LearnXyziosXyzios

Concept

Studio metrics, defined

The numbers a studio runs on, from members and churn to fill and recovered revenue: what each one means, how to count it, and where Xyzios shows it.

The same studio can report three member counts in one week, from the booking platform, the accountant and a marketing deck, and all three are honest: they answer different questions. This page writes each number down once, the way Xyzios counts it: what it means, the formula, where it lives on the boards and the cases at the edges. Where a number is your own arithmetic rather than a figure on a board, it says so.

A few counts follow rules your studio sets with Xyzios: whether a paid-in-full pass makes a member, whether comps count, whether a pause counts as churn, how long a lead has to convert and when an intro’s clock starts. To change one, use Report issue in your account menu, and the boards follow the new rule from their next refresh. How the boards use everyday words, such as a dash or · now, is in The words Xyzios uses. The full list of what the boards count, department by department, is in What Xyzios measures.

Members

Member

Someone holding a membership that has started. Members are counted as people, so someone holding two memberships is one member.

Formula: members = people holding a started renewing membership or paid-in-full pass, each counted once, under your studio’s rules

In Xyzios: Total members on the Overview and Total members · now on Sales are paying members today, with comps and paused members counted beside them, not in them. Which product is which is decided on Classification.

  • Renewing memberships bill on a schedule and count toward MRR. A class pack that renews itself is classified here too.
  • Paid in full (a term pass, a year paid up front) makes a member but never MRR. That is the default; your studio’s rules can leave these passes out of the count.
  • Class packs and drop-ins make customers, not members. Intro offers and first-class trials are not memberships either.
  • Comps (staff, instructors, friends and family) stay out of member counts unless your studio’s rules count them. Their revenue stays in the revenue figures either way.
  • Partner products (ClassPass, Wellhub and the like) stay out of member counts.
  • Paused members are counted apart; see Paused member below.
  • Bought, not started: on Mariana Tek, a membership that starts at the first class is not a member until it starts, and booking a first class starts it. Those people wait on Call Lists under Membership bought, not started yet.

Active member

A member whose membership is running on a given day: started, not ended and not paused.

Formula: active members on a date = people holding a started, running, unpaused membership on that date

In Xyzios: a figure marked · now is today’s count. The Overview’s Members trend and Member growth by studio on Expansion draw paying members at each month end, and Recap shows Active members for the period you pick.

  • A count without a date is not a count: “active” is always as of a day, so compare two counts taken on the same day under the same rules.
  • Months before your studio joined Xyzios are rebuilt from membership dates and can read slightly high; live months come from counts taken each night.
  • How pauses are counted settles many disagreements between two active-member numbers, which is why paused members get a column of their own.

Paused member

Someone with a paused (frozen) membership and no running one: neither paying dues nor gone.

Formula: paused members = people holding a paused membership and no running one, each counted once

In Xyzios: a group of their own, beside paying members. Total members on the Overview shows frozen beside the paying count, Total members · now on Sales says how many are paused, Compare has a Frozen column and KPIs a live count, Suspensions to date. A member’s page on Members reads Frozen, with the freeze’s end date where the booking platform gives one.

  • Someone with one membership paused and another running is a member, not paused.
  • By default a pause is not churn: someone paused has not left, and is counted beside the paying members until they come back. Your studio’s rules can count a pause as a cancellation instead.
  • To judge your freeze policy, count how many pauses end with the member back in class within 30 days of the return date. That is your own calculation.

Net member change

How many more or fewer members a studio has at the end of a period than at the start.

Formula: net member change = new members − cancellations

In Xyzios: Net members on KPIs is new autopays less cancellations over the window you pick, with the number paying now beneath. Member growth · MoM on Expansion sets paying members at the end of the last closed month against the month before. New members · MTD on the Overview gives the net figure beneath the joins.

  • The balance form, members at the end − members at the start, can move further than joins and cancellations do: a cancellation counts when it is made rather than when the contract ends, and someone rejoining is not a new autopay.
  • New members on Sales counts joins only, as people rather than memberships sold, with renewals left out.
  • It is growth, not retention: read it beside the retention rate, never in its place. A month that adds forty and loses thirty grew by ten.

Retention and churn

Churn rate

The share of members who cancelled during a month, measured against the members who could cancel when it began.

