FlowsBook a callSign inCreate free accountSign up
AI employees and automation

An AI agent for gyms: rival studios, office deals and member questions

What an AI agent for gyms can take off a club manager's desk: new studios opening nearby, offices to pitch corporate memberships, and members' questions at any hour.

Part of the guide AI employees

By the xMap team · · 6 min read

An AI agent for gyms earns its keep on three jobs that a club manager rarely has time for: noticing when a new studio opens within a short walk or drive, finding the offices around the club that could buy corporate memberships, and answering members' questions about classes and opening hours late in the evening. In xMap Flows each job is a small AI employee built from a few blocks, and each one reports to you by email or in the Team inbox.

This piece is for owners and managers of independent gyms, boutique studios and small fitness groups. It shows what each job looks like, a made-up week at one club, the traps to avoid, and the exact blocks to put together.

Why the outside view matters for a club

Most gyms track the inside well: check-ins, cancellations, class fill rates. What they miss is what changes on the street. A boxing studio opening two streets away can pull your early-morning regulars within a month, and nobody tells you until they cancel.

The market is crowded and moves. In Toronto, xMap's places database lists 246 gyms within 10 km of the centre (November 2026):

246Gyms listed
72Belong to a chain
174Independent
4.6Average rating
Gyms within 10 km of the centre of Toronto, xMap's places database, November 2026. The biggest chains among them: F45 Training (23), GoodLife Fitness (15) and Anytime Fitness (5).

Of those with an opening date, 14 opened in the twelve months before November 2026. That is roughly one new gym a month inside one city's core. An independent club cannot watch that by driving around.

Job one: a watcher for new studios

The When something changes trigger looks around up to five of your clubs for a new place of the kinds you choose, within the distance you set, and starts the flow only when it finds one. Pick kinds that really compete for your members: gyms and fitness centres, plus yoga, pilates or boxing studios if you run classes.

Behind it, an xMap LLM block can write two lines about each newcomer from what the database holds (its kind, whether it belongs to a chain, its rating once it has one, its distance from your club), and an Email block sends that to you. The new budget gyms opening near my clubs template is a ready start.

The same trigger can also watch for a rival's rating falling below a line you choose. A well-known studio dropping from 4.6 to 3.9 is often a sign of a change of owner or trainers, and a good week to run a trial offer nearby.

Job two: offices for corporate memberships

Corporate deals bring members in tens, not ones. The question is which offices to call first. Two signals help: offices close to the club, and companies that are hiring, since a company adding staff is a company with a perk budget to spend.

  • Nearby places lists offices and business centres around each club, nearest first, within the distance you pick.
  • Job postings finds companies posting roles near the club in the last week or month, one posting per company.
  • xMap LLM ranks them for fit and writes a one-line reason, and Only new keeps companies you already saw out of the next list.

The offices near my gyms for corporate memberships template builds this list every week.

Job three: a chat page that answers members

Members ask the same things again and again: is there a spin class on Sunday, can I freeze my plan, is the pool open on the holiday. A Chat trigger gives your flow a chat page you can link from your website or app. An AI agent after it answers, and Reply sends the answer back.

The agent only knows what you give it. Upload your class timetable and your membership rules, then set the agent to look things up in those files. Under Protection, list the topics it may answer about, so a question about anything else gets a polite no. The class finder chat for my members template has this shape.

A worked example (made up)

Say you run two clubs in the west of a large city, one near a train station and one in a residential street. You set the watcher to a 1.5 km circle around each, checking every day.

On a Tuesday it finds a new pilates studio 600 metres from the station club. Your email that morning says what it is, that it belongs to a small chain, and that it has no rating yet. You put a free reformer week on your noticeboard before the studio's opening offer reaches your members.

On Monday the office list arrives: twelve companies within a kilometre of the station club, three of them hiring. You call the three that are hiring first. Meanwhile, the chat page has answered forty questions over the weekend, and the two it could not answer (a lost locker key, a refund) went to your front desk by email.

Mistakes to avoid

  • A circle that is too wide. In a dense city, 5 km catches dozens of gyms that none of your members would ever visit. Start with the distance your members actually travel.
  • Treating every chain branch the same. A budget chain and a boutique studio take different members. Let the LLM say which of your offers each newcomer threatens.
  • Letting the chat guess. If prices or holiday hours are not in the files you gave it, the agent should say so and point to the front desk, not invent an answer.
  • Calling the same office twice. Without Only new, the weekly list repeats itself and your team stops reading it.

How to set it up in xMap Flows

Each job is its own flow, so you can switch one off without touching the others. This is the order of blocks for each:

  1. Rival watcher: When something changes (your clubs, the kinds of place, the distance, checked daily), then xMap LLM (two lines on each newcomer), then Email to the manager.
  2. Office finder: Schedule (Monday 07:00), then Area of interest around each club, then Nearby places (offices) and Job postings (roles near the club), then xMap LLM to rank, then Only new, then Email with the list.
  3. Member chat: Chat (page switched on), then AI agent (looks things up in your timetable and rules, topics limited under Protection), then Reply.
  4. Run each once from the editor, read what it would send, then Deploy. The watcher and the office list land in your inbox; the chat answers on its own and you can read every conversation in the flow's Conversations.

For the bigger picture of what these workers are and how they report, see the guide to AI employees. If you would like help mapping this to your clubs, you can book a call.

Questions

Can the watcher cover several clubs at once?

Yes. When something changes watches around up to five places in one flow. For more clubs, make a second flow or a linked copy for each group of clubs.

Will the chat page give out members' details?

No. It only knows the files you give it, and Protection takes email addresses and phone numbers out of answers except the contact details you list for the club.

How does it know a company is growing?

From public job postings near your club. A company posting several roles is usually adding staff, which makes it worth a call about a corporate plan.

Do I need to write code?

No. Each job is a few blocks joined on a canvas, or you can start from the templates named above and change the distance and the kinds of place.

Templates that do this

Read next

More on ai employees and automation · Every article · Templates

Put it to work

An AI employee in xMap Flows does this on live data, on its own timetable. Start from a template or describe it in plain words.