01 · Latest & Featured Project
Food Waste & Lost Sales — Store Ordering Optimization (Q3 2026)
A full-quarter operational audit of daily food operations in a high-volume store (1 Jul – 30 Sep 2026): 43 items, 92 days, 3,894 daily stock-movement rows loaded into MySQL. Traced every unit from opening stock to sale, waste (mark out) or stock-out, tested the store's 10%/30% ordering rule against reality, and replayed the quarter under alternative ordering rules to quantify the waste-vs-lost-sales trade-off — ending with a controlled-trial plan instead of a risky rule change.
Verified Outcomes
Quantified 21,436 in waste at cost (3,126 units over 92 days) — sandwiches and pastry alone drive 53%, while salads waste 26.7% of stock handled, nearly equal to the cost of what they sell (94.3%).
Flagged 671 stock-out item-days (17.2%) as possible lost sales, with Friday the worst day (24.9% of item-days) — the mirror image of Monday–Tuesday waste peaks, exposing a weekly ordering rhythm problem.
Proved the stated 10%/30% ordering rule isn't followed — only 22.4% of 2,638 orders match it — and showed via stock replay that a same-weekday 4-week average rule would cut waste ~16% but miss sales on 167 item-days: a trade-off, not a win.
Recommended a 4-week controlled trial on the highest-waste items and a new data-capture standard (order quantities, noon sales, sell-out times, defrost logs) so lost sales can be measured in units, not guesses.
Single Store · Q3 2026 · 92 days21,436 waste at cost