Formula: churn % = people who cancelled a renewing membership during the month ÷ renewing members at the start of the month × 100

In Xyzios: Data reads Monthly churn for the last closed month, with the three-month average and a six-month line. Member churn on the Overview grades the last closed month against your own prior three months. Compare reads the window you pick, a month at a time, and Recap calls it Attrition. The start-of-month count is the one recorded when the month before closed.

  • A cancellation counts in the month it was made, not the month the contract ends.
  • A renewing membership with a fixed term that ends with no next plan bought or begun within two weeks counts as a cancellation in the month it ends. One that carries on into the next plan is neither a cancellation nor a new member.
  • Paid-in-full passes are out of the base: they run out rather than cancel, so counting them would flatter the rate.
  • A pause counts only if your studio’s rules count a pause as churn.
  • A month to date reads near zero early in every month, which is why the boards read the last closed month.
  • It compounds. For example, 5% a month keeps about 54% of today’s members after a year (0.95 to the twelfth power), not the 40% that 5% × 12 suggests.
  • Over a quarter or a year, compare monthly churn. Cancellations over a long window against the members at its start count everyone who joined and left inside it, so that figure grows with the window.

Involuntary churn

Members lost to a failed payment, such as an expired card or a bank decline, rather than to a decision to leave.

Formula: involuntary churn % = members lost to failed payments during the month ÷ renewing members at the start of the month × 100

In Xyzios: no board splits churn this way, so this rate is your own calculation. The boards show the live side: Failed payments on KPIs (members whose payment is failing right now), Payment failures on Compare and Payment failure in Member status mix on Sales. A member whose payment fails gets a payment recovery draft in Approvals when that draft is switched on; see Payment recovery below.

  • Count it in its own column. Mixed into churn, it hides the cheapest recoveries in the building: members who meant to stay.
  • A failed payment is not a cancellation. Net members on KPIs does not subtract failed payments, because a decline that becomes a cancellation would count twice.

Retention rate

The share of the members you started a period with who are still members at the end of it.

Formula: retention % = (members at the end − members who joined during the period) ÷ members at the start × 100

In Xyzios: this form is your own calculation. The boards show churn, above, and 90-day retention, below.

  • Take the joiners out, or a busy month can read as better than perfect. For example, 240 members on the 1st and 242 on the 31st reads as 100.8% until the 18 who joined come out: (242 − 18) ÷ 240 is 93.3%.
  • Paid-in-full passes, comps and pauses move the answer more than the formula does, so count both ends under the same rules.

90-day retention

The share of one group of joiners who are still members 90 days after they started.

Formula: 90-day retention % = renewing members who started 90 to 180 days ago and were still members on their 90th day ÷ renewing members who started 90 to 180 days ago × 100

In Xyzios: Data shows it, and Compare sets it side by side by location. When no group is old enough yet, Data estimates it from churn and says est. from churn.

  • It is the truest retention number because nothing else gets into it: no joiners from other months, and no long-standing regulars to average the new ones out. Compare each group with the ones before it, not with the studio-wide rate.
  • On a location tab, the group follows where the membership was bought.

The fade

The weeks in which a member’s visits thin against their own pattern, before any cancellation is filed.

Formula: fade check = days since the last visit ÷ that member’s usual gap between visits; a result of 2 means the gap has doubled

In Xyzios: a member in the fade is at risk. Retention scores every member each night on signals such as days since the last visit, attendance down by half against their own norm, a usual class missed three times running, no-shows or late cancels climbing, and no future bookings. The list starts at a score of 3, and High is 6 or more. The quick check above is your own. See How someone becomes at risk.

  • Compare each member with themselves, never with the studio average: once a week is normal for one person and a warning for another.
  • The decision to leave shows on the calendar weeks before it arrives as a cancellation, which is why the fade can be won back and a cancellation rarely can. A note sent in that window, while the person is still a member, is a retention save.

First-week activation

The share of new members who took at least two classes in their first seven days.

Formula: first-week activation % = new members with 2 or more attended classes in their first 7 days ÷ new members in the group × 100

In Xyzios: Data shows it under Member engagement, and Compare shows Activation by location. The group is renewing members who started 7 to 90 days ago, so everyone in it has had a whole first week.

