You want your fitness business to run smoothly and still have time for your clients. Yet a large part of your working day goes into looking up information, adjusting plans, processing bookings and preparing the next check-in. You often already know what you want to achieve. Getting it done still takes a series of separate steps.

With MCP and Coach AI, FITsociety is taking the next step: running your business through a conversation. You describe what you need, have the relevant information retrieved and can ask for supported actions to be carried out. You can do this through Coach AI in the coach app, or through your own AI assistant connected to FITsociety via MCP.

This goes beyond having a piece of text written. The AI can work with the context of your business: the client, the existing plan, the calendar or the data needed to answer a question. The aim is practical: move from a question to insight, and from a decision to action, faster, with less searching and manual copying.

We see this as an innovative direction for fitness software. In this article, you will discover what that change means for your work, how the different options fit together and which applications you can explore today. Coaches' experiences from our recent webinar show how they are already putting this into practice.

From operating screens to running your business

Take a client who is going on holiday for two weeks. You want to know their current training plan, what equipment will be available on holiday and what you have previously agreed. You then want a suitable proposal and, after reviewing it, a plan for that period.

That is one coaching question, but the information and the actions involved span several parts of your software. An AI assistant with access to the right data and functions can help you work through those steps together. You start with the result you want and work from there.

The same applies to running your business. Before a meeting, you might want to know which check-ins need attention, what the calendar looks like and which administrative matters are still outstanding. Instead of gathering all that information yourself first, you can request a focused overview and ask follow-up questions.

The value lies in that connection. You can first understand what is happening, then make a judgement and then have the appropriate action carried out. FITsociety remains the environment where your clients, plans and business data are managed. Conversation becomes an additional way to work with them.

By “running your entire business”, we mean that broad use across your work, from coaching to scheduling and administration. The data and actions you can actually use depend on the available functions, your modules and your permissions. This does not mean that AI can already operate every button or complete every process.

Why this is innovative for fitness software

AI can be much more useful when it can work with the current context of your business. General training advice becomes more relevant when the assistant can also take the existing plan and available client information into account. A question about your calendar becomes concrete when the answer concerns your own bookings.

FITsociety therefore brings two options together: AI inside your own coaching environment, and access from an AI assistant of your choice. This lets you choose the approach that suits you. You do not have to build your own software to get started.

You can also provide your own way of working. Think of the structure of a check-in, your rules for progression or the format of a weekly report. This gives the assistant more direction than a single question. Coach Bas's story below shows how valuable that consistent method can be.

For us, this is the innovation: fitness software that helps you work through a complete workflow, starting with what you want to achieve. You explain the outcome, use the information already available in your business and take the next steps with a clear purpose. This creates room for the work where your experience and attention as a coach make the difference.

Where you can use this in your business

The possibilities extend beyond training programmes. Depending on the access you choose, you can bring together information from different areas and carry out supported actions. Here are six familiar areas to help you think about your own applications.

1. Clients and coaching conversations

Have the available information for a conversation brought together: recent check-ups, training sessions, goals and relevant notes. Ask about changes and missing data. This gives you a focused starting point for the conversation. You can then, for example, request a suggested note or next step, where your chosen environment supports this.

2. Training and programmes

Start with the existing plan and describe what is changing: a holiday, different equipment or a new phase of coaching. The assistant can help you develop a reasoned proposal. Supported MCP actions can then also be used to create or assign plans. You assess whether the content suits the client and your approach.

3. Nutrition and progress

Use the available goals and records to prepare a check-in or develop a nutrition plan. Make the intended period and preferences clear. The demonstration below shows how a new nutrition plan is created through MCP and why you also need to check that its days have been filled in.

4. Calendar and bookings

Ask about the client's existing training pattern, find available follow-up slots and have supported bookings added. This helps with recurring appointments and schedule changes. Include the activity, location, time and relevant conditions so that the proposal fits your business.

