Plan · first thirty days
Truth, format, report
What I'd do in the first month, in the order the findings suggest.
Tab 4 · Plan3 phases
No new systems needed
No new systems needed
Thirty days, three phases, nothing that needs a new system — the specials data on lume.com is already structured enough to check.
Days 1–10
Truth
- Run the bundle-vs-single check on all 40 menus; retire or re-price the losing bundles (Baller Jars 2/$98 first)
- Fix the 25%-configured “20% Off” daily deal; rename “Disposable 2g” or fix its SKU list
- Reconcile hours for the 32 stores across Dutchie, CMS and JSON-LD; fix the Big Rapids phone
Days 11–20
Rollout format
- One deal-sheet row per promo (SKU, both prices, stores, dates, limit, funding, disclaimer)
- Store-execution card auto-generated from that row (Tab 3)
- Noindex or 410 the dated campaign pages; remove the Vendor Day card until it has a year
Days 21–30
Report
- Weekly promotion-accuracy report: submitted, cleared first pass, blocked, which check fired
- Store audit checklist tied to the card (price on shelf = price on card)
- Hero-to-landing price match added to the launch checklist
What I'd want to know on day one
Whether checkout already applies the cheapest overlapping deal (that changes finding 01 from a customer problem to a display problem). Whether the CMS hours text reaches the store-selector drawer. Which team owns the specials configuration in Dutchie vs. Contentstack — the fix for a wrong percentage is one field, but only if you know whose field it is. And I haven't worked in cannabis retail; I've run promotions and catalogs at two Michigan wholesalers, which is the same arithmetic with a different rulebook.