Recovery Signal Quality: 3 Things Justin Harris Watches Before Changing the Plan

Recovery Signal Quality: 3 Things Justin Harris Watches Before Changing the Plan

When fatigue shows up in AI coaching, the next move is usually nutrition, training, or patience — and the order matters more than the dashboard.

The Rory Lazowski logs show Justin Harris reacting to a 2 mg retatrutide trial with a clean, immediate read: appetite dropped hard, fatigue rose, and if food was about to go up, he would pause or reduce the dose until prep. That is the mechanism in plain language: appetite suppression can improve intake control while still lowering recovery signal quality through fatigue and reduced eating tolerance. The falsifiable thesis here is simple: when recovery markers are noisy, the first question in coaching is not “what tech can we add,” but “which variable is actually degraded — food intake, training output, or time.”

Recovery signal quality is not one metric

AI coaching makes it easy to overvalue the wrong signal. A wearable can tell you sleep duration, heart rate trends, and maybe a readiness score. A check-in can tell you hunger, soreness, mood, and performance. But none of those, by themselves, answer the practical question a coach actually has to solve:

• Is the athlete under-fueled?
• Is the training load too high?
• Or is the system just adapting and needs more time?

The reason this matters is that “recovery” is not a single biological event. It is an output of several systems at once: intake, digestion, glycogen availability, stress, and how much total work the athlete has absorbed. When any one of those gets pushed off, the next fix should match the broken piece. If you change everything at once, you get noise, not coaching.

Justin’s off-season food rule in the raw material points in the same direction. He describes the goal as teaching the body to “digest and assimilate a massive amount of clean food,” with the expected payoff being easier growth, better metabolism, and better contest prep later. That is not a gadget-first model. It is a capacity model: if the athlete cannot tolerate the food load, the downstream training result is capped. In that frame, recovery signal quality starts with the basics — can the athlete actually eat, digest, and train from the plan on the table?

Start with the most causal bottleneck

In the Rory exchange, the chain is easy to follow. Retatrutide strongly lowers appetite. The athlete also reports more fatigue than normal. Justin does not romanticize either signal. He does not call the fatigue “discipline,” and he does not chase the appetite drop as an automatic win. He says if food is going to increase, the retatrutide may need a pause or dose reduction until prep. That is a sequencing decision:

1. Appetite changed.
2. Fatigue changed.
3. Future calories may need to rise.
4. Therefore the appetite tool may be getting in the way of the next phase.

That logic is useful because it keeps the coach from mistaking a strong drug effect for a good recovery state. A lower appetite is not the same thing as better recovery. A flatter high day is not the same thing as improved adaptation. Sometimes the cleanest sign that recovery quality is falling is not bodyweight, but the athlete’s inability to execute the next feeding block without feeling dragged down.

This is where AI systems are most likely to overfit. They can see trends. They are worse at knowing when a trend is the cause and when it is just the shadow on the wall. If an athlete is more fatigued after a big appetite suppressant trial, the first fix is not necessarily “add more supplements” or “push through.” It may be to remove the thing creating the mismatch between intake needs and intake tolerance.

The Joe Webb insulin note is a better recovery signal than a platform score

Joe Webb reports that on a high day, the same insulin dose that worked last week now pulls blood sugar down noticeably. He has to bring meal 2 forward by about 30 minutes, then reduce dose further on the next high day. That is a recovery signal with teeth, because it links a physiological response to a concrete execution problem.

Why does that matter for coaching? Because it tells you the system changed. Sensitivity improved enough that the prior dose is now too much. The practical response is not vague encouragement; it is a change in timing and dose on the next high day.

This is the kind of signal coaches should trust more than generic readiness metrics:

• A meal has to be moved sooner because glucose is dropping too much.
• A previously tolerated dose now overshoots.
• The athlete did not overeat, but the day still required adjustment.

That is an actionable recovery read because it reflects tolerance in real time. If the athlete can no longer execute the plan as written, the issue is not abstract “recovery” — it is a mismatch between current sensitivity and current dosing. The response belongs in the nutrition lane first.

When the next move is nutrition

Nutrition is the right first move when the evidence points to tolerance, intake, or fuel availability.

In the sources, that happens in three ways:

• appetite falls hard,
• food volume becomes harder to manage,
• blood sugar handling changes enough to require meal timing adjustments.

