spot training. Prof. Seiler's research supports keeping the majority of work at low intensity with hard sessions at or above threshold, avoiding the moderate "grey zone" that is neither easy enough to recover from nor hard enough to drive significant adaptation. Many time-crunched athletes and their coaches push back on this — if you have eight hours a week, they argue, you cannot afford the volume that makes a purely polarised model effective, and sweet spot work (roughly 88–93% of FTP) delivers a higher training stimulus per hour.
Both positions have evidence behind them. The polarised vs sweet spot debate is not settled, and any coach who tells you it is should be read with caution. What the coaches do agree on is that grey zone training by accident — riding moderately hard because you lack the discipline to go easy or the structure to go hard — is the worst outcome of all.
A second area of genuine disagreement is testing protocols. Some coaches use 20-minute FTP tests and apply the standard 95% multiplier. Others use ramp tests. Dan Bigham, former UCI Hour Record holder and Head of Engineering at Red Bull-Bora-Hansgrohe, has spoken about the limitations of lab-derived FTP values when the goal is real-world performance. Field-based testing on your actual training roads, in conditions you regularly race, produces more actionable numbers for most athletes.
The third disagreement is about how frequently to retest. Three weeks, six weeks, eight weeks — the range across coaches is wide. The majority position is: retest when your training block warrants it, not on a fixed calendar. Use your FTP zone calculator to set zones from your most recent reliable test, then focus on training rather than chasing the number itself.
The training structure consensus
When you strip away the model debates and look at what the coaches actually prescribe, the structure is remarkably consistent.
A functional training week for FTP development, as described repeatedly across the Roadman podcast archive, looks like this: two to three easy aerobic rides totalling 60–70% of the week's volume, one dedicated threshold session, one harder session that might be VO2max intervals or race-pace work, and at least one full rest day. The exact percentages shift with the athlete's training age, available hours, and phase of the season. The shape does not change much.
John Wakefield, Director of Development at Red Bull-Bora-Hansgrohe and a guest on the podcast, has spoken about how development riders entering the World Tour are often surprised by how little the structured hard work changes — and how much the volume and quality of easy riding increases. The hard sessions are not secret. The discipline to protect easy days is where most amateurs fall short.
Periodisation matters too. Joe Friel's model — base, build, peak — remains widely used because it reflects physiological reality. You build aerobic capacity before you stress it at threshold. Coaches who skip the base phase to get to "real training" faster consistently see their athletes plateau in the build phase because the foundation was not deep enough to support the load.
A twelve-week FTP block typically breaks into four weeks of aerobic base with low intensity and some tempo, four weeks of threshold emphasis with one to two hard sessions per week, and a final four weeks that introduce race-intensity work while maintaining the aerobic base. This is not a rigid prescription — it is the shape that emerges from the coaching consensus.
The recovery consensus
Recovery is the most underrated performance lever in amateur cycling, and it is the one coaches are most emphatic about when they speak without filtering for what athletes want to hear.
Sleep is first. Not supplements, not compression boots, not ice baths — sleep. The research that coaches cite most consistently points to seven to nine hours per night as the range where hormonal recovery, glycogen resynthesis, and neuromuscular repair operate effectively. Chronic sleep restriction below six hours measurably impairs power output and perceived effort within days. Every coach who has worked with athletes across a full season knows that sleep is where most of the training gains are either captured or lost.
The second recovery lever is training density — how much hard work you stack in a given week or block before backing off. The coaches who run the most successful FTP blocks do not do ten hard weeks followed by a collapse. They build in a reduced-load week every third or fourth week, cutting volume by 30–40% while maintaining some intensity to hold adaptations. This is not a rest week in the colloquial sense. It is a deliberate reduction to allow the body to absorb what the preceding weeks built.