  • It moves in the month someone joins, long before churn can.
  • Two classes is a starting line, not a law. Test it on your own joiners: compare how many of the activated group are still members at day 90 with how many of the rest are. That test is your own calculation.

Money

MRR

Monthly recurring revenue: what your renewing memberships bill in a month.

Formula: MRR = the sum of every running renewing membership’s rate, turned into a monthly figure (a weekly rate × 30 ÷ 7, a yearly rate ÷ 12)

In Xyzios: Finance shows Expected 30-day MRR from active memberships, and so do the foot of the Acquisition funnel on Sales and Compare. KPIs calls it Monthly recurring, Recap shows Recurring at the period’s end, and Membership pricing on Data breaks it down plan by plan.

  • Paid in full is never MRR. The member is a member every month of the term, but booking the cash as one month’s recurring revenue makes a spike that every later month reads as a decline.
  • Class packs, drop-ins, intro offers, gift cards, retail and comps are out. A class pack that renews itself is classified renewing, and counts.
  • A membership bought and not yet started, or paused, adds nothing until it runs.

ARPU

Average revenue per member: what a member pays in recurring dues a month, on average.

Formula: ARPU = MRR ÷ the members paying it

In Xyzios: Membership ARPU on Data, and ARPU under Acquisition cost and member value on Marketing. The average for All memberships in Data’s Membership pricing is the same figure.

  • It divides by the members who pay MRR, not by every member, so paid-in-full holders, who pay no monthly dues, do not pull it down.
  • Revenue / member on Compare is a different reading: all revenue over the window you pick ÷ members now.
  • Watch dues per member, which your prices move, apart from everything else per member (privates, workshops, retail), which your programming moves. That split is your own, and the comparison that tells you something is your own last quarter.

Lifetime value

What a member is worth over their whole stay, after the cost of serving them.

Formula: lifetime value = ARPU × gross margin × average member lifetime, where average member lifetime = 1 ÷ the three-month average monthly churn, in months

In Xyzios: LTV on Data and Marketing, with Avg member lifetime on Data. It uses the Gross margin you set in Settings; until you set one, the LTV note says it is using an industry estimate. LTV per member on Recap is a different reading: cash per member per month × months retained, with no margin.

  • For example, at $150 a month and 5% monthly churn the average stay is 20 months, so a member brings $3,000 in dues; at a 50% gross margin they are worth $1,500.
  • Use the revenue figure to see what one cancellation costs, and the margin figure for a spending decision, such as what you can afford to pay to win a member.
  • The stay is an approximation, since new members leave faster than long-standing ones, but it moves when your churn does.
  • To find your gross margin, take a recent quarter: 100 × (revenue − instructor payroll − processing fees) ÷ revenue, before rent, marketing and admin.

CAC

Cost to acquire a member: what it costs to win one new member.

Formula: CAC = ad spend over the last 30 days ÷ new members in the last full month

In Xyzios: CAC on Marketing and Data, with LTV : CAC beside it, and on Recap. Spend is Google plus Meta. Without Meta’s spend, CAC counts Google only and LTV : CAC stays blank rather than reading high.

  • Read it as a monthly rate: a month of spend against a month of new members.
  • It counts people who joined in your booking platform, not leads and not the purchases an ad platform reports.
  • It is ad spend only. An intro offer that loses money on every class, and staff hours spent following up, are acquisition costs too; add them yourself for the whole figure.
  • For example, $3,600 of ads and 15 new members is $240 a member.

Payback

The months of a new member’s dues it takes to earn back what winning them cost.

Formula: payback in months = CAC ÷ ARPU

In Xyzios: no board shows it. This is your own calculation, from CAC and ARPU on Marketing.

  • For example, $240 ÷ $160 a month is 1.5 months: the member has paid back their cost halfway through their second month.
  • A studio runs on cash, which makes payback in months more useful than any ratio of lifetime value to cost. A member who takes ten months to pay back is a loan to your own marketing, however handsome their lifetime figure.

Switch point

The number of classes a month at which a class-pack holder would pay less on a membership.

Formula: switch point = membership price a month ÷ the pack’s price per class

In Xyzios: no board shows it. This is your own calculation, from your price list; Membership pricing on Data shows what members actually pay, plan by plan.

