Convenience store marketing & events planner - promo margin break-even calculator tool - five promo mechanics side by side on break-even volume multiple, every formula printed on the page instead of a black box, answering only how many times everyday volume a SKU must sell to break even

You are a coding agent. Build a single-file HTML "convenience
store promo margin break-even calculator" for chain marketing
planners: work out, for one promo SKU across five mechanics,
how much margin each gives up and how many times everyday
volume it must sell to break even. Runs locally, no network,
no external dependencies.

Position (design to this; not an all-purpose promo evaluator):
Planning meetings argue one line forever - "how much does
50% off the second unit actually give up, and how many extra
units until we stop losing" - and most answers are gut calls.
This tool answers exactly one question: the give-up ledger
and the break-even ledger of a single SKU. Traffic pull,
cannibalization of neighboring SKUs, brand exposure are all
out of scope - and that boundary is printed on the page, so
the tool is never mistaken for a full promo evaluation; those
ledgers belong to an attribution analyzer.

Inputs:
- Cost, everyday price, everyday weekly volume
- Promo mechanic, one of five: markdown of X off list /
  X percent off / 50% off the second unit / buy-one-get-one /
  add-on exchange at X (BOGO rolls the free unit's cost into
  unit cost; the exchange rolls the price difference into
  unit cost)
- Gift cost (if applicable)
- Expected participating SKU count: the multiplier that
  scales the per-unit ledger into a period total (default 1,
  adjustable per promo)

Four outputs:
1. Per-unit give-up and give-up rate: give-up rate =
   per-unit give-up / everyday per-unit margin - a ratio
   against margin, not against price
2. Break-even volume multiple = everyday weekly margin /
   promo per-unit margin. The formula is printed on the page
   verbatim and recalculates live as inputs change; when
   promo per-unit margin <= 0, print "the more you sell, the
   more you lose" and give no multiple
3. Five-mechanic side-by-side table: anchored at equal
   give-up rate, listing each mechanic's break-even multiple
   and per-unit cash intake - so the planner sees structural
   gaps like "BOGO approximates 50% off, but one hands over
   goods and the other hands back money"
4. Sensitivity table: promo-period margin minus everyday
   margin at 1.2x / 1.5x / 2x / 3x volume - positive or
   negative readable at a glance

Hard constraints:
- Every formula above is printed on the interface, including
  intermediate quantities (everyday per-unit margin, promo
  per-unit margin) - no black box
- Data stored in browser localStorage, exportable /
  importable as JSON, survives a machine change
- One demo SKU preloaded (a bottled drink: cost 2.4, price
  4, weekly volume 150, 50% off the second unit) so all four
  outputs are visible on open
- Comparison and sensitivity tables print cleanly in black
  and white, for use in the meeting itself
- No network, no external libraries

You will not:
- No volume forecasting - "expected sales" never appears in
  the tool; it answers only "how much must it sell to break
  even," and the volume judgment stays with the human
- No multi-SKU or category-period roll-ups - the single-SKU
  boundary is the tool's boundary
- No external data of any kind (platform prices, competitor
  prices, sales-history APIs)
- No accounts, no login

Why it is written this way

Printing "one SKU's give-up and break-even only - no traffic pull, no cannibalization" on the page is this tool's most important gate. A break-even calculator invites a classic misuse: a planner takes "BOGO break-even multiple 2.1" into a meeting and kills a promo that was designed for traffic in the first place, and the argument runs on the wrong ledger. Traffic and cannibalization are real ledgers, but they are not inside single-SKU arithmetic; the on-page boundary reminds the user every time they read a conclusion, so the number is never mistaken for the whole evaluation. Those ledgers belong to an attribution analyzer - each tool owns one stretch and both survive.

Formulas print on the page with their intermediate quantities because numbers in a planning meeting have to survive store managers and finance asking "how did you get that." A black box spits out "break-even multiple 2.1" and the first question kills it; a page showing the formula with everyday per-unit margin and promo per-unit margin visible, recalculating live as inputs change, answers the challenge on the spot. When promo per-unit margin goes to zero or below, the page prints "the more you sell, the more you lose" instead of a multiple - at that point the division has no meaning, and a negative multiple would only be misread.

The five mechanics compare side by side at equal give-up so structural gaps surface on their own. "12 percent off" versus "buy-one-get-one" is usually settled by gut: BOGO's give-up approximates 50 percent off, but it hands over a unit with real cost while a percentage discount hands back cash - completely different cash-intake lines; 50% off the second unit only reaches its headline depth when the customer buys two, and single-unit buyers give up nothing. Written as adjacent columns, the differences sit in front of the planner on every open, instead of buried in a deck nobody re-reads.

The sensitivity table replaces a volume forecast because nobody can predict how many times a promo will sell, and the moment a forecast number appears in a tool, the meeting orbits that invented number. Lay out the margin gap at 1.2x through 3x instead, and "1.5x barely breaks even, 2x starts earning" becomes a lookup instead of a debate; the judgment stays with the people in the room. Printable, because projection screens lose detail - a sheet of paper on the table lets anyone point at a row and argue with it.