I had a rider last year — mid-forties, strong time triallist, the kind of bloke who could sit on 290 watts for the best part of an hour and look like he was reading a newspaper. He did a 20-minute FTP test on his turbo, averaged 305 watts, applied the standard 95% factor, and walked away with an FTP of 290. His training zones were set accordingly. Sweet spot at 261-274 watts. Threshold at 274-290. Everything looked tidy on paper.
The problem was that his sweet spot sessions felt like recovery rides. His threshold intervals were comfortable enough to hold a conversation. He was doing structured training for eight weeks and getting almost nothing out of it, because the number that anchored everything was wrong. His FTP was not 290. It was closer to 300. Maybe 302.
How do I know? Because in a 10-mile time trial three weeks into that block, he averaged 298 watts normalised power for 23 minutes and told me afterwards it felt like a "good training effort." Not race-day pain. Not digging deep. A hard Thursday ride.
His 20-minute test had not lied exactly — 305 watts for 20 minutes was a legitimate result. The lie was in the conversion factor. Ninety-five percent is a population average. This particular rider, with his deep aerobic engine and minimal anaerobic contribution over 20 minutes, held closer to 97-98% of his 20-minute power for a full hour. The 5% haircut that works for most cyclists was too aggressive for him. It shaved off watts he could actually sustain, and his entire training programme was calibrated to a number that sat 10-12 watts below reality.
This is not a rare story. I see it regularly — riders whose FTP test produces a perfectly valid number that then sets perfectly wrong training zones. The test did its job. The conversion factor did not do theirs.
The conversion factor is a guess about you
Every FTP test protocol applies a conversion factor, and every conversion factor is a population average. Coggan and Allen established the 95% factor for the 20-minute test based on data from trained cyclists. The 75% of MAP factor for the ramp test comes from the observed range of FTP/MAP ratios in endurance athletes. Both are sound at the population level. Neither knows anything about your specific physiology.
The 95% factor assumes that a trained cyclist's anaerobic contribution over 20 minutes inflates their power by approximately 5% above what they could sustain for an hour. For most riders, this is close enough. But the actual figure ranges from about 2% to 8% depending on muscle fibre composition, training history, anaerobic capacity, and how the rider paces the test.
Here is where it gets specific.
Time triallists and diesel riders — people who have spent years developing their aerobic system through long, sustained efforts — tend to produce a smaller anaerobic contribution over 20 minutes. Their fibre composition is skewed towards slow-twitch. Their anaerobic ceiling is not especially high relative to their threshold. When they ride 20 minutes flat out, they are riding at perhaps 102-103% of their true threshold rather than the assumed 105%. Apply the 95% factor and you lop off 5% when you should only be lopping off 2-3%. The result: an FTP that sits 5-12 watts below what they can actually sustain, zones that are too easy, and training that does not produce the intended stimulus.
Sprint-oriented or anaerobic riders — people who came to cycling from rugby, football, athletics, or any sport that built significant fast-twitch capacity — tend to produce a larger anaerobic contribution. They can sustain 107-108% of their true threshold for 20 minutes because their anaerobic engine props them up. Apply the 95% factor and the result is still too high. Their FTP reads 10-15 watts above what they can actually hold for an hour, their threshold intervals feel murderous, and their compliance with prescribed sessions drops because they are working in the wrong zone.
These are not edge cases. They are the two most common ways the 20-minute test systematically misleads riders. The same physics applies to every protocol — the ramp test's 75% factor, the 8-minute test's 90% factor. Each embeds a population-average assumption about the relationship between what you tested and what you can sustain.
How the ramp test gets it wrong in two opposite directions
The ramp test is popular because it is short, repeatable, and removes pacing from the equation. You ride until you cannot ride any more. Simple. But the number it produces is not an estimate of your sustainable power. It is an estimate of your aerobic ceiling — maximal aerobic power, or MAP — and then it infers your threshold from that ceiling using the 75% factor.
The assumption: most trained cyclists sustain about 75% of their MAP for an hour. The reality: the FTP/MAP ratio ranges from about 72% to 82% across the population, and that spread is not random. It is physiological.