5. Follow-up and communication

Have information turned into draft feedback, talking points or a follow-up action. Within Coach AI, you can also have supported actions proposed and carried out after confirmation. This connects your preparation with what you want to do next with a client or colleague. You decide what feedback is appropriate.

6. Administration and business insight

You can also include questions about outstanding invoices or other available business data in your preparation. For example, ask for an overview that helps you set priorities. Requesting an overview, drafting a message and carrying out a financial action each have their own capabilities and permissions; asking for insight does not have to trigger a follow-up action.

You can combine these areas into one clearly defined workflow. Think about preparing for Monday: which clients do you want to speak to, what has changed since the previous check-in and what practical matters do you need to arrange? Start with that everyday question and then decide which information you need.

Coach AI, MCP or the API: work where it suits you

You do not have to choose a single technical route for your entire business. FITsociety offers different ways to use data and actions. The main differences are the environment you work in and how much you want to set up yourself.

You use Coach AI in the FITsociety coach app. You ask questions within your own coaching environment and can have supported actions proposed and carried out after confirmation. This is an accessible starting point if you want to keep working inside FITsociety. Explore the options through Coach AI in your coaching environment.

MCP connects FITsociety to an external AI assistant, such as Claude or ChatGPT. MCP stands for Model Context Protocol: a standard that lets an assistant use permitted data and functions. You work in the AI environment of your choice and can also add your own instructions or workflows there. The MCP connection has been available to all FITsociety businesses since 9 November 2026.

At present, MCP lets you carry out a wider range of write actions than Coach AI. If you want to perform a particular action through a conversation, check which route supports it. The capabilities continue to develop. The documentation and the current guidance in your environment can help you choose.

Use the API for a custom integration or automation. For example, when you want to connect an established process to other software. This requires more setup, potentially with a developer's help. The API is the route for a purpose-built integration; MCP makes available functions usable from a connected AI assistant.

An ordinary question in a standalone AI chat only has the information you provide there. With an authorised connection, the assistant can retrieve relevant data directly from FITsociety. This means you do not have to gather and copy that information again for every follow-up question. Your configured access and user permissions still apply.

What you can do with it: three practical examples

During the MCP webinar on Friday 6 November 2026, Thijs Veenhoven demonstrated three applications with a fictional client, Theo. In Claude, he moved from requesting information to processing plans and bookings. The examples show how such a conversation can work in practice. Would you like to watch the recording? Send a request to [email protected].

1. Scheduling extra sessions based on an existing pattern

Thijs first asked which bookings were scheduled for Theo. The assistant retrieved the existing training pattern. It was then asked to add another two weeks of small-group training.

Four new bookings were created during the demonstration. Thijs then checked the days and times in the FITsociety calendar. The assistant had used the previous pattern and location as its starting point.

This immediately shows how important context is. An instruction such as “schedule another two weeks” gains meaning because the assistant can first look at the existing situation. In your own practice, checking the right activity, availability, location and payment or credit arrangement remains part of the process. A single demo does not establish a general rule for how your bookings are charged.

An example of a more careful instruction is:

First review this client's existing training pattern. Find the same activity at the same location for the following two weeks. Show me the available slots and the payment or credit method before adding any bookings.

This is an example prompt that you can adapt to your own situation; it is not a verbatim quote from the webinar.

2. Adjusting a nutrition plan to changing preferences

The second demonstration concerned a new nutrition plan. The fictional client was switching to a vegan diet. Thijs asked for the old plan to be archived and a new plan created with the same goals, taking the recorded allergies into account.

Claude retrieved information from the client profile and developed the plan. The subsequent check revealed something important: the plan existed, but the daily structure had not yet been fully created. As a result, it initially appeared empty in the day view. The days were filled in after a further instruction.

That moment makes the demonstration useful. A confirmation in a chat is a reason to inspect the result. Check that the plan belongs to the right client, covers the intended period and actually contains the days and content you wanted.

