Shownotes

Common Denominators of Successful Innovations

01:40 -

• When thinking about what separates technology that endures from technology that gets abandoned, a few recurring factors stand out. The first is whether it is actually solving a problem. Triathletes tend to enjoy gadgets for their own sake, but that enthusiasm is not durable on its own.
• For one-time purchase products, broad adoption among tinkerers is relatively easy to achieve. For subscription-based models, the product has to demonstrate clear utility, or people will simply cancel. Beyond that, the benefit needs to be fast and unambiguous — users should not be left wondering whether something is working or whether it is placebo.
• Friction matters too, and it is worth separating two different kinds: the day-to-day friction of charging, syncing, and setup versus the onboarding complexity of learning to use something in the first place. Both can kill adoption, but they are distinct problems. And then there is cost, which is always a limiting factor.
• To that list I would add low complexity. This is not about underestimating anyone's intelligence — it is about opportunity cost. I have plenty of sensors gathering dust on shelves because getting anything meaningful out of them required a disproportionate investment of time for what turned out to be a marginal gain.
• When the time spent learning a tool could be better spent on more proven training strategies, most people will eventually make that trade-off.
• On the question of scientific validity, I think it helps prevent a sale more than it drives one. If a product has no evidence behind it, that can be a reason not to buy. But scientific backing alone does not create mass adoption.
• Near-infrared spectroscopy is a good example — it has decent correlations with lactate and ventilatory thresholds and is a genuinely non-invasive marker of metabolic intensity, but that has not translated into widespread use.
• What ultimately matters is whether users feel the product is helping them train better and get more out of their training. The jury is still out on NIRS in that regard.

Why is NIRS struggling?

06:05 -

• The barriers are cost, understanding, and ease of use. A device like Moxie is expensive enough that even someone who actively wants to experiment with the technology cannot easily justify it — particularly when wearing it is awkward and the day-to-day experience adds friction on top of the price.
• The pitch of replacing lactate testing is a compelling one, and at a lower price point it might land. But the cheaper alternatives have data quality questions, and that combination of expense and usability issues keeps it from reaching a broader audience.
• I tried a TrainRed unit and gave it a genuine chance, and while the quality seemed reasonable and they have validation work behind them, I simply could not find practical usefulness in the data relative to the time I was willing to invest in it.

Examples of Innovations: Successes and Failures

08:20 -

• I would not call NIRS a dead end. The more interesting question is whether these companies can survive long enough to build what is actually needed on top of the hardware. Right now, users are getting raw data, and raw data is not really what people want. They want to know what it means.
• Most athletes do not care about oxygenation slopes — they want to be told what they are doing right or wrong. Getting from sensor data to that kind of actionable insight requires a substantial software layer, good UX, and a large enough dataset to draw meaningful conclusions from.
• The problem is that building that dataset requires users, getting users requires sales, and sustaining sales while the product is still maturing is genuinely hard. It is the classic challenge of building the plane while learning to fly it.
• The broader issue applies to physiological sensors generally. Even with something as direct as power data — which is about as close to performance as a metric gets — it is not straightforward to go from knowing your power duration curve to knowing exactly what training you should do to improve it.
• If that translation is difficult with power, it is a bigger stretch to go from upstream physiological signals to concrete training insights. That is not to say it is impossible or without value, particularly in elite settings where small margins matter. But it is a real barrier that all of these products face.

Innovations That Were Regulated Out

11:50 -

• A clear example of an innovation that worked almost too well is the polyurethane swimsuits from 2008 and 2009. Performance improvements of two to three percent were documented even in world-class swimmers, world records fell across the board, and many of those records still stand.
• The product had essentially no friction — you just put it on. But it was banned quickly, and that is one factor that does not get discussed enough: regulation can remove an innovation from the market regardless of how effective it is.
• That said, it may not be the end of the story. Those suits are reportedly being re-engineered and used at the Enhanced Games.
• And in swimming specifically, pools themselves can be modified enough through fluid dynamics — overflow management, pool depth, lane rope design — that some of the drag-reduction benefits the suits provided can be approximated through pool engineering.
• That is part of why we have continued to see records fall even after the suits were banned.