  • For example, a $160 membership and a pack at $20 a class cross at eight classes a month: a pack holder who comes more often than that pays more than a member would.
  • Treat it as a design decision: choose the pace at which a regular should be better off joining, and price the pack and the membership to cross there.

Partner visit

A class taken through an aggregator app such as ClassPass or Wellhub, which bills the customer and pays the studio separately.

Formula: a partner’s rate per visit = a payout ÷ the visits it covers, read off that partner’s own statement

In Xyzios: the visit counts as attendance and fills a seat, while the money arrives as partner payouts: Finance charts them under Partners, and Revenue · MTD on the Overview adds them on All locations, on a cash basis. Operations prices each partner visit at your partner rate in What a class earns, and says so while that rate is still an estimate. Tell Xyzios support what ClassPass and Wellhub actually pay you per visit (Report issue, in your account menu) and it becomes your number.

  • A partner with no known rate shows as Visits not priced: the visit counts, and the money is left out rather than invented.
  • A partner visit never makes a member, and partner profiles are not leads unless your studio’s rules count them.
  • Count the visit in the month of the class and the cash in the month it lands, reconcile the two each quarter, and never apply one partner’s rate to another’s visits.

Leads and intros

Lead

A new person in your booking platform: every new profile, or only one that didn’t buy when its account was set up, as your studio chooses in Studio Setup.

Formula: new leads in a month = profiles created in the month (all of them, or those that bought nothing at sign-up), less partner and gift-card profiles and bulk imports

In Xyzios: Leads · MTD on Sales, New leads today on the Sales Desk, and New leads · MTD and New leads by month on Marketing. On a location tab, leads follow the person’s home location.

  • The monthly chart and the conversion cohorts also leave out walk-ins: people who bought something within a day of their profile being created.
  • Your booking platform’s own lead report counts inquiries, so the two differ.
  • Never blend the sources: paid, organic and walk-in leads behave differently. Marketing keeps each channel’s leads and cost per lead apart.
  • A lead’s clock starts when the profile is created, not at the first reply. The conversion rates below count from that day.

Intro offer

A first purchase for someone new: a single class, a class bundle or a short unlimited window, priced below membership and usually time-limited.

Formula: intro buyers in a month = people whose first intro was bought that month, each counted once

In Xyzios: a product you mark Intro offer or First class trial on Classification is an intro. Active intro offers · now on Sales counts the ones open right now, with how many sold this month and how many expire within five days, and Conversion cohorts follows each month’s intro buyers forward.

  • An intro bought and never used, more than a week past its own window, is not open: it leaves the open and expiring counts and waits on Call Lists under Intro bought, not started yet. On Mariana Tek, which dates an intro only when it starts, that window is taken as 37 days from purchase.
  • Someone who buys two intros is one person in a cohort, judged from the first.

First visit → second visit

The share of first-timers who came back for a second visit within 30 days.

Formula: first visit → second visit % = people whose first class fell in the month and who attended again on a later day within 30 days ÷ people whose first class fell in the month × 100

In Xyzios: Data and Sales show it by the month of the first class. Compare calls it First visit → returned and reads the last closed month for each location, recomputing the combined figure from summed counts rather than averaging the locations. KPIs shows the other side as % NVAFV: first-timers who never came back.

  • Read it on a closed month, once every first-timer has had the whole 30 days. A month still open counts people who still have time to come back as lost; on Data, solid months are closed and dashed ones are still maturing.
  • Two classes on the first day are one visit, not a return. A second visit at any of your locations counts.
  • It moves weeks before intro → member can.

Lead → intro

The share of a month’s leads who bought an intro within 30 days of their profile being created.

Formula: lead → intro % = leads created in the month who bought an intro within 30 days ÷ leads created in the month × 100

In Xyzios: the first card on the Sales Desk reads the newest monthly cohort that has had a full 30 days, company-wide, against its target, which is set under Goals in Settings. Marketing shows it beside Lead → member, and Compare by location. See From lead to member.

  • A younger month would read low only because it is still maturing, which is why the card shows an older one.

Intro → member

The share of intro buyers who became members.