Your professional judgement is still needed for dietary preferences and allergies. An AI retrieving and mentioning a point of attention does not prove that every part of its proposal is suitable. Review the products, quantities and other client agreements before using the plan.

3. Creating a holiday plan with different equipment

Thijs then asked for a training plan for a holiday period. The hotel only had dumbbells up to 15 kg and a treadmill. The plan also needed to take account of the points recorded in the fictional client's profile.

The assistant created training days using the available equipment and assigned the plan for the specified period. The demo also covered scheduling: a plan for a future period is not automatically the same plan you see in the app today.

A duplicate assignment also occurred because Thijs had tested the instruction beforehand. This is a useful lesson for repeated requests. Specify whether you want an existing plan adjusted or a new plan created. If in doubt, first ask the assistant to find out what already exists.

For a holiday plan, for example, it is therefore better to describe the period, the existing plan, the available equipment and the desired result. Then ask for a proposal first. This helps you keep track of what the assistant will add and why.

How coaches are already using it

The value of this development becomes concrete in coaches' working practices. During the webinar, Bas, Karina and Marcel explained how they connect AI to their own practice. Their experiences are examples of what you can set up; a personal dashboard or scheduled report is not automatically created for every account.

Bas: start with your own coaching method

Bas Zuiderent spoke about his work in physique and bodybuilding coaching. He works with individual, complex programmes and considers it important that progression and phases align with his own approach.

For him, the connection became particularly interesting when he combined his method with fixed working instructions in Claude. He called those instructions “skills”. Among other things, he set out when someone can progress, when a different phase is needed and which limits he normally applies.

At a new check-up, he then asks the assistant to look at both the client's situation and those rules. The conversation focuses on a specific recommendation: does the proposed adjustment fit his approach, is an exception needed and does Bas actually want the change carried out?

His summary during the webinar was:

“It has saved me a great deal of time, but above all, it has brought consistency.”

That consistency is an important part of his story. A coach often has clear ideas in mind, but it takes attention to apply them afresh to every client. Writing them down gives the assistant a concrete framework for assessing a situation.

A skill like this is an instruction in the chosen AI environment. It does not automatically create a new FITsociety function. The instruction helps the assistant decide how to work with the information and actions available through the connection.

A starting point for your own working instructions might look like this:

  • Progression: which information you review before increasing load or volume.
  • Missing data: when the assistant should ask a question first.
  • Exceptions: which deviations must always be put to you.
  • Feedback: how you want to explain a proposal to a client.
  • Execution: which changes may only be processed after your approval.

Describe these rules concretely enough to explain them to a colleague. An instruction such as “coach like I do” leaves plenty of room for interpretation. Examples of your reasoning give the assistant firmer guidance.

Karina: less searching before a check-in

Karina Sparreboom shared a different use case. Much of her work goes into preparing online and hybrid coaching. Normally, she opens each client's training sessions, check-ins and her own notes to build a complete picture.

Using ChatGPT and the connection, she had built a dashboard that brings this information together. On Monday mornings, this lets her see what she needs for the check-in more quickly. The AI then helps her prepare feedback.

She made her own role clear:

“I do check it.”

In the workflow she described, feedback is therefore prepared and reviewed. Her story is not an example of messages being sent automatically to every client without being checked.

For many coaches, this is an accessible first application. You do not need to build a complete dashboard straight away. You can begin with a standard overview for one client: completed training sessions, the latest check-up, the most important notes and the points you want to explore further.

If that overview fits your conversations well, you can expand it. For example, to multiple clients, a fixed format or a comparison between weeks. Keep the period covered and any missing data visible. A compact summary only becomes useful when you know what it is based on.

Karina also said she was working with the API and Make on an automated start for new clients, including a getting-started guide and a welcome message. She added that this was still being tested. This is her own automation project in development, rather than a standard process that automatically switches on for every coach when MCP is connected.

