Case Study · Retail
Optail Initial Allocation: first distribution of new product with a master plan, store grouping and capacity
When a new product leaves the warehouse there is no sales data yet; the decision rests on similar products' history, store groups and capacity. For Derimod we built the Master Plan screen where planning is done before and throughout the season, store clusters, fill and dispatch priority management; footwear and bags went live first, then apparel. In 2026 Flow Through, open-quantity allocation and simulation are being added.
- 3
- Categories
- Master plan
- Decision input
- Fill
- Capacity
- Flow Through
- 2026
Footwear → bags → apparel, in that order
Option × store group × depth × date
Category-level ceiling; no dispatch to a full store
To the store without waiting in the warehouse
Problem
Replenishment and transfer work with sales data: where a product sells, where it does not. In initial allocation that data does not exist. When a new collection enters the warehouse, which option goes to which store, in what quantity and which sizes, must be decided before the product reaches the shelf. At Derimod this decision was made store by store, in the planning team’s Excel, at the start of the season; once made, which store was full and which product suited which store group was never asked again systematically.
The master plan
Initial allocation became Optail’s third and most input-hungry module. At its centre is the Master Plan screen: where planning is done before and throughout the season.
| Input | What it sets |
|---|---|
| Active plan and product hierarchy | Which options are in the season’s plan, in which gender–category–subgroup |
| Store grouping and priority | Which store groups an option goes to; groups are built by gender, category and price range (clusters) |
| Dispatch depth | Config (assortment) and single dispatch depths — how many units per store, which sizes |
| Dispatch dates | When an option becomes shippable |
| Dispatch approval, bans, recall orders | Approval limits; bans on specific store–option pairs; recalls |
The planner looks at the plan and changes it; the screen shows order quantity, warehouse stock, store stock, in-transit stock, sales and revenue next to the plan. The plan discipline from Excel moved inside the system; the plan is no longer a file but a table the algorithm reads.
The algorithm
For every option entering the warehouse the algorithm takes the master plan’s store groups, orders the stores by priority and builds the size distribution for each store according to depth. Two checks on top:
- Capacity and fill. Every store has a ceiling per gender × category; nothing is dispatched to a full store, and as capacity opens the next option goes. The store capacity dashboard shows these limits live.
- Collection management. An option does not go in initial allocation to a store that does not carry it in its collection; out-of-collection movement is transfer’s job.
The result is a proposal at store × option × size level; the approved proposal goes to AX as a dispatch order, the warehouse picks, the goods leave for the store.
We went live in sequence: first footwear — Derimod’s core category and the clearest size structure; then bags — sizeless, with a different assortment logic; finally apparel — women’s and men’s in the 2026 season, with separate size sets. In every category the first results were compared line by line with the planning team; “how different is the system’s distribution from our Excel, and why” turned each time into either a parameter or a rule.
2026 developments
- Flow Through. Goods accepted at the warehouse have their stores determined in a holding area and are made ready to ship the same day, without entering the put-away, storage, picking and handling steps. A pre-dispatch order is created before the product appears in stock; the ERP and the warehouse operations system are being changed accordingly.
- Open-quantity initial allocation. Giving a system that only ships assortments the flexibility to build an assortment from single units.
- Initial allocation simulation. Seeing the season’s stock distribution as a scenario before shipping it.
Why this way
- The plan must live inside the system. Had the master plan stayed in Excel, the algorithm would read a different file every season and the gap between plan and result could not be tracked.
- Capacity is a limit, not a target. The fill ceiling protects a full store; but dispatch priority is built on sales potential, not on capacity.
- Category by category. The trust earned in footwear carried to bags, and from bags to apparel; had all three opened at once, none would have settled.