Eighty active 1:1 clients. Twelve hours a week of WhatsApp DMs. A coach burning out on the admin work that used to feel like care. Then Maya — a Squidgy-built accountability agent — absorbed the volume in three weeks and added a recurring revenue stream nobody planned for.
The starting point
Sam (we're anonymising the coach) runs a strength practice with 80 active online clients on £200/month programs. Standard cycle: programming on Sunday, weekly check-in DMs on Wednesday and Sunday, ad-hoc "quick form check" messages every day. The admin alone — not the coaching — was eating twelve hours a week.
The check-ins were the part that broke. Each one was three or four messages: how did the session feel, any pain, what's the bar speed, how's sleep, what about diet compliance. Multiplied by 80 clients, twice a week, every week. Sam was doing the same conversation 160 times a week.
What Maya does
Maya pings each client three times a week post-session. Asks the same five questions Sam would ask. Logs responses. Flags concerns: missed sessions, pain reports, sleep dropping below five hours, drift in compliance. Does notgive advice — that's Sam's job. Maya's job is to gather, surface, and escalate.
Sam reviews Maya's dashboard once daily — about 20 minutes — and acts on flags.
I genuinely thought my clients would hate it. Half of them haven't noticed it's not me. The other half explicitly told me they prefer the consistency.
— Sam, three weeks in
The build
Sam spent about an hourin conversation with Ace (Squidgy's build agent) describing what Maya should do, what tone to use, what to escalate. Approved the design, connected to TrueCoach (Sam's programming software) and Stripe (for client billing sync), and Maya was live the same afternoon.
Voice calibration took longer — about a week of edits as Sam reviewed Maya's outgoing messages and tweaked. By week two, Maya's tone was indistinguishable from Sam's.
The numbers, 90 days in
12 hrs
Saved per week
£4.3K
MRR added (own clients)
+4
New 1:1 clients taken
Sam priced Maya into existing client packages as a £20/month uplift — £49 → £69 for the check-in tier. 80 clients × £20 = £1,600/month new revenue from existing clients alone.
With twelve hours a week reclaimed, Sam took on four more 1:1 clients (£500 ARPU) = £2,000/month additional. Total recurring revenue uplift from Sam's own practice: £3,600/month.
The unexpected part
About sixty days in, two other strength coaches in Sam's circle asked if they could use Maya for their own clients. Sam listed Maya in the Squidgy marketplace under coaching, white-labeled (each coach's clients see theircoach's brand, not Sam's).
Twelve other coaches subscribed within thirty days at £499/month per coach. £6,000/month additional recurring, with zero additional client-facing work for Sam.
Total Maya revenue, ninety days in: £10,300/month recurring, of which £6K requires basically nothing from Sam beyond reviewing Maya's flagged messages once a day across all the licensee accounts (about 30 minutes total).
What Sam learned
- Voice calibration matters more than feature scope. Get the tone right and clients won't notice. Get it wrong and they will, immediately.
- Approval workflow earns trust fast. For the first month, Sam reviewed every Maya message before it sent. By week three, spot-checking was enough.
- The white-label opportunity wasn't planned. It came from sharing what worked. Now it's the highest-margin part of the revenue.
- Customers paid more for less of Sam, not less. The £20/month uplift wasn't resented — it was rewarded with better retention.
Could you build this?
If you have a vertical practice with repeated client touchpoints — coaching, consulting, agency, professional services — yes. The pattern Sam used (an agent that gathers, logs, and escalates without giving advice) maps cleanly to almost any expertise-driven practice.
The harder part isn't building Maya. It's deciding which forty hours of your week could be done aboutas well by an agent that's seen your every move for the last six months. Most operators have an answer to that question. Most just haven't been asked.