Most lifetime value arguments are really data arguments wearing a mathematical costume. Two people look at the same figure, disagree, and discover after twenty minutes that one of them counted gift card redemptions as revenue and the other did not.
This checklist exists to close those gaps before the calculation starts. Work through it once in full, then keep a short version for each period you report.
Start with the definition of a client
- Write down what makes someone a client. A completed paid visit is the usual answer; a consultation that was never booked is not.
- Decide how to treat a household that shares one phone number. One record with several people, or several records, but the same rule every time.
- Decide how to treat a client who transferred from another location in the same group. New client here, or continuing client for the group.
- Record the date of first visit as a real field, not as something reconstructed later from the earliest appointment on file.
The first-visit date is the spine of the whole exercise. If it is missing, cohorts cannot be built and the benchmark becomes a guess about the present.
Fix the period and the revenue basis
- Choose the period and write both end dates down, including whether the final day is included.
- Choose revenue or gross margin, once.
- If margin, decide how consumables are allocated to a service. A per-service estimate is fine as long as it does not change mid-year.
- Decide where discounts land. Reducing revenue at the time of the visit is the cleaner choice; applying them later means the figure has to be restated.
Audit the visit records
- Look for duplicate client records created by a new phone number or a changed surname.
- Confirm that cancelled and no-show appointments are excluded from the visit count.
- Confirm that appointments moved to a later date are counted once, at the date they happened.
- Check refunds and credits, and decide whether a partially refunded service stays in revenue at the amount collected.
- Confirm that visits recorded before your system went live are either imported or explicitly out of scope.
- Check whether staff appointments, model calls or training sessions were booked under a client record. They are visits in the calendar and not clients in the business.
Separate service from retail
- Keep service revenue and product revenue on separate lines.
- Treat gift card sales as a liability until redeemed, then record the redemption against whatever was purchased.
- Decide whether package sales are recognised at purchase or spread across the visits they cover. Spreading is more accurate and more work.
Mark dormant and single-visit clients
- Choose the dormant threshold per service group, not across the whole menu. A brow appointment and a colour correction return on different cycles, and one number will mislabel one of them.
- If you would rather keep a single figure, take the longest cycle on the menu and accept that faster-returning services will look healthier than they are.
- Flag clients with no visit past the threshold, and apply the same threshold to every cohort.
- Keep single-visit clients inside the cohort. They are the honest bottom of the distribution.
- Note how many clients in the cohort have moved away or asked to be removed. They are neither retained nor lost in a way you can influence.
Sign off before comparing
- Have one person confirm each of the four choices: client definition, revenue basis, period, dormant threshold.
- Save the definitions beside the file, not in a message thread.
- Only then run the comparison.
Elevoi helps the inputs stay clean at the source. Every call it answers produces a record with the caller's name and number, the appointment lands in the live calendar, and SMS confirmations sit against that booking, so the visit either appears in the export or it never existed as a booking at all. Urgent calls are escalated to a person rather than left as a note.
When the list passes, move to the template and fill it in, or read the calculator page for how the divisions are applied.