Marcel: prepare for conversations with scheduled reports

Marcel from Blueprint Bodyplan particularly highlighted being able to work ahead. He described how he can have a report ready for Monday morning at half past eight, containing the information he wants to see for his conversations.

His enthusiasm centred on the room that creates: once again “having time for your clients”.

The idea is familiar. Before a coaching conversation, you want to know what has happened, what questions you have and which client needs extra attention. If the information you need is gathered in advance, you can organise your preparation differently.

You do need to decide what such a report is for. An overview of every available piece of data is rarely the most helpful preparation. Instead, think of a fixed structure: relevant changes, source data, missing information and a few possible talking points.

You set up the schedule for such a report within your chosen AI tool or your own automation. MCP provides authorised access to FITsociety. Whether a task can run independently at a particular time depends on the environment, access and setup you use. During the webinar, Thijs noted, for example, that some workflows require an active computer or additional permission.

Test a scheduled workflow at the time you intend to use it, too. Check whether the data is retrieved then and where you can find the result. This turns a recurring report into a workable part of your week.

Six instructions to get started with

Below are six newly written example instructions that you can adapt to your own practice. They are intended as a starting point; execution depends on the available tools and the access you have configured.

Prepare for a coaching conversation

Prepare an overview for my conversation with this client. Review the available training sessions, check-ups and relevant notes from the past four weeks. List the main changes, missing information and three questions I can ask. State what each point is based on. Do not make any changes.

This instruction combines a defined period with a clear result. You receive talking points that you can review and use in your own preparation.

Find clients with possible points to address

For this group of clients, review the available training and check-up data from the past two weeks. Show where information is missing or the pattern differs from the previous period. Distinguish between a missing entry and a deviation that has actually been recorded. Give me an overview to review first.

That distinction matters. A missing entry may mean something has not been filled in. By itself, it does not tell you why. Use the overview to prepare more focused contact.

Support a proposed training adjustment

Compare the latest check-up with the current training plan and my working instructions. Describe the adjustment you would consider, why and which information you still need. Leave the current plan in place until I have reviewed your proposal.

This follows the use case Bas described. You make the reasoning visible before deciding to change anything. Add the relevant rules from your method to your own instructions.

Extend an existing schedule

First find this client's training pattern and check whether the desired follow-up sessions have already been booked. Show me the options that are still missing, including date, time, location and the relevant booking conditions. Add only the bookings I subsequently confirm.

Checking existing bookings first reduces the chance that a repeated instruction will create a second set of appointments. Check the result in the calendar afterwards.

Check a nutrition plan for completeness

Check that the new nutrition plan belongs to the intended client, covers the right period and has been filled in for the required days. Identify any missing parts and the choices I need to review professionally. Leave existing plans unchanged.

This example follows the moment in the demo when a plan existed but its days still needed to be created. Ask about the concrete result that the client should ultimately be able to use.

Get a clear overview of outstanding invoices

Provide an overview of outstanding invoices within my permitted scope. Group them by client and show the relevant dates and status. Do not send messages or perform payment actions; I want to review the overview first.

Administrative information can also be part of your preparation. Keep the intended result clear: an overview, a proposal and an actual action are different instructions.

What can you configure for access and write permissions?

The MCP connection has its own defined access. You decide which parts of FITsociety are available and whether the assistant may only read data or may also carry out certain changes.

For example, you can start with client information and bookings if your main aim is to prepare your calendar and coaching conversations. Other applications may need access to other areas. Health data and private communication require additional permissions.

Write access is a separate choice. A connection with read access does not automatically gain the ability to adjust a plan or create a booking. The relevant capabilities must be explicitly permitted for supported write actions. Your own business and user permissions continue to apply as well.

This gives you room to start small. You can first check whether a summary is correct. You can then add an appropriate write action and test it in a controlled way. A more extensive instruction does not automatically require access to every area.

You can also revoke access. The guide to areas, write permissions and revoking a connection explains how to manage these choices.