Barefoot Running

13:25 -

• The barefoot shoe phenomenon is another failure worth examining, though not because the underlying concept is wrong. Born to Run came out around 2010, triggered a wave of adoption, and then the results simply did not bear out the promises.
• Injuries followed, often stress fractures, and the craze faded. I was working as a podiatrist at the time the book launched and saw it play out directly — patients throwing away their orthotics, then coming back with fractures.
• The concept itself has real merit. There is genuine value in spending time stimulating the foot and ankle musculature. The Stanford track team had athletes doing barefoot strides for years, and Nike developed the Free range partly out of that tradition.
• There is also good work around using more minimalist footwear or altered foot strike patterns as a tool for rehabbing knee injuries — one of my former colleagues wrote extensively on this. So the stimulus is real and the application can be valid.
• The failure was the all-or-nothing framing, which is ultimately a marketing choice. Injuries are multifactorial — you cannot reduce them to a single cause, and you cannot solve them with a single intervention. Load cannot be created or destroyed, it just shifts.
• Carbon-plated shoes raise the same kind of debate now. The question with barefoot running was never whether foot stimulus has value, but how much, when, and how it fits into total training load. Throwing your shoes in the bin was never the answer.
• In practical terms, the lowest-hanging fruit for most people is doing their gym work barefoot or in minimal footwear — that is probably the safest and most accessible form of that stimulus.
• A little barefoot walking on grass after a run, or around the house, can also be useful. Beyond that, most people are not going to tolerate much more, particularly if they have any history of foot or ankle issues.

Continuous Glucose Monitors

18:50 -

• I should flag a conflict of interest here: I worked for Super Sapiens and I am currently consulting for a company called Perform that has a CGM component, so I have skin in this game.
• The CGM story for non-diabetic athletes is another example of data without insights. What people actually cared about was not their glucose numbers — they cared about whether they were fuelling correctly.
• At Super Sapiens we were trying to bridge that gap with tools like Glucose Effectiveness, which essentially gave a binary answer: was your glucose tracking well with your workload, and therefore was your fuelling on point?
• That is the kind of feature layer that turns raw data into something actionable. There is also a potential role for CGMs in detecting overtraining signs — suppressed and unresponsive glucose, particularly when paired with a lactate sensor, can mirror the kind of suppressed lactate response associated with overtraining syndrome.
• But again, that is an insight built on top of the data, not the data itself. The other major issue with Super Sapiens specifically was cost — running sensors continuously is expensive, and that limited how broadly it could be adopted.
• The risk with this category more broadly is companies rushing to claim predictive insights before the underlying assumptions are solid. Recovery metrics, training readiness scores derived from sleep stages — many of these involve logical leaps that are not always validated.
• I use an Oura Ring and its AI illness detection triggers false positives for me roughly 70% of the time. In this particular case, I would rather it be overly sensitive than miss something, because the cost of a false positive is easing off training for a day, whereas a missed illness during a hard block is a bigger problem.
• But ideally the tool would let me set my own sensitivity preferences based on my risk tolerance, and ideally it would be trained on my data specifically — not just a broad population model, but something that knows what my biometrics look like when I am actually getting sick.

The Risks of Outsourcing Judgment to a Wearable

24:35 -

• If anyone is fully outsourcing their thinking to a wearable, they should throw it away. These tools should help tune your own sense of how you are feeling, not replace it. The wearable is the input; you are still the one processing it.
• That said, there are two real downstream risks when a tool gets things wrong too often. The first is the boy who cried wolf problem — if the alert fires constantly and inaccurately, you start ignoring it entirely, at which point it has no value.
• The second is the nocebo effect, and there is research behind this: if you give someone sham data telling them they slept badly, their performance will actually decline. So a wearable that incorrectly tells you that you might be getting sick can, in a meaningful sense, make you worse.
• These are not trivial concerns, and they deserve more attention in how this technology is designed. The problem is you cannot fully control how people use what you build — anyone who has spent time around product design knows that. You try to design for the right use, but you cannot prevent misuse.

Power Cranks and Transcranial Stimulation

26:10 -

• Power cranks — independent cranks where each leg pedals without mechanical connection to the other — were a notable failure from the early 2000s. There was early research that looked promising, and sponsored athletes gave them visibility, but the hype curve was steep and the results ultimately did not hold up.
• Another example from that era, and one closer to my own experience, is transcranial direct current stimulation. I spent time working for a medical device company in this space, where the primary clinical indication was depression, and the underlying science there is genuinely solid.
• But the technology also attracted attention for endurance performance, with some early studies — flawed in several cases — suggesting it could improve output.
• A company even built a headset product with integrated electrodes and got pro triathletes to use it. It never reached wide adoption.
• The problem was that in endurance sports you can measure everything precisely — pace, power, times — and the long-term performance improvements simply were not showing up. A good swim session after sitting in a lab with one of these devices on your head is not evidence that it works.

