Allocation that reflects
what is actually arriving.
Inventory and allocation teams are downstream of every planning, sourcing, and production decision — and they absorb the consequences of every slip. The allocation engine runs against the buy that was committed, not the inbound that is actually coming.
RetailNorthstar joins the buy plan, in-DC dates, and store-level demand into one model — so allocation reflects current reality and replenishment respects the seasonal arc.
The week that everyone reconciles. Every week.
Allocation is the function that ties planning intent to store reality — and the function that gets blamed when the supply side does not match. Most of the failure is not allocation logic; it is the gap between the plan and the inbound.
Inbound delays force reactive allocation
You build the allocation plan against the planned in-store date. The PO slips two weeks. Now stores get whatever shows up first — the original allocation logic is moot. The cycle repeats every season.
Replenishment systems run on actuals — without demand-side context
The replenishment engine sees that Store 47 is selling well and ships more units. It does not know the season is two weeks from transition. The reorder lands during markdown, the inventory liability shifts to the wrong store, and the markdown calendar absorbs the cost.
DC capacity surfaces at receiving — not at PO commit
Three vendors deliver in the same week because the original plan had different ex-factory dates. The DC is overwhelmed; some shipments wait at the dock. Stores receive late, allocation breaks, and nobody saw it coming because PO commitments were never roll-up checked against DC capacity.
Cross-functional reconciliation eats half the week
Monday: allocation team updates the spreadsheet. Tuesday: planning sees it differs from their view. Wednesday: buying confirms what was actually committed. Thursday: logistics flags shipments that already sailed. Friday: a meeting to align everyone on what is true. Repeat next week.
Allocation logic that knows what is actually inbound.
Five decisions allocation teams make every week — and what RetailNorthstar surfaces for each.
Each decision pulls from a specific signal set. When those signals live in different tools, the decision becomes a meeting. When they live in one model, it becomes a click.
| Decision | Signals Joined in the Platform |
|---|---|
| Initial allocation by store group | Buy quantities by style-color-size, store cluster definitions, planned in-store date, prior-season sell-through patterns by store cluster. |
| Re-allocation after inbound slip | Updated ex-factory and in-DC dates, current store on-hand, weeks-of-supply by store, sell-through against plan to date. |
| Replenishment trigger | Store on-hand vs. min/max threshold, days remaining in season, markdown calendar position, vendor lead time, current WIP for the same style. |
| Markdown timing by store | Sell-through against plan, weeks of supply, store cluster performance, prior-season transition timing, end-of-season exit target. |
| Cross-store transfer | Store-level on-hand vs. demand, transit cost vs. markdown alternative, remaining selling weeks, size-level coverage gaps. |
Questions from inventory and allocation teams
How does this connect to our WMS and TMS?
RetailNorthstar is the planning and allocation-decision layer; your WMS handles physical inventory and your TMS handles transportation. We integrate via API or scheduled imports — receipts, on-hand by location, shipment status flow into the allocation view, and allocation plans flow out to the WMS for execution. We do not replace these systems; we provide the demand-side context they need to support the season.
Can we use this if we do not yet have store-cluster definitions?
Yes. Most teams start with the cluster structure they have (often something like A/B/C tiers by volume or geography) and refine it once the platform is live. RetailNorthstar can also propose clusters from sell-through patterns — grouping stores that performed similarly on prior-season styles. Cluster definitions are configurable per category and per season; you do not have to commit to one taxonomy forever.
How does replenishment respect the seasonal arc without manual intervention?
Each style has a season assignment with a planned start, peak, and exit window. The replenishment engine checks the current week against that window before proposing a reorder. If the style is within 4-6 weeks of its planned exit, replenishment proposals automatically scale down or stop. Markdown windows trigger a hard suppression. Teams can override at the style level for staple/replenishment categories that do not follow a seasonal arc.
Allocation that reflects the inbound, not the original plan.
See how RetailNorthstar joins buy quantities, in-DC dates, and store-level demand — so allocation is dynamic and replenishment respects the seasonal arc.
Connected apparel planning — live in weeks, not quarters.