We encoded THETA's coaching into software because a human coach can only hold a few dozen athletes at once, while the demand for good HYROX guidance is enormous. And the method that produces results is a system that can be written down, tested and scaled. Selling more hours would help a handful of people; encoding the method helps thousands without diluting it.
- THETA BLUEPRINT is built on 3 years at the Elite 15 sharp end and 1,000+ teardowns of publicly logged elite training.
- Elite 15 athletes commonly train ~20–27 hours a week, which is exactly the kind of pattern a system can encode and adapt down for amateurs.
- Race data shows compromised running and the roxzone decide results, principles that generalise, so they can be programmed.
Why not just take on more athletes?
Because it doesn't scale, and I've seen this bottleneck before. In product-building, the answer to "more people need this" is rarely "hire more people to do it manually" it's "turn the repeatable parts of the work into a system". Elite human coaching is inherently limited: hours in a week, attention per athlete, and a waiting list that only grows. Selling more one-to-one hours would just make good coaching scarcer and more expensive, which is the opposite of what a young sport full of first-timers needs.
What part of coaching can actually be encoded?
The method. Setting paces off a 5K benchmark, biasing training toward a measured weakness, keeping most running easy, periodising base to build to race, deloading and tapering. These are rules, not magic. THETA arrived at them by reverse-engineering how the best athletes actually train and testing what survived contact with real athletes. Rules that consistent can be written into software that reads your assessment, applies them, and rebuilds your plan block to block, which is precisely what a good coach does between sessions anyway.
| Selling more coaching hours | Encoding the method in software |
|---|---|
| Limited to a few dozen athletes | Reaches thousands at once |
| Quality varies with coach fatigue | Same tested method every time |
| Expensive, long waiting lists | Accessible, no queue |
| Hard to update everyone at once | Improves for all as evidence moves |
What can't be encoded, and how do we handle that?
Some things resist automation: the relationship, in-the-room technique cues, the judgement calls around a messy life or an unusual injury history. I'm not pretending software replaces those. What it does is handle the 90% that is systematic (the programming, the progression, the adaptation) so that sound coaching reaches people who would otherwise train off a random PDF or nothing at all. For most amateurs, the gap isn't a lack of a world-class coach's intuition; it's the absence of any structured, adaptive plan built for the actual race.
How we built it, step by step
The process mirrored how I've built products: study the problem deeply, encode what's proven, then iterate.
- Analysed publicly logged elite training to find the patterns that repeat at the top.
- Distilled those into programming rules. Intensity distribution, phasing, station and transition work.
- Built an assessment so the system starts from your real fitness, not an average.
- Made the plan adaptive so it rebuilds block to block as you progress.
- Kept updating the method as new race data and athlete outcomes arrive.
"I've spent my career turning hard, repeatable work into systems. Coaching a first-timer well is largely systematic, so encoding it, rather than rationing it by the hour, was the honest way to get good HYROX training to the people who need it," says Michael Snook, CTO, THETA.
Does encoded coaching actually work for amateurs?
For the majority, yes. Because most amateurs aren't limited by the fine judgement only a human provides; they're limited by not having a real plan at all. An adaptive system that sets your paces correctly, builds your aerobic base, rehearses compromised running and progresses sensibly will outperform guesswork and static templates for nearly everyone. As of 2026 that's the gap BLUEPRINT was built to close: not to replace the best human coaches, but to make sound, race-specific coaching available to athletes who could never reach them.
Common questions
Why did THETA build software instead of coaching more athletes?
Because a human coach can only hold a few dozen athletes, while demand for good HYROX guidance is huge. Encoding the proven method into software reaches thousands without diluting quality, whereas selling more hours would just make good coaching scarcer and more expensive.
Can coaching really be turned into software?
The systematic parts can, setting paces off a 5K, targeting your weakness, polarised intensity, periodisation, deloads and tapers are rules, not magic. Software can read your assessment, apply those rules and rebuild your plan block to block, which is what a coach does between sessions anyway.
What parts of coaching can't be encoded?
The relationship, hands-on technique cues, and judgement around unusual injuries or messy schedules resist automation. Software handles the roughly 90% that's systematic so sound coaching reaches people who'd otherwise train off a random PDF or nothing.
Is encoded coaching good enough for a beginner?
For most beginners, yes. Their limiter is usually the absence of any real plan, not the lack of a world-class coach's intuition. An adaptive system that sets paces correctly and builds the aerobic base will beat guesswork and static templates for nearly everyone.
How was THETA BLUEPRINT built?
By analysing publicly logged elite training to find repeating patterns, distilling them into programming rules, and building an assessment so the plan starts from your real fitness. It's adaptive, rebuilding block to block, and the method keeps updating as new race data arrives.
Does software make human HYROX coaches unnecessary?
No, the best human coaches remain valuable for elite athletes and complex cases where judgement matters most. Encoded coaching changes access, extending sound programming to the majority of amateurs who could never fit into a top coach's roster.
How is an adaptive plan different from a PDF plan?
A PDF is fixed and generic, while an adaptive plan sets intensity off your own benchmarks and rebuilds each block as you progress. That ongoing adjustment is the part of coaching that actually drives improvement, which is why it's the part worth encoding.
Sources
- HYROX official race format and results (hyrox.com)
- THETA's analysis of publicly logged elite training (Strava, race splits, published programs), 2023–2026
- THETA coaching data, 2024–2026
Want this programmed for you? THETA BLUEPRINT is our coaching method encoded into an adaptive HYROX plan, built from a 2-minute assessment and rebuilt block to block, with the first week of every block free. Build my plan.