Altitude Tents

29:00 -

• Altitude tents belong in the failure category, though with some important caveats. We know altitude works — it is essentially a prerequisite for competing at the highest level in endurance sport.
• The tent concept should work in theory, but the dosage problem is significant. At actual altitude you are there around the clock. In a tent you are getting perhaps ten hours on a good night, which may simply not be enough stimulus, meaning you would need to extend the block considerably to compensate.
• And then there are the practical problems: poor sleep quality, noise, and the obvious issue if you share a bed with someone.
• Sleep is probably the biggest barrier. If you are already getting the minimum eight hours and the tent is degrading that quality, you are not recovering properly, and then whatever altitude adaptation you might be gaining is being offset by impaired recovery.
• The net effect on performance may well be neutral or negative — not because altitude does not work, but because the conditions for it to work are not being met.
• Cost is a factor in how widely it could spread through the amateur field, but I do not think it is the primary barrier.
• Athletes who really want to perform will spend money if they can feel and measure an improvement. The problem is that many using tents are not seeing that improvement clearly enough.

DNA-Based Training and Genetic Testing

33:10 -

• Genetic testing for training and talent identification peaked in hype around the time of the Human Genome Project in the early 2000s, and it has not delivered.
• The core issue is that phenotype matters more than genotype — what you actually express is the product of both genetics and environment, and the interactions between them are currently too complex to make reliable individual predictions from.
• There is also a psychological dimension that these tests ignore entirely, and psychology is a substantial driver of performance. I have looked at what these genetic reports produce and it is simply not there yet.
• Maybe large enough datasets with AI applied to them eventually get us somewhere useful, but that is a fairly optimistic framing of where things currently stand.

Ketones

34:24 -

• Ketones attracted enormous attention around Chris Froome and the 2012 Olympics and have remained a topic of ongoing research since. My honest view, and I should note I have a conflict of interest in this space, is that we are not close to the end of the research curve.
• The work on carbohydrates began in the 1990s and three decades later we are still refining our understanding. Ketone research is nowhere near that stage — probably another ten to fifteen years of quality work needed before we can draw firm conclusions.
• There are recent papers showing some interesting signals: one looking at 31 days of high-dose supplementation found improvements in cognitive function and some oxygen uptake measures, though 5k race performance was not improved; another found positive mitochondrial adaptations over eight weeks of periodised training with ketones.
• There is enough smoke here that the research may eventually support the product. But if the taste remains poor and the cost stays high, mass adoption is not coming regardless of what the science shows.

Triathlon Wetsuits

36:20 -

• Wetsuits are about as clean an example of successful innovation as you will find in the sport. Adoption was rapid from the late 1980s onward because the benefits were immediate and obvious on two fronts: comfort in cold water and a clear performance gain.
• The friction of using one amounts to five or ten minutes putting it on before the race, and nothing else about your preparation or behaviour needs to change.
• What makes wetsuits particularly interesting compared to most performance innovations is that they are not top-heavy in their benefit. Most technology disproportionately helps the faster athletes.
• Here it is arguably the opposite — slower athletes spend more time in the water, so the warmth and buoyancy matter more to them, and the rules reflect this with a lower wetsuit cutoff temperature for professionals than for age groupers.
• Economies of scale also mean that a good-enough wetsuit costs a fraction of a top-end one without a dramatic difference in real-world performance, which has helped drive adoption across the full range of ability levels.

Aero Bars

38:28 -

• Aero bars had their defining moment in 1989 when Greg LeMond used them in the final Tour de France time trial and won the race by eight seconds. That viral moment drove rapid mass adoption.
• The friction is minimal — in the early days you simply bolted clip-on bars onto a road bike and that was it. The speed benefit is immediate and completely unambiguous.
• You do not need science to confirm it; a stopwatch tells you everything. That combination of near-zero friction and instantly perceptible benefit is exactly the pattern you see in most successful innovations in this space.