If you have strong anaerobic capacity relative to your aerobic base — you are punchy, you can attack, you can sprint — you push deeper into a ramp test than your aerobic engine alone would allow. Your MAP reads higher than it should from a pure aerobic standpoint, and the 75% calculation overestimates your FTP because it is taking 75% of an inflated ceiling.
This is the rider who finishes a ramp test feeling like a hero, plugs the number into TrainingPeaks, and cannot get through a 2x20 at threshold without blowing up at the 12-minute mark. The test says FTP is 280. The actual sustainable threshold is closer to 260-265.
If you are a diesel — deep aerobic engine, modest anaerobic punch — you reach failure on the ramp earlier than your threshold fitness deserves. You lack the fast-twitch power to push through the final high-wattage stages. Your MAP tests lower than it would if you had a bigger kick, and the 75% conversion sets your FTP below what you can sustain on the road.
This is the rider who gets an FTP of 250 from the ramp, then goes out on the weekend and rides 60 kilometres at a normalised power of 265 without any sense that they are at threshold. The test missed them.
Both of these failures are predictable. They are not bugs in the ramp test — they are features of a protocol that measures one thing and infers another. If you understand where you sit on the anaerobic/aerobic spectrum, you can interpret the result accordingly. (The VO2max Estimator can help you gauge your aerobic ceiling independently, which gives useful context for interpreting a ramp-derived FTP.)
What your rides are already telling you
The best FTP validation tool is not another test. It is the data you are already collecting on every ride.
Normalised power from sustained hard efforts. If you have a power meter and you ride hard — group rides, time trials, sportives, even solo efforts where you pushed for 45-70 minutes — you already have the data. Your normalised power from a maximal or near-maximal effort lasting 45-70 minutes is a direct estimate of your FTP. Not a conversion. Not a factor. A measurement of the thing itself.
Look at your last three or four hardest sustained efforts. If your normalised power from a 50-minute road race sits at 275 and your tested FTP is 260, one of those numbers is wrong, and it is probably not the race. Nobody accidentally rides 15 watts above their threshold for 50 minutes. If the race felt like a race — hard, sustainable, right at the edge — then 275 is a better estimate of your FTP than whatever the test produced.
The FTP Test Calculator will let you plug in efforts of different durations and see where the numbers land. Cross-reference it against your formal test result. If there is a gap, the ride data is usually more honest.
Heart rate at supposed threshold power. You should know your lactate threshold heart rate (LTHR) — either from a lab test, from a standalone 30-minute heart rate test, or from the heart rate you consistently see during sustained hard efforts. If your FTP is correct, riding at your FTP power should produce a heart rate within a few beats of your LTHR.
If you are riding at supposed threshold power and your heart rate sits 8-10 beats above LTHR, one of two things is happening: you are fatigued, or your FTP is set too high. If it happens consistently — not just on a hot day or after a hard week, but across multiple sessions — the FTP is the problem. Your body is working harder than threshold intensity to produce the prescribed power, which means that power is above your actual threshold.
The reverse also holds. If you ride at threshold power and your heart rate sits well below LTHR, your FTP is set too low and you are not getting the intended training stimulus.
Workout compliance rates. This is the one most people overlook, and it is the most telling. If your training plan prescribes 2x20 minutes at threshold and you consistently complete those intervals without significant fade, without RPE climbing above 8/10, and without heart rate drifting to or beyond LTHR — your FTP is either correct or slightly low. The intervals are doing what they should.
If you consistently fail those intervals — you cannot hold the power for the full duration, you need to drop the intensity mid-interval, or your RPE hits 9.5/10 by the 15-minute mark — your FTP is too high. Not your fitness. Not your mental toughness. The number.
And if you consistently smash those intervals — finishing with power to spare, feeling like you could do a third set, heart rate barely budging — your FTP is too low and you are wasting time at an intensity that is not producing the adaptation you are training for.
When your zones feel wrong
There is a specific pattern that shows up when FTP is set too high, and it is worth describing precisely because most riders feel it before they see it in the data.
Sweet spot feels like threshold. Sweet spot training — typically 88-94% of FTP — should be hard but manageable. You can talk in short sentences. Your breathing is up but controlled. You could sustain the effort for 30-60 minutes without a survival mentality. If sweet spot efforts feel desperate, if your breathing is ragged by the 15-minute mark, if finishing a 20-minute sweet spot interval feels like a victory, your FTP is too high. That intensity is not sweet spot for you. It is threshold or above.
