FoodOwner of a 20–60 seat eatery, cooking or managing in person

When my staff leave after a few months, I want to keep people long enough to become good, so I stop retraining from scratch four times a year.

Khi nhân viên của tôi nghỉ sau vài tháng, tôi muốn giữ được người đủ lâu để họ thạo việc, để tôi thôi phải đào tạo lại từ đầu bốn lần một năm.

Job context

Who: An owner with 3–8 staff, mostly part-time
When: Each time someone quits, usually with a few days' notice
Trigger: The third resignation in a quarter
Situation: A new hire needs six weeks to become useful and stays five months on average
Constraints: Margins are thin, so an across-the-board raise is impossible

Does not know why people leave
High
The reason given at the exit is almost never the real one.
Root cause: Honesty on the way out benefits the leaver in no way
Six weeks of training lost per replacement
High
The true cost of churn dwarfs whatever the lower wage saved.
Root cause: Training cost appears in no ledger, so nobody sees it
The roster is always short on the busy days
Medium
Everyone wants weekends off, and weekends are when the shop lives.
Root cause: The shop's demand and staff preference are systematically out of phase
Nothing to promise the people who stay
Medium
A single shop has no career ladder to point at.
Root cause: A small shop has two rungs: staff and owner
Raising one wage is known by everyone
Medium
Kitchens have no wage privacy, so every individual adjustment becomes precedent.
Root cause: Shared workspace means wage information travels instantly

A solution is not the same thing as a product — a customer can hire a behaviour or a workaround too.

Match the local wage
Behaviour
Effective and expensive, and holds only until the shop next door pays more.
Rostering app
Product
Solves scheduling, and never touches why people want to leave.
Shift-revenue bonus
Workaround
Owners invent a career ladder out of money because no other ladder exists.
Hire friends and family
Behaviour
Retains through relationships, and makes firing nearly impossible.
A group chat for swapping shifts
Workaround
The trade's actual rostering system, running on everyone's phone.

Churn costs far more than the wage saved, yet it appears in no ledger.
Why existing solutions fail: Wages are a visible monthly line; six weeks of retraining is the owner's own time, and that time has never been priced.
Potential opportunity: Putting a number on churn so it can be compared with the cost of retention.
The only person who knows why they are leaving has no reason to say.
Why existing solutions fail: Leavers still need a reference and still meet again in a small trade, so honesty carries only risk.
Potential opportunity: Asking while people are still there and can stay anonymous, rather than at the exit.