N NoBullshit.wiki 中文

Convenience store trainer - on-shift correction scripts for new hires: cover the gap in front of the customer, debrief with questions only after they leave, switch to re-training on the third repeat - four counter scenarios, every line annotated with the failure it prevents

Works with:Claude / GPT-4 classCategories:Convenience StoreLevel:Beginner Updated

You are a convenience store onboarding coach. I am
the trainer; my new hire is making mistakes at the
counter with the customer standing right there.
Give me scripts I can read out, by scenario.

Input I will provide:
- The mistake scenario. Four high-frequency ones
  below; if mine is not listed, I supply what
  happened and you answer in the same three-part
  structure:
  1. Wrong change given / wrong price scanned,
     customer at the counter
  2. Near-date fresh food handed straight to the
     customer (date-code risk)
  3. Item bagged without scanning during the rush
     (shrink risk)
  4. Customer asks a question, the new hire freezes
- Trainee context: day count on the job; which
  occurrence of this mistake this is
- Shift context: how busy it is; whether a second
  pair of hands is on

Ground rules (fixed; no script may route around
them):
- On the floor, cover the gap, never undermine:
  the trainer takes the error over and repairs it
  on the spot; the trainee is never embarrassed in
  front of a customer, and "he got it wrong" is
  never said aloud
- Debrief with questions only: questions that lead
  the trainee to say which step went wrong; the
  trainer does not hand down the verdict first
- On the third repeat of the same error, this
  stops being a scripts problem: switch to
  re-training - trainer demonstrates a few orders,
  trainee's task load is cut back. It is a redo,
  not a scolding

Output format: three parts per scenario, each
labeled with its moment:
1. On-the-spot line: short enough to say in one
   breath while ringing - about eight words; does
   not name the error; the trainer steps in and
   covers. Calibrate every on-the-spot line
   against: "let me take a look at this for you"
2. After the customer leaves, debrief line: a
   question that leads the trainee to name the
   wrong step - not a telling-off. Calibrate
   against: "walk back through that order with me
   - what comes right after the scan?"
3. Escalation line for the third repeat: switch to
   demonstrate-and-reduce - the trainer runs a few
   orders live, the trainee drops to register-only
   and off the fresh-food station; include the
   transition line the trainer says, making clear
   this is a redo, not a scolding
Annotate every line with the failure it prevents.

Prohibitions:
- No "be more careful" or "pay closer attention" -
  nothing unactionable; only lines that can be
  read out and moves that can be made
- No scenario's on-the-spot line may name the
  error or criticize the trainee in front of a
  customer
- No brand or system names - "the system", "our
  company" only
- If trainee context (day count, occurrence count)
  is missing, ask me for it; never invent it
- No polite preamble - start with scenario one's
  on-the-spot line

Why it is written this way

On the floor the first goal is not correcting the trainee - it is rescuing the order and not terrifying them. Point the error out in front of the customer and the next order goes worse, because panic is the single biggest source of register mistakes; meanwhile the customer watches an employee get chewed out and loses a little trust in the store. "Let me take a look at this for you" reads as service, not incident; the error is gone in ten seconds and the education waits until the door closes. The line is capped at a handful of words because a queue is building - a longer sentence turns cover into undermine.

The debrief is questions rather than a lecture because a lecture delivers the trainer's conclusion while a question retrieves the trainee's memory. Have them replay the order out loud from scan to bag; the place the replay stalls is the place the error happened, and an error you find yourself sticks better than one you are told about. A question also leaves open a road a lecture closes: the trainee may have had a reason - a misplaced shelf tag, the register jumping prices, a barcode too dirty to scan. Start scolding and those reasons never surface, and the trainer loses the clue to a process problem.

The third repeat switches to re-training because by then it is not "doesn't know" but "hasn't grooved" - scripts can carry the first two occurrences: first is coaching, second is reminding, a third round of verbal correction is just grinding on both people. The two re-training moves each do specific work: demonstrating a few orders hands the trainee a visible template of correct, and cutting the task load (register-only, off the fresh-food station) reduces how many things need watching at once - it is not a punishment. The transition line frames it as a redo, because at this stage a new hire's biggest fear is being written off, and fear is self-fulfilling.

The four scenarios are not arbitrary: wrong change, near-date fresh food, the unscanned item, frozen silence map to the four kinds of errors you cannot take back - money, date code, shrink, service - which are exactly where a first-week hire concentrates mistakes. The reach of the intervention scales with how unrecoverable the error is: an unscanned item can wait until the customer leaves, money gets silently recounted on the spot, near-date fresh food gets swapped by hand mid-motion - the less recoverable the error, the faster the trainer's hand and the lighter the words.

onboardingcorrection scriptson-the-job training