Super Shoes and Coaching Platforms

39:58 -

• Super shoes surprised me initially. Carbon plates had been around for nearly twenty years, so I was sceptical when the performance claims started coming in. Then you put them on.
• There is interesting research on the placebo component — perception of the shoe matters — but the performance data is real, not placebo. The improvement is something you feel immediately and can measure straight away, which is the recurring theme in everything that has succeeded.
• Coaching platforms like TrainingPeaks represent a different kind of success but follow the same logic. Compared to what came before — faxed training programs on a Sunday night — the reduction in friction for both coaches and athletes was enormous.
• Communication, programme delivery, session analysis: all of it became vastly more accessible. Even self-coached athletes now have something like Strava or Garmin Connect as a baseline, giving them at minimum a view of training volume and session data.
• The adoption is close to ubiquitous for anyone who is coached, and widespread even among those who are not.

Power Meters and GPS Watches

42:03 -

• The power meter and the GPS watch belong in the same category — tools that measure what you are doing and make both prescription and execution of training more precise.
• The power meter took a long time to reach mass adoption, partly because of cost and partly because early versions were too heavy to race with. It appeared in the late 1980s but did not reach critical mass until the 2010s.
• A big part of what unlocked that was a framework for using the data — Training and Racing with a Power Meter gave people a practical structure for interpreting what they were seeing, and that made the difference between a metric that confuses people and one they can act on.
• That lesson applies directly to some of the newer innovations that are still struggling today.
• The GPS watch followed a slightly faster trajectory from its arrival around 2003, helped by falling costs and a strong network effect — once enough people were using them, the social pull became part of the adoption story.
• I got my first one in 2011, a Nike model that had its strengths and limitations, but by that point the question was becoming not whether to get one but when.
• Today most people starting out in running feel almost immediately that they need to track what they are doing, and a watch is the default answer.
• Hunter Allen has just released a book called Training and Racing with a Glucose Monitor, which is an obvious parallel to the power meter playbook. It is a smart move, but I do not think it will achieve the same ubiquity.
• The distance from glucose data to race performance is considerably greater than the distance from power to performance — probably an order of magnitude or two further removed.
• Nutrition getting it right matters, but it is a less direct lever than your power duration curve. There may be broader adoption through the health angle beyond pure athletes, but the same level of penetration as power meters seems unlikely, at least for now.

Sports Gels

47:20 -

• Gels might seem like an odd entry in a list of successful innovations, but they belong there. The first ones appeared in the late 1980s, and they solved a real and immediate problem as the understanding of carbohydrate needs in endurance sport developed.
• The friction of carrying and using a gel is about as low as it gets. I should note a conflict here — this podcast has been sponsored by Precision — but setting that aside, the adoption has been essentially total.
• There is a growing movement towards simpler alternatives like real food or candy, and there are legitimate arguments around cost and environmental impact, but the convenience factor is almost impossible to argue against.

Predictions: Bespoke Footwear

48:05 -

• Looking ahead, I think bespoke footwear is coming, though it is a longer-term prediction rather than something imminent.
• The concept would involve scanning your foot with your phone using a marked sock as a reference, then having a shoe 3D printed to your specific geometry.
• Over time, feedback from the user would allow the prescription to be refined, and different versions could be produced for different types of running.
• In the near term it probably starts more modular — choices across upper, midsole, outsole, rocker profile and plate — before moving towards something fully individualised.
• Embedded or external sensors feeding back into the system over time is a natural extension of that. It is speculative, but the direction of travel is clear.

Predictions: Continuous Hydration and Sweat Sensors

49:18 -

• Continuous sweat rate and sodium sensors are an interesting case. The underlying idea is sound — building an individual sweat and sodium profile that can inform hydration planning, and potentially adjusting that in real time during long events.
• The questions are around sensor accuracy over the duration of a race, particularly as salt crusts on the skin or sunscreen degrades the reading, and around how much precision is actually needed over and above what you can establish through simple pre- and post-session weigh-ins combined with trial and error.
• There are already products in this spacE. Both use algorithms, which introduces its own layer of uncertainty.
• My honest view is that these end up as features rather than standalone products. The "octopus problem" applies here — there is only so much real estate on the human body for sensors, and a sweat sensor worn in isolation is a hard sell.
• The more likely trajectory is acquisition: a company like Core, for example, could buy one of these businesses and roll the sweat measurement into their heat adaptation platform, producing richer insights as part of something people are already wearing. That model makes more sense to me than a standalone device reaching mass adoption.
• There is, however, a safety angle worth taking seriously. More people are doing more extreme events with less preparation than ever before, and hydration failures are a genuine medical risk.
• Having worked as an event medical officer at ultras, where we would weigh athletes at checkpoints, I can say that real-time sweat data would add useful context to those decisions even if it would not be the whole picture.
• Climate trends make this more relevant over time, not less. So while I do not see mass adoption for performance purposes, the safety use case could drive broader uptake in a way that pure performance metrics alone would not.

