From Metrics to Skills: Carv’s New Coaching Pathway

By John MooreHead of Product

October 28, 2025

8 min read

John Moore is Carv’s Head of Product, which means he spends his days flying around the world testing, talking to Carv users, and navigating Figma at 5am to design our latest features across timezones. But most importantly, John’s world is about figuring out how to make Carv more effective and more fun.

This season, John and the team have made some of the biggest changes to Carv in years: moving away from a focus on metric details, toward coaching whole-body skills that make progress simpler, and clearer for every terrain.

We sat down to dig into what wasn’t working, what changed, and how skiing with Carv is going to feel different this winter.

John is Carv’s Head of Product

From Granular Data to Whole-Body Coaching

Looking back at last season, what didn’t work so well?


JM: Carv has always been incredible at measuring the fine details of skiing. After years refining our motion models, we can score them with real accuracy.

But our coaching was always built directly on those metrics. Carv could analyze what was happening in your turn, but coaching didn’t always click. We were basically poking and prodding at tiny details one at a time: “Move your boot angle here,” “rotate there.” It was quantitative, and it worked well if you were willing to dig into the data.

But for many of us, a reactive, metric-driven approach didn’t make it obvious what to do next. And we heard your feedback loud and clear in our end of year survey. Skiers love chasing a better score, but when they get stuck, they’re not sure how to fix their skiing to get the next breakthrough.

“We want you to understand your skiing at a higher level: what should your whole body be doing? That’s what actually clicks.”

The coaching system is moving from reactive metrics-driven system (left) - to a system where you pick the goal, and you get a step by step pathway of the movements you need to make (right)

Did you feel that yourself as a skier?


JM: Definitely. Last season, I had a big breakthrough in how I initiate turns when carving. Carv had identified it, but I was struggling to make the changes. I actually learned it through exploring other coaching materials that explained whole-body movements at turn initiation.

Once I made that change, everything unlocked: early edging, better grip, more control — and crucially, I could now do it on steeper terrain, not just easy slopes.

We have known for a while we needed to bring a more holistic coaching approach into Carv. I mean, it’s probably not that controversial, human coaches don’t just focus on small parts of the puzzle. They look at the bigger movement pattern and build that up.

This approach becomes even more important as we grow. Carv’s community includes people with very different knowledge, goals, and attitudes to skiing. The system has to show you what to do next without requiring deep metric understanding.

Why Technique is the right lens for Terrain

How has Carv’s approach to terrain changed?


JM: Last year we launched terrain detection with Carv 2, where Carv’s motion AI detected the snow surface — smooth vs. uneven. That was a great step, but it wasn’t the whole problem.

Skiing isn’t just about terrain, it’s about technique. You don’t ski moguls the same way you carve a groomer, so it never made sense to score them with the same metrics. Detection is part of the issue, but scoring and coaching was the main problem.

Carv was using the same metrics and a generalizable algorithm to coach each terrain. That meant the algorithm ended up with feeling less interpretable, and some of the metrics didn’t make sense (G-force didn’t feel as relevant in powder as it does in carving).

This year, Ski:IQ and coaching are built for the specific technique you’re working on, giving you feedback that matches how you ski.

  • We’ve built different metrics for each Ski:IQ pathway, so the system detects what terrain you’re on
  • It figures out what good technique looks like there, and coaches you toward it.


PS: Thanks to a much bigger skiing video library we’ve also made big leaps in terrain detection itself — powder and mogul accuracy improved by 30% thanks to you guys out there!

Trust us, the Carv team spared no energy hunting down the powder Ski:IQ testing in Chile this summer

So what does this actually mean for Carv skiers?

JM: Now Carv can ask what your goals are, and deliver focused coaching that matches them. Instead of hacking away in a sea of metrics, you’ll know exactly what to work on for the terrain you care about — and you’ll feel progress.

Generally, most of us learn by doing. Unless you’re lucky enough to work with an instructor, skiing improvement is often “fake it till you make it.” Now Carv can guide you straight to the skills that matter. Well that’s the plan. It’s giving skiers more dimensions to improve.

In the words of Tom Gellie, top level instructor: “even on a cloudy or choppy-snow day, when you can’t carve or ski powder, there’s now something to play for. This year’s Carv can coach that versatility with short turns, moguls and parallel pathways.”

The Carv app shows you which skills you need to tackle next

Behind the scenes

So how did you actually create these skills?

JM: We went back to the foundations: how human coaches actually teach. We spoke to top-level skiers and instructors all over the world. Everyone had slightly different opinions, but there were strong points of consensus.

We whittled it down to four to six essential skills for each technique. Enough to make real progress, but not so many that people get overwhelmed. Then we tuned our metrics to measure those skills effectively, and iterated endlessly

What’s been the hardest part for the product team?


JM: Balancing different learning styles. Some people need audio cues, some visuals, some written coaching. You need to serve all of them without overwhelming anyone. And then there’s the sheer content load — creating enough examples, videos, and cues — and tuning it so it feels supportive for every skier, but not noisy. That’s been the marathon.

The team tuning the metrics and skills with Reilly McGlashan in Los Colorado to ensure that each skill score goes up as you improve.

What don’t people see about how we actually build this stuff?


JM: Honestly, the logistics. To build year-round, we’re traveling to far-off places, skiing with new people, praying for the right conditions. You go out to test, and suddenly something’s broken. You sprint back to the laptop, crank out a fix, and rush back to the hill.

We move fast because we’re small, seasonal, and ambitious. That means constant hiccups and constant fixes. Sometimes Carv says something totally wrong in a test, and you’ve got to explain to testers “this is changing” while still getting the feedback you need.

We have some rough moments building Carv, and some pretty special ones.
The view over Santiago after skiing until dark is a pretty good one.

Was there something you thought would be good but just wasn’t?


JM: Carv has always had a powerful real-time training mode. Each year we try to improve it though. This year tried all kinds of clever levelling systems based on the skills — levels going up and down, streaks, metric ceilings.

But on snow, it was overwhelming.

What worked was the simplest version: levels only go up. Done right, with well-tuned audio and visuals, that’s more effective and more fun. Skiing is hectic enough — Carv needs to fit into that.

What small detail are you secretly proud of?


JM: In training mode, we’ve added a little heads-up display showing what level you’re on, your best level this season, and your progress through the day. It’s a gamified tracker that feels elegant and motivating but keeps you connected to the journey without overloading you

The Carv app shows you which skills you need to tackle next

How will skiing feel different this year

If everything works perfectly, how will skiing feel different this year?


JM: It’ll feel more accessible. Skiers will know exactly what to work on when they want to improve, instead of just hacking away and hoping their metrics jump up.

But we don’t always want to push every run, or even every day. So when you don’t want to train, you can just cruise with Track and Learn modes (just tips or just Ski:IQ), still get value from Carv, and have fun with friends.

If we nail it, Carv won’t just measure your skiing — it’ll give you choice, clarity, and a much more fun way to get better. I hope these changes will put a big smile on your face this winter.


The update rolls out on iOS and Android in the first week of November. Simply update in the App Store to see a refreshed UI, upgraded Ski:IQ, and the full coaching pathways.

Once you’ve tried it, we’d love to hear what you think. Drop us a message on the website or in the app, and let us know how you find it. Your feedback is what keeps shaping Carv.

00

Written by: John Moore

Head of Product