A note from the founder

One of the reasons I started doing things differently was because I got tired of seeing good coaches run their whole business out of five apps, none with their name on it, and gyms lose the members they paid to bring in the day a trainer walked out.

Fifteen years in gyms, on both sides of the desk. Here is what I saw, what we see now, and how we decide what to build.

The founder of Marrow, a working coach, smiling in a gym shirt.

The story

The coach is the product.

Fifteen years of coaching in gyms and partnering with the people who own them taught me the same lesson from both sides.

The software was never built for the coach. It was built for an admin, so a coach who had never run anything like this before got a screen full of buttons, a client who could not work out how to join, and a payment they had to chase by text.

On the owner's side, the member record lived in the trainer's phone. Nobody decided who owned the relationship until the day it mattered, and by then it had already left the building.

I did not think the fix was a smarter app. I thought it was a simpler one. One app with the coach's name on it. The gym keeps the members it brought in. Every number on the screen is one the coach can trust, or it is labelled as an estimate. That is what we built. I run my own in person clients on it, on the same account any coach gets.

The founder of Marrow.Coach number one on the platform, verified by the same process as everyone.

Contrarian, on purpose

Five mistakes we see coaches in gyms make.

We would rather tell you what goes wrong than list our features. These are the five we see most, and the fix for each one is usually the opposite of what looks professional.

  1. Running on five apps and calling it professional.

    Programs in one app, booking in another, payments in a third, messages everywhere. Your client does not see a professional. They see five logins and none of them with your name on it.

    Instead

    One app, with your name on it.

  2. Assuming the invite is the easy part.

    The most common one star review in coaching apps is a client who opened the app, saw Sign in, and could not find a way to join. The coach never hears about it. The client just stops. The first screen your client meets matters more than the program behind it.

    Instead

    The first screen your client meets is the way in.

  3. Tracking more to look thorough.

    Amino acid rings, red numbers and warnings the day someone falls behind. The more you ask a client to track, the less they log. Track what you will actually read on Sunday.

    Instead

    Track what you will actually read on Sunday.

  4. Making every log wait on you.

    Confirm the meal, approve the photo, sign off the check in. You become the bottleneck, the client waits, and nobody wins. Your job is to correct, not to approve.

    Instead

    Correct, do not approve.

  5. Not deciding who owns the member until a trainer quits.

    On most systems the relationship lives in the trainer's phone, and the owner finds out when it leaves. Decide it now, in writing, inside the software, while everyone is still friends.

    Instead

    Decide it now, in writing, inside the software.

Judgment, not claims

Three calls we made that most coaching apps would not.

Before we built, we read 1,475 recent reviews across seven coaching apps, 544 of them at one or two stars, to learn where coaches and clients actually get stuck. Then we made some calls.

The ask: AI macros from a photo

We would not pay a model to guess.

The ask was the one every app ships: snap a meal, get macros, print an accuracy figure. What we built instead: the photo is the member's memory and the coach's context, and the numbers come from structure. The coach's own plan meals, the member's past meals, a public food table, a barcode or a label, and hand portions with one extra question about oil, sauce and drinks, the row the research keeps flagging as the biggest miss.

A member who already pays for their own AI can paste its answer in, and the entry is tagged as their AI's estimate. We publish no accuracy percentage, because it would be a guess about a guess.

Your coach can correct any estimate in one tap, and Marrow remembers it.

The member's food diary on a phone with sample data: a lunch logged from a photo in hand portions, tagged Estimated from photo, with Edit and Delete beside it.
The real food diary, shown with sample data
The ask: coach approves every log

We took the coach out of the loop.

"Way too much work for a coach."

One of the first coaches on Marrow said that about a confirm step on every meal. He was right. So the rule is simple. What a member logs for themselves lands at once, tagged, and the coach may adjust it, never must.

What goes out in the coach's name waits for the coach. A check in reply is drafted from the client's week and sent by the coach, never auto sent, never scored.

The coach's check-in screen on a phone with sample data: a client's week, what went well, what was hard, a note for the coach, and a reply box the coach sends.
The real check-in screen, shown with sample data
The ask: pressure for retention

We kept the alarm off your numbers.

Every retention playbook says pressure keeps people: red numbers, warnings, a count you are afraid to break. The reviews say the other side. Being punished for one late log is a reason people quit.

So a number under the range is never red. It shows how far you are from the range, and nothing about your progress is for sale.

The member's nutrition screen on a phone with sample data: 1,200 calories against a range of 2,100 to 2,400, and protein, carbs and fat each under their range, drawn as calm rings with no red, each with how far it is from the range.
The real nutrition screen, shown with sample data

Every screen above is the real product, shown with sample data.

What we can show

Evidence, not adjectives.

  • Verified

    The founder is a working coach.Spartan SGX, FOCUS Personal Training Institute, StrongFirst Barbell. Fifteen years coaching and partnering with gyms. Verified on Marrow by the same process as every coach. See the verification

  • Standard

    Every coach is checked by a real person.Before a coach is listed anywhere on Marrow, a person confirms the credential with the issuing body. There is no unverified tier.

  • Rule

    We do not claim what we cannot stand behind.No certifications we do not hold. No accuracy figures. No client counts. The same rule covers us and the numbers on your screen.

  • Product

    Every screen on this site is the real product.Shown with sample data, never a mockup. See the coach screens
Pick your plan

One app. Your name on it.

See pricing

Solo Coach $99 a month. Head Coach $349 a month. Gym $500 a month per location, unlimited trainers. Marrow takes 8 percent of the payments it processes.

Coach inside a gym?

The Gym plan covers every trainer at the location, so adding you costs your owner nothing extra. Send them this page.