Every other studio dashboard tells you a class is emptying out and stops there. Yours drafts the schedule change, sizes who to nudge, and waits for your yes. One number tells you which stage of the member journey is weakest — and that stage hands you something to approve.
18 members booked once and never came back. First-month attendance is where you're weakest — I've drafted a check-in message to the 12 who joined in the last three weeks.
Approve & runRefine
What’s inside Analytics
Three ideas, and the third is the unusual one.
It acts rather than reports. It reads as a journey, so a low number points somewhere. And it refuses to show you a figure it can’t stand behind — which sounds like a small thing until you’ve made a decision on a fabricated zero.
01The signal-to-agent loop
Analytics that hands you something to do.
Five signals watch the things that quietly cost you money — fill sliding, members drifting, cards about to fail, a teacher carrying too much of the schedule. When one fires it doesn’t raise a flag; it drafts a concrete move, grounded in your data, for you to approve or bin. The dashboard is the last resort, not the product.
Fill declineMember driftFailing renewalsKey-person riskSoft-class detection
Legacy studio analytics stop at surfacing the problem. You still have to work out the fix, then go and do it.
app.studiosensei.ai/health/signals
Detected
Tue 6pm fill down 4 weeks
Drafted
Move to 6:30pm, nudge 23
Your call
Approve, refine, dismiss
Signals watching now
Fill decline · Tue 6pm Vinyasa
4 weeks down · schedule_editor drafted
drafted
Member drift · 9 quiet 30+ days
re-engagement agent · awaiting you
drafted
Renewals · 3 cards will fail
dunning note drafted for Thursday
drafted
Key-person risk · Dani teaches 41%
no action drafted — flagged only
watching
Soft classes · none this week
nothing to report is a valid answer
quiet
02The Vitality Score
One number that points somewhere.
A single score is useless if it just goes down. This one is built as the member journey, so a fall names the stage that caused it — and every stage carries the levers that move it. Click a stage in the panel above to see the diagnosis and the drafted fix.
·20
Getting them in
Leads, intro offers, first bookings.
·20
Getting them started
Whether a new member comes back.
·25
Keeping them
Retention, cohorts, quiet members.
·15
Keeping promises
Cancellations, subs, staff reliability.
·20
What it earns
Revenue as the outcome, not the input.
On the ranges. Each signal is compared to a healthy range, and the panel shows you where that range came from — some are published industry figures, some are our own directional estimates until real history calibrates them. We label which is which rather than calling them all “the industry standard”.
03Honest by construction
A blank week is a gap, never a zero.
A new studio has thin data, and most software fills the hole with a confident-looking zero. Ours draws the missing week as a hollow bar, marks a signal “still accruing” and excludes it from the score rather than dragging it down. A row we can’t classify reads “—”. You never make a decision on a number we invented.
Hollow bar for a missing week“Still accruing”, never a fake 0Weight redistributed, not zeroedRange provenance shown
The weakest studio for this is a brand-new one — exactly the studio most likely to be misled.
app.studiosensei.ai/classes
Classes · state and 8-week trend
Every row answers “how is this one doing?” in the list itself.