Threshold feels like VO2max. Threshold work — at or near 100% of FTP — is hard. It is not comfortable. But it is sustainable for 20-60 minutes with the right pacing. If your threshold intervals feel like VO2max efforts — short, explosive suffering that you can barely hold for 5-8 minutes — then you are working above your actual threshold and your FTP is inflated.
Recovery rides produce TSS that seems too high. This is a subtler signal. If your easy spin produces a TSS of 60-70 for 90 minutes, and you know you were riding at zone 1 effort, the issue might be that your FTP is set too low — the system thinks your easy power is a larger fraction of your FTP than it really is. Check using the TSS Calculator with your current FTP versus a hypothetical adjusted number and see if the result makes more sense.
The FTP Zone Calculator and the Sweet Spot Calculator both depend on your FTP being close to reality. If your zones feel wrong, recalibrating the input is the fix — not changing the zone model.
Efficiency Factor and decoupling: your ongoing FTP audit
If you use TrainingPeaks, WKO, or Intervals.icu, you have access to two metrics that function as continuous FTP validators: Efficiency Factor (EF) and aerobic decoupling (Pw:Hr).
Efficiency Factor is normalised power divided by average heart rate. It strips away the noise of variable-power riding and gives you a number that represents how much output your cardiovascular system is producing per beat. At a given intensity, EF should be relatively stable from ride to ride — assuming similar conditions, similar fatigue state, and similar hydration.
If your EF at a supposed zone 3 effort is significantly lower than it was three weeks ago — same course, same conditions — something has changed. Either you are carrying more fatigue, or the intensity is no longer zone 3 for you. The opposite also works: if your EF at a given power keeps climbing over weeks, you are getting fitter and that power now represents a lower fraction of your threshold than it used to. Your FTP has probably risen.
Aerobic decoupling measures how much your heart rate drifts upward (or your power drifts downward) during the second half of a sustained effort compared to the first half. Dr. Joe Friel popularised this metric as a measure of aerobic fitness, and it works well as an FTP check.
In a steady-state ride at a true sub-threshold intensity, decoupling should be less than 5%. Your heart rate stays coupled to your power output. The aerobic system is handling the load.
If you are riding at supposed sweet spot (88-94% of FTP) and your decoupling exceeds 8-10% — heart rate rising significantly in the second half while power stays constant — you are working above your actual aerobic threshold. The effort is not sustainable in a true steady state. Your FTP is likely set too high, and what you think is sweet spot is actually threshold or above.
These metrics turn every ride into a soft FTP test. You do not need to set aside a day to retest. You need to look at the data you are already producing and ask whether the numbers are internally consistent.
The 95% factor is not universal — and what to use instead
Dr. Andrew Coggan, who co-developed the power training framework that most of us use, has been clear on this point: the 95% factor is a starting point, not a rule. In Training and Racing with a Power Meter, he and Hunter Allen present it as the typical relationship for trained cyclists. The word "typical" is doing a lot of work in that sentence.
Here is a rough guide based on what I see across the athletes I work with and what the published data supports:
93-94% — Riders with strong anaerobic systems. These are the people who can punch above their weight on 1-5 minute efforts, who came from explosive sports, or who have a high anaerobic work capacity relative to their aerobic base. Their 20-minute power is inflated by a larger-than-average anaerobic subsidy. Using 93-94% brings the FTP closer to what they can actually sustain.
95% — The default, and correct for the middle of the bell curve. If you do not have a strong reason to adjust, 95% is the right place to start.
96-97% — Diesel riders and time triallists. Deep aerobic engines, modest anaerobic capacity. Their 20-minute power is almost purely aerobic — the anaerobic contribution is minimal because they do not have much to contribute. Using 96-97% gives an FTP closer to their true sustainable output, and their training zones will actually produce the intended stimulus.
How to find your personal factor: Test with the 20-minute protocol, set your FTP using 95%, and then pay attention. If your zones feel wrong in the ways described above — sweet spot too hard, threshold unbearable — drop to 93-94% and try again. If your zones feel too easy — threshold is conversational, sweet spot is a rest — try 96-97%. Give each adjustment 2-3 weeks of training before deciding whether it fixed the problem. Your body will tell you.