Formula: intro → member % = intro buyers who bought a membership within 30 days of their intro ÷ intro buyers in the group × 100

In Xyzios: the card on the Sales Desk reads intros from 30 to 90 days ago, so everyone in the group has had at least 30 days to decide, and a location’s tab on Sales reads its own intros the same way. Conversion cohorts on Sales follows the rate month by month, by the month the intro was bought, and Compare reads the last closed cohort for each location.

  • A membership here means a renewing membership or a paid-in-full pass.
  • By default the card dates an intro from its first class after the purchase, so an intro not yet used waits outside the group. A studio whose intros run from the day they are sold can have them dated from the purchase instead.
  • Four decisions change the rate: whether you divide by intros sold or intros used, the window, what counts as converting, and whether the group is by intro or by join month. Compare it with your own last quarter, never with a figure measured under someone else’s four decisions.

Lead → member

The share of a month’s leads who became members within 30 days of their profile being created.

Formula: lead → member % = leads created in the month who bought a membership within 30 days ÷ leads created in the month × 100

In Xyzios: Lead → member on Marketing reads the last closed lead cohort. The Acquisition funnel on Sales and Conversion by location on Compare show it too.

  • It is a different number from intro → member, with a different denominator: a lead who never booked is in this one and not in that one. So it reads lower, and the two never share a goal.
  • The 30 days is your studio’s conversion window, one of the rules set with Xyzios. A shorter window gives fresher numbers and lower percentages, and a cohort already closed does not reopen.
  • A membership here means a renewing membership or a paid-in-full pass.

Classes and the schedule

Fill

Booked seats as a share of the seats a class offered.

Formula: fill % = seats booked ÷ seats offered × 100, over the classes that ran

In Xyzios: Programming shows Avg class fill, Peak fill and Off-peak fill, graded against the class-fill target you set in Settings. KPIs calls the same figure Utilization, and Compare calls it Fill rate. Peak means weekdays before 9am or from 4pm, and weekends before noon.

  • Cancelled classes are left out, and so are sessions that are not one of your class formats, such as a sauna slot or a workshop (Class formats on Classification).
  • Seats offered is the capacity your booking platform records. If a room’s capacity is set high there, fill reads low.
  • A combined figure is recomputed from the totals, never averaged across locations, so a 30-seat room does not weigh the same as a 90-seat one.
  • A studio-wide average hides the schedule: a packed evening can carry a mid-morning that has never worked. Read it by slot and format: Top classes by fill, Bottom classes by fill, Fill by format and the Schedule heatmap.
  • A booking is an intention, so a sold-out class can still run short. Attended fill, next, counts who came.

Attended fill

Attendance as a share of the seats a class offered: the second reading of fill.

Formula: attended fill % = people who attended ÷ seats offered × 100

In Xyzios: no board prints it, but you can work it out from two that do: fill × the show rate (attended ÷ booked, Show rate by weekday on Operations) over the same classes, or a slot’s attended average ÷ its seats. On Programming, each slot in Top classes by fill and Bottom classes by fill shows its attended average beside its booked fill, with the See location costs grant.

  • Where fill runs high and attended fill runs short, the gap is no-shows and late cancels nobody took, not demand.

Attended average

The average number of people who actually came to a class, over a set window, for one slot or the whole schedule.

Formula: attended average = attended visits ÷ classes run, over the same weeks

In Xyzios: with the See location costs grant, each row of Top classes by fill and Bottom classes by fill on Programming carries a bar for what the slot averages in attended heads, with a tick where break-even falls.

  • It is the headcount to price and plan by, not the size of the room. Cost per class ÷ attended average is the least a visit can earn before the class loses money.
  • Judge a slot over a long window, such as the last quarter, not a week.

No-show rate

The share of expected attendance lost to people who booked, never cancelled and never came.

Formula: no-show % = no-shows ÷ (attended + late cancels + no-shows) × 100

In Xyzios: No-show rate on Operations, for the window you pick, with its trend against the prior 30 days on Last 30d. Show rate by weekday (attended ÷ booked) is on the same board.

  • No-shows, late cancels and attended share one base, so the three always reconcile. Your booking platform’s own booked count is taken at a different moment and need not match.
  • Check the check-ins first: a roster nobody checked turns people who came into no-shows. Missed check-ins on Operations lists those classes.
  • Work it out slot by slot and read it beside the waitlist, because a high rate can cost nothing: a no-show in a class nobody was waiting for loses no seat anyone wanted. The slot-by-slot reading is your own.