Predictions: Continuous Metabolic Insights

55:18 -

• I should flag again that I work for Performa, which is relevant context here. The prediction is for multi-analyte continuous sensors — devices that measure glucose, lactate, ketones, hormones, and other markers simultaneously and in real time.
• The potential, if that data can be integrated with training and health data on a single platform, is genuinely significant for both health and performance.
• The important caveat is that continuous data does not look like discrete data. Continuous lactate readings will not resemble finger-prick lactate values — they are measuring something related but different — and that is going to create a difficult adoption curve.
• There will be an initial wave of enthusiasm, followed by a backlash when people conclude the data is inaccurate, before a more mature understanding eventually develops. The learning curve is real, and anyone entering this space needs to anticipate it.

Predictions: Ventilation Measurement

56:50 -

• Devices that measure breathing rate, ventilation, and respiratory volume during exercise sit in a similar category to NIRS — physiologically meaningful in principle, with real question marks around practical adoption.
• In theory you can derive a reasonably accurate picture of VT1 and VT2 from ventilation data, which makes it a potential tool for setting training zones without lab testing.
• Timeware is the company currently sponsoring Visma in this space, and I have spoken to them directly about it. The problems are several. Cost is one — a hardware purchase plus a subscription is a significant commitment if the main output is training zones you could establish other ways.
• The chest strap form factor is another: people increasingly dislike wearing chest straps, and launching a chest strap-dependent product now is a timing problem that would not have existed in the 1990s.
• There is also a fundamental physiological complication in that breathing can be consciously overridden — you can intentionally hyperventilate or hypoventilate — which introduces noise that heart rate does not have.
• And in a running context, the movement artefact problem has not been convincingly solved, at least not in any studies I have seen. Respiratory rate, measured well and in combination with heart rate, is probably one of the most powerful internal load signals available.
• But the "measured well" part remains the challenge, and I think the product needs to be integrated into a broader sensor ecosystem rather than sold as a standalone device.

Predictions: Hyper-Individualised Training

1:00:40 -

• The longer-term prediction I find most compelling is hyper-specialised training driven by a combination of population-level data and deep individual data from multiple sensors.
• The vision is something like: given your current power duration curves, your metabolic sensor data, and your physiological responses over time, here is the training that would work best for you right now, and here is how it should evolve as you adapt.
• That is not science fiction — it is arguably what the Norwegians were already attempting manually with athletes like Blummenfelt. The question is when it reaches broader adoption.
• My honest view is that full automation of training decisions is probably a decade away, and even then the limiting factor will be confidence as much as capability.
• Athletes already modify their coach's programme constantly based on how they feel — that is not going to disappear with AI. The more realistic near-term trajectory is modular: discrete analytical tools sitting within existing platforms like TrainingPeaks, each handling a specific domain. Aerodynamic analysis from power and speed data.
• Stroke analysis from motion capture in swimming. These silos gradually become integrated as the data infrastructure matures.
• Full outsourcing of training decision-making to an AI platform is something quite different from self-coached athletes using AI-informed insights, and I think that distinction matters. The first category is much further away than most people assume.

