Customer Case Study · Buying
From blanket "flat 50%" end-of-season sales to an option-level markdown for every style, set early and set right. Built with GenAI and sitting on the buyer's own sales and stock data.
Figures indicative; validated metrics available on request.
Every season, buyers had to clear thousands of options across concepts and categories. The markdown was the single biggest lever on margin, and it was mostly set by gut feel.
One rate per department, "flat 40%" or "flat 50%". Fast sellers were discounted they didn't need to be; slow ones weren't cut enough.
Markdowns waited for the sale calendar. By then the remaining weeks were too few, so the only option left was a deep cut.
Each buyer stitched together sales, stock and cost extracts by hand. A full markdown plan took around three days.
Whatever didn't sell went to outlets or jobbers at a fraction of cost, quietly eating the season's margin.
Buyers stay in charge of every decision. The calculator shows them what each markdown is likely to do before they commit.
Select a step to see what happened.
We sat with buyers and merchandisers across concepts and replayed the previous end-of-season sale.
Deliverable: Markdown leakage diagnostic by concept and category.
Simple enough to trust, sharp enough to beat gut feel.
Deliverable: Markdown logic signed off by buying and finance.
Claude wrote the elasticity analysis and the calculator from buyer-approved specs.
Deliverable: Single-file calculator plus weekly data refresh.
The calculator replaced the spreadsheet on the big screen.
Deliverable: Weekly markdown cadence and approval workflow.
Same engine, more decisions.
Deliverable: Season-on-season learning loop.
Pick an option from the portfolio or the plan. Move the markdown slider to see sell-through, stock and margin change, or apply the recommendation. Change weeks left, salvage value or the clear-out target and every option is re-priced.
Drag the pink marker or click the curve. ★ is the recommendation.
Units left each week at full price, your markdown and the recommendation
Weeks of cover at current rate of sale vs sell-through to date. Bubble size is stock value at cost; colour is the recommended action. Click a bubble to load it.
| Option | Concept | Stock | Full price | Markdown | New price | Season ST | Margin vs no markdown | Action |
|---|
Options, prices and sensitivities are illustrative. The production version runs on the retailer's weekly merchandising extract; figures indicative.
Earlier, option-level markdowns cleared more of the buy at better prices. The gain came from four places, not one.
% of season buy unsold at season end, by concept
Gross margin % on clearance options, season on season
Figures indicative; validated metrics available on request.
A 20% markdown with eight weeks left often earns more than 50% with two. Timing is worth as much as depth.
Fast sellers keep their price; slow ones get cut harder. Blanket rates give margin away on both.
When buyers see margin against clear-out on one screen, they make better calls than any fixed rule.