Fitness & Studios · Personal Trainers & Solo Coaches

AI Agent for Small Group Training

Fills semi-private slots, keeps each group above the number that makes it worth running, and warns you early when one is falling apart.

Start from this template
Edit it — the agent is built from this briefBuild this agent
How it works
1 Step
Watch the count
2 Step
Fill the gaps
3 Step
Escalate the decisions
The agent tracks bookings for every group session against the minimum that makes it worth running and the maximum you can coach.

Overview

Semi-private only works at the right number, and the number moves weekly.

An AI agent for small group training manages the thing that makes semi-private coaching profitable or pointless: how many people are in each session. A group of four at a reduced rate earns a trainer considerably more per hour than a one-to-one, which is why the model exists. A group of one earns less than a one-to-one and costs the same hour to deliver. The difference between those outcomes is a handful of individual bookings each week, currently managed by a trainer messaging people between sessions and hoping. The economics are unusually sharp and the administration is unusually fiddly, which is a bad combination for somebody who is coaching six hours a day.


Capabilities

What the Small Group Agent does

Keeps group numbers where the model actually works.

01

Tracks who is booked into each group session against your minimum and maximum

02

Offers open places to clients who have asked for that time

03

Warns you early when a session is heading below the viable number

04

Handles the drop-out and tries to refill before the session runs

05

Suggests merging or moving a session that is consistently under-filled

06

Explains the group format and pricing to anybody asking about it

Why you should use the Small Group Agent

Small group training is often recommended to trainers as the way out of the hours-for-money trap, and it works, but only when the sessions are full. The failure is not strategic — trainers understand the arithmetic perfectly well — it is operational. Keeping six groups at four people each requires constant small adjustments, and every one of those adjustments has to happen while the trainer is mid-session with somebody else. Sessions therefore run at two or three because filling them takes attention nobody has, and the model gets abandoned as not working when what was not working was the admin.

Before
Groups run at two or three because filling them takes attention nobody has
A drop-out on the day is discovered when three people arrive instead of four
Open places are offered ad hoc, to whoever the trainer happens to think of
Consistently empty sessions run anyway because nobody reviews them
The model gets written off as unprofitable when the problem was scheduling
After
Open places are offered systematically to the people most likely to take them
A drop-out triggers a refill attempt while there is still time
You know on Monday which sessions are short for the week
Persistently under-filled sessions are surfaced for a decision
The model runs at the numbers that made it worth doing
Process

How it works

A three-step flow that watches the number in each session.

Step 01

Watch the count

The agent tracks bookings for every group session against the minimum that makes it worth running and the maximum you can coach.

Step 02

Fill the gaps

It offers open places to clients who want that time or have sessions left on a block, rather than announcing to everybody at once.

Step 03

Escalate the decisions

When a session stays below the minimum it tells you in time to merge, move or cancel it rather than running it at a loss.


Example

Example workflow

A realistic week across six semi-private slots.

Scenario: a trainer running six semi-private sessions a week at a maximum of four was averaging under three per session and considering dropping the format. On Monday morning the agent reports the week: four sessions full, Wednesday 6am at two, Thursday 7pm at three. It offers the Wednesday places to the two clients who have asked for early mornings and to one with block sessions expiring; one takes it, bringing the session to three. On Wednesday at 5:10am somebody drops out with a sick child. The agent offers that place to the Thursday group's waiting client, who takes it at 5:25. Thursday evening still sits at three by Tuesday, so the agent flags it, and the trainer merges it with the Friday group rather than coaching three people for an hour at group rates.

Missed-Call & Booking Automation Google CalendarTwilio SMSStripeMindbody AI Agent flow

Audience

Who can benefit

Coaches trying to earn more per hour without adding hours.

✍️ Self-employed personal trainers

Semi-private is the main route out of trading hours for money, and it fails on admin.

💼 Trainers with a full one-to-one diary

Group sessions are the only way to grow once every hour is already sold.

🧠 Studio owners renting to coaches

A group session that runs at two occupies the same space as one that runs at six.

Outdoor and park-based trainers

Weather-driven drop-outs make the numbers move more than in any other setting.

🎯 Coaches running programs in blocks

A group formed for a six-week program has to hold together for six weeks.

📋 Trainers testing a new time slot

Knowing quickly whether a slot fills is the difference between a trial and a habit.

Integrations

Watches group bookings and moves people between them.

Google Calendar

Holds the group sessions and the individual bookings inside them.

Twilio SMS

Offers open places and handles drop-outs, both of which need answering within minutes.

Stripe

Charges the group rate or deducts from a block when somebody takes a place.

Mindbody

An alternative system of record for coaches running groups inside a gym or studio.

Instagram Direct

Answers questions about the group format from people who are not clients yet.

Google Sheets

Tracks average attendance per slot, which is the number that decides what to keep running.

Applications

Best use cases

Where group numbers decide whether the hour was worth it.

Sessions sitting one or two below the viable minimum
Same-day drop-outs from a small group
Clients who have asked to be told about a particular time
New slots being trialled to see whether they fill
Six-week programs where the group must hold together
Deciding which under-filled sessions to merge or drop

FAQ

FAQ

Questions about running small groups at viable numbers.

An AI agent for small group training manages the number of people in each semi-private session: it tracks bookings against your minimum and maximum, fills open places, handles drop-outs and flags sessions that are not viable.

Whatever number makes the hour earn more than a one-to-one at your rates, and it is worth calculating rather than assuming. Many trainers run groups at a size that is worse for them than coaching one person, without having done the arithmetic.

Sometimes, for a new slot you are building or a group you do not want to disrupt. What matters is that it is a decision rather than a default, and that you can see how often it happens.

From rules you set — clients who asked for that time, clients with sessions expiring, clients whose usual slot was canceled. Announcing an open place to everybody produces a scramble and several disappointed people.

Yes, and that is a different job: keeping the same people attending for six weeks rather than filling a rolling slot. Missed weeks matter more, and the flag should come earlier.

It is arguably most useful there, because weather moves the numbers unpredictably and often on the morning. Handling a wet Tuesday well is the difference between a group that survives winter and one that does not.

Use it to decide the timetable. Average attendance per slot over a couple of months tells you which times genuinely work in your area, which is a far better basis for scheduling than the times that suited you when you started.


AI Agent for Small Group Training

Fills semi-private slots, keeps each group above the number that makes it worth running, and warns you early when one is falling apart.

Start from this template
Edit it — the agent is built from this briefBuild this agent