06 — Case study
What the model found at Alkami Barbershop.
Alkami gave us full access. Here’s exactly what Rydra’s analysis identified.
The situation
Six barbers. A loyal client base. And no visibility at all.
Alkami Barbershop in Canberra runs six barbers. A well-run, well-regarded business with a consistent client base. And no visibility whatsoever into which of those clients were drifting.
The business was paying a scheduling platform tens of thousands of dollars a year for the ability to book and confirm appointments. We found that they have actually lost over $200,000 to marketplace fees — and in return, all they got was a calendar. No data on which clients were loyal versus at risk. No churn signal. No predictive intelligence of any kind.
Alkami knew clients were leaving. They could feel it. They couldn’t measure it, and they couldn’t act on it systematically.
What the Rydra model identified
Every client classified into a Markov state — First Timer, Loyal, On the Fence, At Risk, Churned — in real time.
A proportion of clients sitting in the At Risk state at any given moment, consistent with the 20–35% of active databases seen across comparable personal care businesses.
A precise intervention window of 45–90 days for the majority of at-risk clients — the period where a single personalised message can recover them.
$50,000+ in identifiable annual margin across platform fee elimination, no-show reduction, and systematic churn recovery.
A projected rebooking rate improvement from below the 22% industry standard toward 34%+ with the intervention model running consistently.
The arithmetic
average appointment value
cost of a single recovery SMS
barbers — each with their own dashboard: rebooking rate, at-risk clients, intervention opportunities
the intervention window before a client churns permanently
total annual margin identified in the analysis
messages written by a human. The model writes them all.