Late cancels

Bookings cancelled inside your booking platform’s cancellation window, too close to class for the seat to be sure of a taker.

Formula: late-cancel % = late cancels ÷ (attended + late cancels + no-shows) × 100

In Xyzios: Late cancels on Operations gives the count, with its share of the same base the no-show rate uses beneath it.

  • A late cancel can still be rescued: if someone on the waitlist takes the seat in time, it costs nothing.
  • Set the window from your own releases: note how many hours before class each seat came back and whether the waitlist took it, and close the window where released seats stop being taken. That is your own calculation.

Seats someone wanted

The seats a class lost to no-shows and untaken late cancels while people were still on its waitlist.

Formula: seats someone wanted = no-shows + late cancels nobody took, up to the number still waiting when class started

In Xyzios: no board counts it. This is your own calculation, from each class’s roster and waitlist in your booking platform.

  • The two rates say how often; this says what it cost. For example, a 6pm class with three no-shows and four people still waiting lost three seats, and a noon class with three no-shows and nobody waiting lost none.
  • Judge a late-cancel policy by it, slot by slot, before and after a change: fewer seats lost that someone wanted, with the classes still booking full.

Waitlist

The queue of people who tried to book a class that was already full.

Formula: waitlist depth = people on the waitlist for one slot, averaged over four weeks of that slot

In Xyzios: no board shows waitlist depth; read it from your booking platform. Your AI team watches it too: when one recurring slot has had five or more people waitlisted over the last 14 days, the Analyst can raise it in Your focus on the Overview, tagged This week.

  • It is the plainest demand data a studio owns: someone tried to book a particular class at a particular price and was told no.
  • Read it by slot over four weeks, so one busy night does not become a schedule change. Where it stays deep, that is the case for another session, and the test afterward is that attendance across both sessions rises.

Cost per class

What one class costs to run: payroll plus rent, divided by the classes that ran.

Formula: cost per class = (payroll + rent over the last three closed months) ÷ classes run in those months

In Xyzios: Operations shows it as est. class cost beside Est. break-even, under What a class earns, for owners and anyone with the See location costs grant. It comes from your books on Finance; with fewer than two closed months it reads Not enough data.

  • Payroll is instructor pay where a payroll register is connected, otherwise all payroll at the site. Rent is rent alone, no utilities, and owner pay is never in it.
  • Where your books do not split payroll and rent by site, one company-wide figure is used, and the board says so.
  • Months the bookkeeper has not closed are left out. Count the classes that ran, not the ones scheduled.
  • Working it out yourself, take the costs from the books rather than the rate card, so payroll taxes and sub cover are in. For example, $72,000 of payroll plus $18,000 of rent over 900 classes is $100 a class.
  • Month by month, a location’s tab on Finance shows revenue and cost per class and per attendee under Class economics, with the same grant.

Revenue per visit

What an attended visit earns, on average.

Formula: revenue per visit = revenue in the period ÷ attended visits in the period

In Xyzios: rather than divide total revenue, Operations prices each attended visit from the attendee’s membership, and a partner visit at your partner rate, then shows Average per class and Per available seat under What a class earns. That value per visit is what the break-even headcount divides by. Visits it cannot price are counted as Visits not priced and left out.

  • Divide by visits that happened, not bookings: a no-show earns nothing.
  • It averages very different visits: an intro, a drop-in and a heavy unlimited member are not worth the same. A partner visit reaches your booking platform with no revenue attached, so an average taken from the platform alone reads low.
  • For example, $180,000 over 9,000 attended visits is $20 a visit.

Break-even headcount

The number of people a class needs, at what an average visit earns, to cover what the class costs to run.

Formula: break-even headcount = cost per class ÷ revenue per visit, rounded up

In Xyzios: Est. break-even on Operations, and on Programming a tick on every row of Top classes by fill and Bottom classes by fill, set against that slot’s attended average: green clears it, red falls short. Both need the See location costs grant. There is one figure per location, never one per slot, and with fewer than two closed months it reads Not enough data, never a number.

