ProtocolCRM

The challenge

In early 2025 we interviewed successful online fitness coaches about how they actually run their businesses. The answer was ten or more disconnected tools: Calendly for scheduling, Zoom for sessions, MyFitnessPal for nutrition, Trainerize for programming, Stripe for payments, ActiveCampaign for email, Google Sheets for progress, WhatsApp for client chat, Typeform for intake, and whatever else had been bolted on that quarter.

The cost was not the $400–600 a month in subscriptions. It was that coaches were spending 60–70% of their time on administration instead of coaching, and that their clients experienced the business as a scattered set of logins rather than a premium service.

One coach summarised it: he became a fitness coach to change lives, not to be a part-time IT administrator.

What I built

Rather than adding another tool to the stack, we asked what the platform would look like if you started the online coaching industry from scratch in 2025, with today's AI capabilities available from day one. As co-founder and CTO I led both the technical strategy and the hands-on build.

AI as the foundation, not a feature. Competitors bolt AI onto systems designed before it existed, which caps what it can do. We built the AI infrastructure as a core component of the architecture. The program builder generates personalised training from 50+ parameters in seconds. Meeting prep analyses a client's recent data and produces talking points before the call. Check-in analysis processes client updates and flags concerns the coach should look at. Retention analytics identify at-risk clients before they churn rather than after.

One platform, one login. A unified dashboard replaces the ten tools: a full CRM with visual sales pipelines, integrated scheduling with native video conferencing, program and nutrition planning with macro tracking, progress tracking across 100+ data points, wearable device integration, and payment processing. Data flows between all of it because it is one system, not an integration layer pretending to be one.

White-label mobile as standard. Fully native iOS and Android apps carrying the coach's branding, plus a branded client portal. No "powered by" badges. For a coach charging premium rates, the client-facing experience is the product.

Healthcare-grade infrastructure. Cloud-native multi-tenant architecture on AWS, with isolated environments per coach on shared infrastructure, an API-first design for future integrations, and HIPAA-compliant handling of health data. That last choice is a bet: fitness coaching and preventive healthcare are converging, and platforms that cannot handle clinical data will be locked out of that market.

We ran it in five phases: domain research, architecture, core platform, AI integration, then beta with four elite coaches whose real-world usage drove the refinement before wider launch.

Results

Early-access coaches report:

• 15–20 hours per week saved on administrative work
• Consolidation from 7–10 tools down to one
• $200–400 per month saved on subscriptions
• Materially better client satisfaction from a single branded experience

Projected from beta usage: a 70% reduction in administrative time, 2–3× the client capacity per coach, 40–60% lower tool costs, and improved retention as a second-order effect of the better client experience.

Market signal has been strongest among coaches already doing $3K+ in monthly revenue, the segment where administrative drag is the binding constraint on growth rather than demand.

What I took from it

Consolidation plays are only defensible if the integration is real. Ten tools behind one login is a worse product than ten tools, because you inherit every seam. Being one system is what makes the AI layer possible: the program builder is only useful because it can see the check-ins, the wearable data and the session history at once.

The project is ongoing. Current focus is onboarding the founding cohort, building wearable partnerships, and exploring clinical and corporate wellness applications, which is where the HIPAA decision starts paying for itself.