Predictions: Aerodynamic Sensors for Cycling

1:05:10 -

• I have been bullish on aerodynamic sensors for six or seven years now, and they remain a niche product that most athletes are not yet getting real value from.
• My prediction is that they follow a power meter-like trajectory — a long runway to maturity, driven more by improvements in usability and data integration than by cost reductions alone.
• The reason I am more optimistic about this than my other predictions is that aerodynamics is not a marginal gains story for most amateur athletes. The upside is large. Position changes can save meaningful minutes over a 40-kilometre bike leg, and that is a lever worth pulling.
• The problem at the moment is that the sensor alone is not enough. Seeing your CDA number is close to useless without a proper testing protocol, and most users are left to figure that out themselves.
• What is needed is a guided, stepwise process built into the software: start on an indoor trainer working on a sustainable position, establish a baseline, then move outside to test specific changes against that baseline using a structured out-and-back protocol.
• AeroTune was doing something along these lines with a Garmin ConnectIQ app years ago, and that approach is essentially correct. The sensor companies need to help users decide what to test and in what order — most people would be better served addressing stack height and position before worrying about whether calf guards save two or three watts.
• A tool like Robbie Ketchell's Aero Chat app, which uses a phone or webcam camera to estimate CDA during an indoor session and suggest position changes, points at one accessible end of this.
• I downloaded it, used it once, and then did not go back — which is its own data point about friction. But the concept is right, and combining that kind of position guidance with outdoor sensor testing and a proper data dashboard that pulls in power, heart rate, and temperature alongside CDA would be a genuinely useful product.
• The data dashboard point matters because improving your CDA at the cost of a suboptimal hip angle might lower your threshold power and end up costing you more on the run than you saved on the bike.
• You need contextual synthesis to make sense of those trade-offs, and no product is currently doing that well.
• Used without that guidance, an aerodynamic sensor in the wrong hands is easy to misuse. That is not a reason to dismiss the category — it is a software and onboarding problem, and it is solvable.
• I think this one does reach meaningful adoption eventually, not at the level of power meters, but significantly beyond where it sits today.

Exogenous Lactate: Cost and Physiology

1:12:33 -

• The obvious question with exogenous lactate supplements is what they would cost to produce and therefore to buy.
• My sense is that to reach anyone beyond the top one percent of the performance market, pricing would need to sit somewhere between a premium gel like Maurten and perhaps three to five times that — closer to the cost of a Maurten gel than to ketones.
• The dosing model matters here too. Ketones might be used once or twice across an ultramarathon. Gels are taken multiple times an hour.
• If lactate supplements need to be dosed at anything like that frequency, the per-unit cost has to be in a similar ballpark to an expensive carbohydrate product, or adoption simply will not scale.
• On the physiological side, the most intuitive framework is to think of it as a third fuel pathway alongside glucose and fructose.
• We already understand the dual-transporter model for carbohydrates — glucose first, then fructose unlocking a secondary pathway that allows higher total carbohydrate absorption.
• Lactate may function as a third pathway, though it is more complex than a straightforward carbohydrate source because lactate is also a signalling molecule with broader metabolic roles.
• The research needed to understand how this plays out in practice simply has not been done yet.
• One angle worth raising is organ-specific fuel use. Lactate is a primary fuel source for the heart, and possibly the brain as well.
• Over the course of a very long event like an Ironman or an ultra, cardiac output inevitably declines — some of that is attributable to muscle fatigue limiting the demand, but some may reflect the heart itself fatiguing as a muscle.
• There is research showing measurable cardiac stress and transient reduction in cardiac function following ultra-endurance events.
• Whether providing exogenous lactate could attenuate some of that stress or extend the period over which higher cardiac output is sustainable is genuinely unknown at this point.
• It is a reasonable hypothesis, and it is exactly the kind of question that will only get answered once these products are on the market and being tested properly.

Testing innovations for its success

1:18:10 -

• To close out, we each named one prediction we got right and one we got wrong. I have been calling aerodynamic sensors as an imminent breakthrough for six or seven years now, and they still are not where I said they would be.
• That is a clear miss. I remain convinced the trajectory is there, but the timeline has been consistently further out than I expected, and I have to own that.
• On the other side, I was never persuaded by CGM for non-diabetic athletes. My view was that glucose is too far upstream from actual performance and too indirect a marker of something that is not that hard to get right through simpler means.
• That has broadly played out. And to be fair to the other side of that argument, the CGM story is my own failure too — I worked for Super Sapiens and genuinely believed in what we were building.
• Just because there is value in the data does not mean a product succeeds at scale. Those are two different things, and that distinction is worth sitting with.
• The prediction I would put on the success side of the ledger, even if it is slightly early to declare it, is blood flow restriction training filtering into endurance sport.
• I said on a previous appearance that endurance athletes would be doing isometrics and BFR training, and I am seeing that trend develop.
• A company called Hytro — no affiliation, though I know some people there — has made a meaningful contribution to this by building the cuffs directly into shorts with a simple velcro tension system.
• That removal of friction is exactly the kind of thing that takes a niche practice and starts moving it toward broader adoption.
• It is still mostly at the high-performance end for now, but the practical use cases — travel, reduced fatigue compared to conventional strength training, recovery applications — give it a clear path to trickling down through the field.

LINKS AND RESOURCES

• David’s website, Substack, Podcast and Instagram 
• David’s previous appearance on That Triathlon Show: David Lipman | EP#414