How to start your first workflow

Choose one recurring task where you can easily assess the result. Preparing for a familiar client often makes a better first test than an instruction that immediately changes all kinds of data across your business.

  1. Open AI connections in your business settings. Use the connection details and guidance shown there by FITsociety. If you do not have access to these settings, have the connection approved by the authorised administrator.
  2. Connect your AI assistant. Follow the instructions for your chosen application and complete sign-in and authorisation. The exact capabilities can differ between AI providers and accounts.
  3. Select the access you need. For example, start with a limited, read-only connection. Add other areas and write actions when your workflow needs them.
  4. Ask a clearly defined question. Specify the client or group, the period, the desired result and what the assistant should do if data is missing.
  5. Check the answer in FITsociety. Compare a few results with the source data. Also check that the assistant used the same period and the same client.
  6. Then expand. Add fixed working instructions or test one write action. Record which parts you want to keep confirming yourself.

For the specific screens and connection options, follow the guide for Claude or ChatGPT. The general guide to MCP and the API provides more background.

You give the work direction

A useful result starts with a clear instruction: which client or group are you referring to, which period do you mean and what do you want to see? Also specify whether you want a proposal, an existing item adjusted or something new created.

The questions during the webinar underlined this. Kadir asked about the desired structure of a nutrition programme: separate templates and one continuous plan are different outcomes. Mark mentioned answers that seemed to use different assumptions about coach availability. Ask the assistant to state the relevant assumptions so you can ask follow-up questions about the answer.

Not every action is available yet, and AI can misinterpret something. Check important changes in FITsociety, such as a booking or an assigned plan. Also ask the assistant to say which data is missing or which action it cannot perform. This helps you move forward with a clear purpose, without mistaking a plausible-sounding answer for a completed action.

Less routine work, more attention for your clients

The examples differ, but the aim is familiar. Bas wants to apply his method consistently. Karina wants to get a complete picture for a check-in more quickly. Marcel wants his preparation ready on time. In all three cases, the setup helps organise recurring work more effectively.

This is also the opportunity for your whole business. When you spend less time searching, copying information and repeating the same steps, you can devote more attention to conversations, decisions and the quality of your coaching. How much difference that makes depends on your workflow. Choose a task that comes up regularly and compare your preparation and results with how you do it now.

You do not have to change your entire way of working today. Start with one useful application, make it reliable and build from there. The combination of your expertise, the data in FITsociety and an assistant you can direct makes this interesting both for everyday coaching and for the further development of your business.

Is there a capability you are missing in MCP, Coach AI or the API? Submit a request through the FITsociety wishlist. Describe the task you want to carry out, where you currently lose time and the result you want to see. Concrete applications like these help us expand the capabilities further.

Frequently asked questions

Can I run my entire fitness business through AI?

You can use AI across different parts of your business, from coaching and scheduling to administrative insights. The actions you can actually carry out vary by route, supported function and configured access. Describe your desired workflow and check which steps are already available. For now, you will still carry out some parts through the regular interface.

Do I need to know how to code?

You do not need to develop software to use Coach AI or connect a supported MCP assistant. You follow the connection steps and ask questions or give instructions. A custom dashboard, advanced scheduling or an API integration may require additional setup, potentially with a developer's help.

Can the assistant immediately change everything?

Your user permissions and the access you choose determine what is available. With MCP, you select areas and write actions; you can also permit read-only access. For supported changes, Coach AI works with proposals and confirmation. Start with access appropriate to the task you want to carry out.

Can I automatically receive a recurring report?

This requires an AI environment or automation that supports scheduling and has access at the intended time. MCP provides the connection to FITsociety. Connecting MCP does not, by itself, enable a weekly report. Test your chosen setup before relying on it in your schedule.

Where can I read more or watch the webinar?

Read the MCP and API documentation, the help centre guide or open Coach AI in your coaching environment. The webinar recording is not publicly available. Would you like to watch the recording, which runs for approximately 41 minutes? Email [email protected] to request access.