Leave
Churn factors
These can hit an account at any stage, so they're not mapped to a specific layer. They're hypotheses until tested against real reason-code data, which is what the Validation pilot opposite is for.
Undercut on price
A competitor, large or niche, wins the account on cost alone.
Local roaster charm
An independent roaster offers a more compelling story and more personal service.
Poor service experience
Order errors, late delivery, or stock-outs quietly erode trust.
Rep discontinuity
The rep the cafe knew leaves the business, the relationship resets to zero.
Ownership change
The cafe changes hands, the new owner brings their own supplier preference.
Neglect
No proactive contact for a long stretch, the account quietly goes cold.
Terms friction
Minimum order volumes, pricing, or payment terms stop working as the cafe grows or shrinks.
Cafe closure
The business itself shuts down. Not a relationship failure, and not something a loyalty programme fixes.
Love: advocates, refers other cafes, resists competitors. The reference account.
Invested: branding adopted, staff trained, real switching cost starts to exist.
Like: regular and satisfied, but replaceable. "Safe, trusted" lands the account here.
Onboarded: signed, first orders placed, the relationship is still thin.
Aware: knows Toby's Estate as an option, sales team's first point of contact.
Test & Learn
Pilots to run
Small, bounded moves that generate real evidence, on both the Learn axis and the Leave hypotheses opposite. Recommended running order below, each pilot builds on what the last one tells us.
1Validation
Exit-reason capture
A lightweight practice for any cafe that does leave, a quick call or short survey to find out why. This is what actually tests the Leave hypotheses opposite.
2Test cohort
At-risk account cell
Pick a small group of accounts showing early warning signs and run a hands-on relationship push against them, see if attention actually shifts the outcome before rolling out wider.
3Pre-rep support
AI pre-visit briefing
Brief reps automatically before a visit using account history, so a junior rep shows up as prepared as a tenured one.
4Churn signals
Ordering pattern analysis
Mine existing order history for leading indicators, volume trend, frequency drift, basket change, before a cafe ever calls in or visibly churns.
5Sales reporting
Voice-to-text pilot
Post-visit voice memo capture with a trusted rep group, transcribed and tagged against the loyalty layers above.
6Cafe tools
8Seats pilot
Sponsor 8Seats' team messaging and management app for a small group of cafes. A genuine value-add for the cafe, and a chance to put Grade's own technology partnership network to work on a real account problem.