Once you've connected your gym with AI Connect, you can ask Claude business questions in plain language and get answers pulled from your own gym's data in Kilo GMS. This article gives you example questions to start with, grouped by what you're trying to figure out, along with tips for reading the answers.

You'll need an active AI Connect connection before you can ask any of these questions. If you haven't set one up yet, start with Set Up AI Connect in Kilo GMS.

What can I ask Claude with AI Connect?

You can ask Claude about:

• New members who haven't attended a class yet

• When members cancel and which packages lose the most members

• Packages that are expiring without a renewal in place

• Revenue at risk from expiring packages

• No-shows and late cancels by class type and by coach

• Revenue by class type, by package, and per member

Every question below can be copied into Claude as written. Change the time periods and numbers to fit your gym.

Questions about new member activation

These questions help you spot new members who signed up but haven't attended a class.

• Of the members who joined in the last 90 days, how many have never attended a class, and how long have they been signed up?

• Compare new-member activation for the last 30 days vs. the prior 30. Is time-to-first-class getting better or worse?

• Flag anyone who signed up more than 7 days ago and still hasn't booked a class.

You can change the 7-day window to match how quickly you expect new members to get started.

Questions about cancellations and churn

These questions help you see when members leave, which packages lose the most members, and whether members had already stopped attending before they canceled.

• Where do cancellations cluster by tenure week this year?

• Which packages have the worst cancellation rate, and how many members actually hold each one?

• Did the members who canceled this year still attend in their final two weeks, or had they already gone quiet?

• How does churn in the last 90 days compare to the previous 90?

Questions about expiring packages and revenue at risk

These questions build a save list of members whose packages are ending without a scheduled renewal, so you can reach out before they lapse.

• Who has a package expiring before November 15 that isn't set to auto-renew? Bucket them by urgency.

• Which of my highest-lifetime-value members have expiring packages or have gone quiet?

• What's the revenue at risk from packages expiring in the next 21 days?

Questions about class and coach performance

These questions show where no-shows and late cancels happen most.

• Which class types have the worst no-show and late-cancel rates over the last 90 days?

• Are any coaches' classes showing unusually high no-show rates?

Detailed report prompts

These longer prompts ask Claude for a full report in a single question. Copy a prompt into Claude and replace anything in brackets with your own dates or numbers before you send it.

Class revenue performance

Using my Kilo data for [date range], generate a class revenue performance report. For each class type, show:
Total revenue generated
Number of sessions held
Average revenue per session
Utilization rate (reserved / available spots)
No-show and late cancel rate
% change vs. the prior period

Rank classes from highest to lowest total revenue. Flag any class with a utilization rate below 60% or declining revenue trend. Suggest which classes to expand or cut based on this data.

Cancellation patterns

Pull all athletes who cancelled in the last [30 / 60 / 90] days from my Kilo data. For each cancellation, I need:
Package type and how many weeks/sessions they completed before cancelling
Whether they had attended in the 2 weeks prior to cancellation
Time between their last class and their cancellation date

Then summarize:
The most common drop-off points (e.g. week 3, week 8)
Which package types have the highest cancellation rates
Whether there's a pattern by class type, day of week, or coach

End with 3 specific retention actions I can take based on these trends.

Packages expiring without a renewal

From my Kilo dashboard, show me all athletes who meet ALL of these conditions:
Have an active package right now
Have no scheduled renewal or billing contract on or after [date]
Have not been flagged as cancelled or on hold

For each athlete, include: package name, package end date, days until expiry, last class attended, and total revenue to date.
Sort by days until expiry (soonest first). Group into urgency tiers:
Critical (< 7 days)
At Risk (7–21 days)
Monitor (22–45 days)

Recommend a re-engagement message for each tier.

Time to first class for new members

Using my Kilo data, calculate time-to-first-value for all new members who joined in [month/quarter]:
Days from sign-up to first class attended
Days from sign-up to first renewal

Show averages and also flag members who signed up more than [14] days ago and have NOT attended a single class yet. Include their package type.
Compare these numbers to the previous period. If time-to-first-class is increasing, flag it as a risk. Suggest one operational change to reduce it.

Missed revenue from uncredited reservations

From my Kilo data for [date range], calculate the missed revenue from uncredited reservations:
Total reservations that resulted in a no-show or late cancel
Of those, how many were NOT charged a cancellation fee
Estimated missed revenue (use the average class value or package rate)
Which class types or time slots have the highest no-show rates
Which member segments (package type, membership length) no-show most often

Present results as a simple table. End with a recommendation: based on the dollar amount, is it worth enforcing a stricter cancellation policy?

Monthly revenue breakdown

Using my Kilo data for [month or date range], give me a full revenue breakdown:
Total revenue month-to-date, split by package category
Top 5 packages by total revenue generated
Average revenue per paying member (ARPU)
% of revenue from new vs. returning members
Upcoming scheduled revenue (next 30 days)

For each package category, show: # of active members, avg. price, total revenue, and % of total gym revenue.

Flag any package category where revenue dropped more than 10% vs. last month. Suggest whether to promote, reprice, or retire underperforming packages.

Frequently asked questions

Can Claude see my revenue if it can't see billing information?

Yes. Claude can report on revenue, but it never sees payment details. Payment methods, card numbers, and bank account details are never shared, so Claude can't tell you which members pay by card or by bank account, or find a member by the last four digits of a card.

Why didn't Claude return any results for my coaches?

Coach results depend on payroll being generated for the time period you asked about. If no payroll report exists for that period, Claude has nothing to return, which is different from your coaches having zero sessions. Generate the payroll report for that period in Payroll → Payroll Report, then ask again.

Why does one package show a 100% cancellation rate?

A cancellation rate depends on how many members hold the package. If only one member has ever held a package and that member canceled, the rate is 100%. Ask Claude how many members hold each package so you can tell a real trend from a single cancellation.

Can Claude give me my members' phone numbers or email addresses?

No. Phone numbers are never shared, and member emails are masked. Claude sees the first two characters of an email followed by asterisks and the full domain. Member names are shown in full, so you can look up contact details for anyone on a list directly in Kilo GMS.

Can I ask Claude to cancel a package or update a membership?

No. AI Connect is read-only. It can return information about your gym but can't create, update, or cancel anything. Make those changes directly in Kilo GMS.

Can I ask about more than one gym in the same question?

No. Each connection covers one gym. If you manage several gyms, each one needs its own connection, and Claude answers from the gym that connection was authorized for.

Helpful tips

• Include a time period in every question, such as the last 90 days or this year

• Check how many members a percentage is based on before acting on it

• Watch results based on only a few members over a longer period before you act on them

• Ask follow-up questions to narrow a list or change the time period

If you have questions about AI Connect in Kilo GMS, our Support team is happy to help. Reach us at hello@usekilo.com.