Customer Retention

This month's signal

23 of your regulars are about to walk away. Their lifetime value adds up to £14,820. A 60-second text to each one typically saves 6 in 10.

Model trained on 18 months of bookings, no-shows, message tone and email engagement. It learns each client's personal rhythm, not a one-size-fits-all rule. The list updates every morning at 6am.

CLIENTS

Active Customers

847

Visited in last 12 months

PREDICT

At Risk This Month

+4 vs last month

Revenue At Risk

£14,820

12-month lifetime value combined

SAVED YTD

Recovered Revenue

£38,440

82 saves · 64% success rate

Risk Distribution

Active client base · today's score

847

Total clients

Critical

High

Medium

112

Low

664

Saves YTD · Interventions That Worked

Clients flagged, contacted & rebooked within 14 days

Saves

128

Flagged

64%

Success rate

£469

Avg saved (LTV)

What Predicts Churn Best

Signal strength across 18 months of historical data

Visit gap vs personal cycle

94%

Email/SMS engagement drop

71%

No-show or late-cancel pattern

63%

Sentiment in last 3 messages

52%

Stylist switch or unavailability

41%

Spend per visit declining

28%

What's Driving Risk Today

Patterns the model spotted this week

Pattern noticed
Customers who missed your April 22nd "new colour menu" email are 3.1× more likely to be in the at-risk group right now. A second send to non-openers might be worth running.

Biggest single save opportunity

Sarah Mitchell, 11 yrs as a customer, £2,140 lifetime value. Risk score 87. Hasn't been in 11 weeks (usually 6). Worth a personal call from Mel.

Stylist pattern

Jodie's chair shows the highest retention (94%). Liam's chair has 19% churn, usually clients who tried him once after their regular stylist left.

Timing tip

Saves work best when contact is made between weeks 8 and 11 of silence. Past 14 weeks the success rate drops from 64% to 22%.

Customer

Lifetime value

Last visit

Risk score

Why they're at risk

Action

How the model actually works, no jargon

Built around your business, not someone else's

Step 01

Learn each customer's rhythm

The model watches every customer for a few months and works out their personal booking cycle. Sarah comes every 6 weeks. Tom every 11. Megan whenever she fancies. Once it knows the rhythm, it knows what "late" looks like, for them.

Step 02

Watch the small signals

Then it tracks the quiet things humans miss. Did they stop opening your texts? Did their messages get shorter? Did they cancel and not rebook? None of these matter on their own. But two or three together is your warning.

Step 03

Tell you why, not just who

Every alert comes with reasons in plain English. "Eight weeks since last visit, usually six. Hasn't opened the last 3 emails." So when you pick up the phone, you know what to say. No guesswork.