Ethan Rogers
← Work

IRIS TPO

Scenario planning for a year of retail promotions. Planners branch scenarios, edit them on a live 52-week calendar, and see what every change is worth before choosing the plan to back.

Client
Kimberly-Clark
Year
2026
Role
UX Design, Interaction Design, Front-end Development
Stack
Next.js, React, TypeScript, Tailwind CSS, shadcn/ui, Recharts, Motion
IRIS TPO

Challenge

Trade promotion planning is dense by nature.

One plan covers a retailer for a whole year: 100+ promotional events across sixteen product groups, each with a depth, price, timing, lift, redemption rate and fees, all feeding revenue, profit, ROI and a trade budget. Planners don't make one plan, they weigh several — today's baseline, what the optimizer recommends, and their own variations — and until now that comparison lived in spreadsheets. They needed to try ideas without losing track of where each version came from, see a year of events as a shape rather than a list, and walk into an approval meeting able to explain why one scenario beat another.

Approach

Scenario planning is the spine of the product.

Every plan starts from a read-only Original Scenario, the baseline everything is measured against. An optimization branches an Optimized Scenario from it, and planners branch Simulations from either. The scenario library nests them by lineage, so you can always see what a version was built from, and every revenue, profit, spend and ROI figure carries its change against the baseline. A simulation measures itself against its parent: edit depth, price, timing, inclusion, lift or fees and the live impact updates instantly, a condensed impact bar stays pinned as you scroll, and a changes drawer lists every edit with its own impact and an undo. Each event is tagged as an earner or a burner against the trade budget, so the trade-off shows up at the row, not just in a total. A one-line guide on every plan page shows which of four steps you're on and the one button that moves you forward. The promo calendar turns the year into a picture. Every product group is a row across 52 fiscal weeks, and every event is a bar you can recolour by what you're asking: the optimizer's recommendation (deeper, shallower, hold, locked, removed), discount depth, promo strength by ROI, or tactic. Coloured by ROI, a year of events becomes a heat map: weak promotions glow red and strong ones deepen to teal, so the events worth pausing jump out without reading a number. Edited events are outlined, excluded ones dashed, and committed weeks are frozen and locked. The calendar is editable, not just a view — drag an event to move it, drag an edge to resize it, and it snaps to whole fiscal weeks and stops at its neighbours — and the same edits flow back into the grid and the live impact. You can zoom to any quarter, and the colour choice is kept in the URL so a shared link opens on the same view. The compare view closes the loop. Planners pick two to four scenarios and a baseline, and the decision comes first: a recommendation card with the suggested scenario, an optional note for approvers and one Save button. A KPI matrix tags the leader on each metric, and a gains-and-losses summary spells out what each scenario wins and gives up. Every screen also carries built-in design notes (purpose, layout, components, MVP versus future), so the prototype doubles as the hand-off spec.

IRIS TPO, screen 1
IRIS TPO, screen 2
IRIS TPO, screen 3
IRIS TPO, screen 4
IRIS TPO, screen 5
IRIS TPO, screen 6

Outcome

IRIS TPO went from Figma frames to a product stakeholders can click through end to end, from opening a plan to saving the recommended scenario.

Scenario lineage and the colour-coded calendar let planners compare versions of a full-year plan without opening a spreadsheet, and every recommendation arrives with its reason and its price attached. Writing the design notes into the product kept MVP and future scope explicit, and gave engineering a living spec instead of a slide deck.

Next projectDRCS →

Open to design technologist & UX engineering roles

Let’s makesomething good.