You can also calculate backwards from a known sustained effort. If you averaged 295 watts for a 55-minute time trial and that felt like a maximal but sustainable effort, that is your FTP. If your 20-minute test produced 310 watts, your personal factor is 295/310 = 0.951. If it produced 305, the factor is 295/305 = 0.967. Now you have a conversion factor calibrated to your actual physiology, not to a population average.
The word "functional" is doing all the work
The F in FTP stands for functional. Not maximal. Not theoretical. Not what you wish you could sustain. Functional — as in, the number that makes your training function properly.
The right FTP is the one that sets your zones so that each zone produces the intended adaptation. Zone 2 should feel sustainable and conversational. Sweet spot should be hard but manageable for 20-40 minutes. Threshold should sit at the edge of sustainability. VO2max intervals should be maximal efforts lasting 3-8 minutes. If those relationships hold, your FTP is close enough. Your legs do not know whether the screen says 268 or 271.
Where it matters is when the number is off by 10-15 watts. That shifts your zones enough to put you in the wrong training stimulus. Sweet spot becomes threshold. Threshold becomes VO2max. The entire programme is calibrated to a version of you that does not exist.
If you suspect this is happening, use the FTP Zone Calculator with your current FTP, then again with your FTP dropped by 5%. Compare the zone boundaries against where your recent training actually fell. If the adjusted zones match your RPE and heart rate data better, the adjusted FTP is closer to reality.
A practical protocol for ongoing validation
Here is what I tell every rider I work with. It is not a one-off procedure. It is a habit.
After every formal test, set your FTP using the standard conversion factor for your protocol. Train with it for two weeks. Pay attention to how sweet spot and threshold intervals feel. Check your heart rate at threshold power against your LTHR. If everything aligns, the number is good. If something feels wrong, adjust by 2-3% and give it another week.
After every hard sustained effort, check your normalised power. Any time you ride hard for 40 minutes or more — a race, a fast group ride, a solo effort where you were properly motivated — compare the normalised power to your FTP. If it is within 5 watts, your FTP is validated. If it exceeds your FTP significantly and the effort felt sustainable, your FTP is too low and your training is leaving watts on the table.
Once a month, look at your Efficiency Factor trend. Pull up three or four rides at similar intensity — ideally steady endurance rides or tempo efforts on the same route. If EF is climbing, your fitness is improving and your FTP may need to come up. If EF is dropping, you may be accumulating fatigue or your FTP has drifted above reality.
Whenever you change your FTP, check the downstream effects. Run your new number through the FTP Zone Calculator and the Training Load Calculator. Make sure your planned TSS for the coming block still makes sense with the updated number. A 10-watt FTP increase changes the TSS of every ride you do — check that your weekly load is still in a range you can absorb.
The number is a tool, not a trophy
Riders treat FTP as a score — something to be maximised, bragged about, defended against the suggestion that it might be lower than they think. I understand the impulse. But an inflated FTP does not make you faster. It makes your training worse. Every watt of error above reality means intervals that are too hard, recovery that is insufficient, and fatigue that accumulates without producing proportional adaptation. A rider training with a correct FTP of 270 will get fitter than a rider training with an inflated FTP of 285, because the first rider is hitting the right intensities and the second rider is surviving sessions rather than adapting to them.
If you suspect your FTP is wrong — in either direction — test the hypothesis. Drop it by 5% and see if training quality improves. Or raise it by 3% and see if your intervals start to feel like what they are supposed to feel like. Give it two to three weeks. The answer will be obvious.
Your power meter does not lie. Your RPE is imperfect but honest over time. The only thing that lies is the conversion factor — a factor designed for an average rider and applied to a specific one.
Find the number that makes your training work. That is your FTP. Not the test result. Not the Strava PR. Not the number you tell your mates.
If you want to dig into the mechanics of different test protocols — what each one measures, when to use which, and how to execute them properly — the companion piece covers that in detail: Every FTP Test Protocol Compared.
And if you want to talk through your numbers with people who have been through the same questions, the Roadman Cycling community on Skool is where we do that.