  • It counts attended heads, where fill counts booked seats.
  • What a visit is worth moves with the window you pick; the class cost does not.
  • A slot with many visits that could not be priced is not judged.
  • A Location Manager sees their own sites’ figure, never the company-wide one.
  • For example, a $100 class at $20 a visit needs five people. For a slot near the line, redo the sum with its own instructor’s pay and its share of rent: that is your own calculation. Before you cut a class, count the regulars whose only class it is.

The ledger

Every entry on Recovered Revenue starts when an agent proposes. Its clock starts when the email goes out, or when the desk marks the member Handled on Retention; an approval alone does not start it. A draft nobody decides within 14 days expires, and a rejected draft never earns, even if the outcome happens anyway. See Proposes, approves, proves.

Retention save

A note to a current member whose visits are slipping against their own pattern, sent while they are still a member.

Formula: one month of dues credited at each check the member passes, on day 60, 90 and 120 after the clock starts: three months at most

In Xyzios: your AI team drafts a save for a High member on Retention who slipped within the last 30 days, whose payment is failing, or who is new and has never come in; everyone else stays on the list for a personal call. It waits in Approvals, and nothing reaches the member until a person approves it. Retention saves is switched on or off in Settings → Features (Turn on your AI team’s drafts).

  • A note to someone who has already cancelled is not a save, and Xyzios does not draft one.
  • A member who has left by day 60 earns nothing, and the entry reads Expired. One who leaves later keeps what was already credited and earns no more.
  • What a save may be worth is every month the member goes on to stay; what the ledger counts is only the months it can confirm. For example, a save on $150 dues earns $150 at each check it passes, $450 at most.
  • A Handled tap counts the way a sent email does, and no new save is proposed for that member for seven days.

Payment recovery

A failed payment collected after a note asks the member to fix it.

Formula: one check, on day 21 after the clock starts: credited what failed (a month of dues for a failed membership payment, or the fee itself for a failed fee) if it has been collected, and $0 if not

In Xyzios: your AI team drafts a kind note for each member whose payment is failing, when the failure is newer than the Payment-recovery window in Settings; older failures were chased before Xyzios saw them. No new note is drafted for a member someone spoke to in the last seven days: one marked Handled on Retention, one whose failing-card flag was marked done on Front Desk, or one reached by a call logged on the Sales Desk’s list of negative account balances. It waits in Approvals, not on the Sales Desk. If you sort the payment out yourself, mark it Fixed there and its clock starts then. Payment recovery is switched on or off in Settings → Features.

  • Judge it at day 21, not day seven: a card updated on day ten still counts.
  • As a rate, dues collected by day 21 ÷ dues that failed. That rate is your own calculation.
  • Keep these losses apart from members who chose to leave: they are involuntary churn.
  • Your booking platform’s own retries carry on. Xyzios sees that a payment failed, not why, so the note assumes an innocent cause such as an expired card. It never mentions an amount, never threatens access and never retries a card.

Intro rescue

A follow-up to someone whose intro offer ends within five days, recorded on the ledger at $0.

Formula: one check, on day 14 after the clock starts: if the person now holds a membership bought after the note went out, the entry reads Confirmed at $0; if not, Expired

In Xyzios: a confirmed entry shows the plan the person bought and adds it to nothing.

  • It is a receipt, not revenue: a note before a purchase does not prove the note caused it.
  • The expiring intros are on Call Lists, in the Intro expiring call sheet, for your team to call.

Recovered Revenue

The running total of what your AI team’s sent or handled actions verifiably earned, each dollar traced to the action behind it and counted only after a dated check against your booking platform.

Formula: confirmed recovered revenue = payment recoveries collected by day 21 + retention saves whose member is still a member at day 60, 90 and 120 (a month of dues each, three at most) + intro rescues at $0

In Xyzios: Recovered Revenue, which the rail calls Recovered: Confirmed · all time, Awaiting outcome, Expired and Rejected, then the entries themselves. Recovered · 30 days on Approvals is the money confirmed in the last 30 days.

  • Estimates and lifetime value are never credited. Every rule under-claims, so the total holds up when someone checks a line.
  • Entries that earned nothing stay on the list at $0 beside the ones that did; showing them is what makes the confirmed total believable.
  • Each entry reads Proposed, Clock running, Confirmed, Expired · $0 or Rejected · never earns. The checks run each night against what your booking platform records.