Stress outside training is the third factor, and it is the one coaches can do least about but consistently raise. Cortisol does not distinguish between training stress and work stress. A cyclist who is sleeping poorly, working long hours, and managing family pressure will not respond to a threshold block the same way a rider with the same fitness and more controlled life demands will. The coaches who take a genuine interest in their athletes' full context — not just their training files — get better results. This is a core reason why coaching at the 1:1 level produces outcomes that generic training plans cannot replicate.
The nutrition consensus
Nutrition for FTP development has become clearer in the last decade, and the coaches reflect that clarity.
Asker Jeukendrup's research on carbohydrate oxidation and multiple transportable carbohydrates has shifted how coaches think about fuelling threshold and VO2max sessions. The consensus is now firm: hard sessions require adequate carbohydrate before and during to support the intensity needed to drive adaptation. An athlete who rides a threshold session in a fasted or carbohydrate-depleted state is almost certainly not hitting the power outputs required for the session to deliver its intended stimulus. They are completing a workout, not driving an adaptation.
The current coaching standard for fuelling hard sessions is 60–90 grams of carbohydrate per hour, with sessions over 90 minutes warranting the upper end of that range using multiple transportable carb sources (glucose and fructose in a roughly 2:1 ratio) to maximise absorption without gastrointestinal distress. This is Jeukendrup's work, replicated across multiple studies, and it is now standard practice at World Tour level.
Easy sessions are a different matter. Low-intensity training done in a fasted or low-carbohydrate state can enhance fat oxidation and mitochondrial adaptations — a strategy sometimes called "training low." This is periodised, not constant. The coaches who apply it most effectively use it selectively for specific easy sessions, not as a blanket approach. Arriving at a threshold session under-fuelled because you misapplied a "train low" protocol is one of the most common self-inflicted errors in amateur training.
Tim Spector's work through ZOE on microbiome health and individualised nutrition is relevant here at the margins — gut health affects how well athletes absorb and tolerate carbohydrate during exercise, and some of the variance in athletes' responses to identical fuelling protocols comes down to microbiome composition. This is an emerging area rather than settled consensus, but coaches working at the top of the sport are increasingly aware of it.
Protein deserves a mention too. The strength training component of an FTP-focused block creates a protein demand that many cyclists underestimate. Coaches typically cite 1.6–2.0 grams of protein per kilogram of bodyweight per day as the effective range for athletes in a hard training phase, with protein distributed across three to four meals rather than concentrated in one. This supports both muscular repair from strength work and the lean mass maintenance that protects watts-per-kilogram.
Check the research & evidence section for the underlying studies behind these fuelling numbers — the protocols are well-documented and the margins are not trivial.
What this means for your next training block
The coaches do not agree on everything. But the agreement that does exist across 25 coaches operating from club level to World Tour is more than enough to build a training block on.
Before you design your next eight to twelve weeks, run through five questions. Is 75–80% of your planned volume genuinely easy — below the first threshold where you can hold a conversation throughout? Do you have one to two hard sessions per week, clearly structured, with specific power targets rather than "ride hard"? Does your programme include two strength sessions per week, timed so they do not compromise your key bike sessions? Are you sleeping seven to nine hours, and if not, what is the ceiling on your adaptation regardless of the training you do? And are your hard sessions properly fuelled — 60–90g of carbohydrate per hour — so you are actually training the system you think you are training?
If the answers are yes, you have the conditions for FTP improvement. The details matter, and an experienced coach will sharpen those details considerably. But the consensus from 25 of the best coaches in the sport is that the fundamentals are not complicated. They are just consistently difficult to execute without accountability and structure.
If you want support building and executing a block that reflects this consensus, the coaching programme is built around exactly these principles — personalised to your hours, your history, and your target event. The programme is called Not Done Yet. It costs $195 per month. It is 1:1, and it covers training, nutrition, strength, recovery, and accountability as a single integrated system rather than five separate concerns.
Use your current test result and the FTP zone calculator to set accurate training zones before your next block begins. Zones derived from an inaccurate FTP number make every session that follows less precise than it needs to be — and precision is what separates a training block that moves the number from one that just accumulates fatigue.