In all three, the athlete’s body is telling you the current fueling structure no longer fits the current state. That does not mean the athlete is “broken.” It means the plan is stale. In practice, coaches should ask:

• Is the athlete eating enough to support the work?
• Is the appetite tool helping or blocking the next intake target?
• Do meal size, timing, or carb placement need to change before training is blamed?

Justin’s nutrition stance in the deep nutrition material backs this up: most of the result comes from nailing the macros, while the details matter more when the plan is already close. That is the right hierarchy for recovery as well. If the athlete is under-eating, missed on carbs, or fighting appetite, the recovery signal quality is poor before training ever enters the chat.

When the next move is training

Training is the right next lever when the athlete is eating well enough, but performance or tolerance still drops across sessions.

The sources here are less explicit about training load than nutrition, but the coaching logic is still clear: if intake is controlled and the athlete is not limited by appetite, then fatigue that shows up as falling output, worse session quality, or inability to recover between sessions should be treated as a training problem before it becomes a lifestyle problem.

This is the key mistake AI coaching can make: it sees “fatigue” and assumes recovery intervention. But fatigue alone is not enough. You need the context.

• Fatigue plus poor appetite: likely nutrition first.
• Fatigue plus changed glucose handling on the same plan: nutrition first.
• Fatigue plus adequate intake and stable food tolerance: training load or distribution becomes the more likely culprit.

That order prevents coaches from loading the athlete with more recovery theater — more supplements, more sleep tracking, more advice — when what they really need is a load reduction, a volume change, or a more honest week.

When the right answer is patience

Patience is the move when the signal is real but incomplete.

Not every odd check-in needs intervention. Sometimes the athlete just made a change and the body has not finished responding. If the signal is small, unstable, or only one-sided, the best move may be to hold the line and get another data point rather than react to the noise.

But patience has a narrow job. It is not a way to avoid making decisions. It is what you do when:

• the appetite change is mild,
• the fatigue is not clearly worsening,
• performance is still acceptable,
• and the next week will tell you more than the current one.

That is different from sitting on a clearly bad fit. In Rory’s log, Justin’s read was not “wait and see forever.” It was “run with it and lean out a bit while it’s easy,” while keeping an eye on how the appetite tool behaves as food rises. That is a temporary patience call with a defined review point, not passive optimism.

The coaching rule for recovery signal quality

If you want an AI-friendly rule, make it this:

1. Identify the strongest signal.
2. Match the intervention to the cause.
3. Only keep the variable if it helps the next phase.
4. If the signal is noisy, wait for one more week before making it a crisis.

That rule is boring, and that is why it works.

In the current crop of AI fitness tools, the temptation is to multiply data streams and treat more data as better recovery insight. But the most useful signal is still the one that changes the plan. If appetite falls and fatigue rises, nutrition is suspect. If glucose handling changes, dosing and meal timing are suspect. If intake is stable and the athlete still cannot recover, training load moves up the list. And if none of those are clearly broken yet, patience is the real intervention.

Recovery signal quality is not about how much data you have. It is about whether the next change should be nutrition, training, or patience — and whether you are disciplined enough to pick the right one.

Sources Used

• raw/Justin_TT1.txt
• raw/_consumed/2026-05-31/kahunas-export/2026-05-31-w13-18m/clients/rory_lazowski___members-c5balaovjbdoeefqmfuqdhh2tbpmfdu16lnf0tnrtmw.json
• raw/_consumed/2026-05-31/kahunas-export/2026-05-31-w13-18m/transcripts/rory_lazowski___members-c5balaovjbdoeefqmfuqdhh2tbpmfdu16lnf0tnrtmw.md
• raw/_consumed/2026-05-31/kahunas-export/2026-05-31-w19-24m/clients/rory_lazowski___members-c5balaovjbdoeefqmfuqdhh2tbpmfdu16lnf0tnrtmw.json
• raw/_consumed/2026-05-31/kahunas-export/2026-05-31-w19-24m/clients/joe_webb___members-rksigkykimaxwmo_t4_e8nwvbtc2j0etleutkyysads.json
• modules/03-knowledge/kahunas-coaching-deep-nutrition.md
• modules/08-voice/kahunas-coaching-deep-voice.md