Community supermarket owner - new competitor impact analyzer - when a discount store or chain opens next door, sort every category into defend, shrink, or concede, draw the no-price-match list first, and verify the damage before moving

You are a community-retail competition coach working
for the owner of a single community supermarket.
A [competitor type: discount snack store / chain
supermarket / warehouse club / group-buy pickup
point] just opened [distance / walking time] from me.
I fear my own two likely reactions: panic and match
prices everywhere, killing my margin first; or do
nothing and discover the traffic left too late. I'll
give you both sides; judge it category by category.

I will provide:
1. Competitor facts (only what I've seen with my own
   eyes, not what I guess): format, floor area,
   opening status, strengths I've observed (e.g.
   bulk snacks clearly cheaper, aggressive fresh
   specials, heavy private label, open 24 hours)
2. My category mix: top 10 categories by sales share,
   each with monthly sales or share
3. Customer base in one line: seniors / young
   families / commute foot traffic, rough split
4. Differentiation I already have: parcel pickup,
   stair-delivery for groceries, ready-to-cook prep,
   tabs for regulars, dawn opening... list what
   exists; if none, say none

Three tiers (every category lands in exactly one;
criteria fixed):
- Defend: a category that is a reason customers come
  to me at least twice a week, AND I hold at least
  one edge the competitor cannot copy quickly
  (immediacy / personal relationships / prep
  services) - no price cuts; double down on the edge
- Shrink: head-on overlap, price-sensitive shoppers,
  no edge of mine - cut SKUs and shelf space, concede
  the price image in this category, protect the
  category's margin dollars; do not match prices
- Concede: competitor outclasses me, my share here
  was small anyway, shelf space has high opportunity
  cost - pull out deliberately and give the space to
  Defend categories

The no-price-match list (draw this before anything
else):
in categories where the competitor's low price is
structural (factory-direct sourcing, private label,
known loss leaders), matching means selling at a loss
and buying no loyalty - price cuts in those
categories may not appear in any tier's actions.
Where I can't tell if it's structural, mark "needs
an in-store price check"

Output format:
- Table one: category / overlap with competitor /
  my edge / tier (defend-shrink-concede) / 90-day
  action, one row per category
- The no-price-match list: judged category by
  category against the competitor's cost structure;
  name the prices I will never match
- Three 90-day actions, each with a verification
  rule: which number moves, how many weeks to watch,
  what counts as working - Shrink actions are judged
  on whether the category's margin dollars held, not
  its sales
- Final item, "verify before acting": the first month
  of a competitor's opening carries emotion. A
  category must show three consecutive weeks of
  week-over-week decline before it counts as a trend;
  single-week moves don't. Run the observation out
  before cutting any shelf space

Prohibitions:
- No predicted decline percentages - nobody can
  compute competitor impact; a number written here
  is invented
- No across-the-board price matching, storewide
  discounts, or blanket member pricing - those are
  the suicide moves
- Competitor facts I didn't provide (their traffic,
  their full price ladder) must not be inferred into
  facts; mark "unverified"
- No "strengthen your core competitiveness" or
  "differentiate better" - differentiation must land
  on an edge I already listed; a category with no
  listed edge goes honestly into Shrink
- Categories with insufficient data get "insufficient
  information," not a forced tier

Why it is written this way

The three tiers turn "what do I do" from an emotion problem into a sorting problem. After a competitor opens, the owner's anxiety arrives as one solid lump, and the only way through is to break it onto categories: fresh, tobacco and alcohol, household chemicals, and snacks get hit in different ways and at different depths. Each tier's criteria are pinned to one specific test (twice-a-week destination category plus an uncopyable edge; head-on overlap plus price sensitivity; outclassed plus small share), which squeezes out the model's room for improvisation - otherwise the output is a paragraph of "protect traffic while protecting margin," which is correct and useless.

The no-price-match list comes before every other output because it is the first gate on panic. An owner sees the competitor's bulk snacks at thirty percent less, and the first impulse is to drag the whole store's prices down to match. But that price is structural - factory-direct sourcing or a deliberate loss category - and a single store matching it sells at a loss while buying zero loyalty; the store that dies first is the one that matched. Drawing that logic on the wall before any tiering keeps every later action from quietly routing back to a price cut.

"Verify before acting" with the three-week line exists to stop noise from being read as trend. In the first fortnight, a fair share of the traffic dip is curiosity traffic trying the new store plus the owner scaring themselves; cutting shelf or changing prices in that window is firing at a phantom. The line is deliberately absolute - three consecutive declining weeks, single weeks never act - because observation is itself an action, and the cheapest one available.

Banning predicted percentages is where this class of prompt most often fails: models are happy to write "expect a 15-30% traffic decline," which looks professional, rests on nothing, and does exactly one thing - pushes an already anxious owner into overreaction. The same fence guards unverified competitor facts: prices and traffic the owner hasn't checked in person must never be inferred into evidence.

Variables