Skip to main content

Understanding Soft Allocation in Dynamic Replenishment

Explains how Soft Allocation enriches replenishment target-location suggestion with other open tasks' virtual placements, how Replenishment Algorithm rule order shapes the outcome, and what behavior changes to expect once it's enabled.

Overview

Soft Allocation changes what the replenishment target-location suggestion logic treats as "existing inventory" when it evaluates a candidate location. Normally, suggestion only looks at committed inventory — what's actually sitting in a location today. With Soft Allocation enabled, it also factors in the target locations already suggested to other non-completed replenishment tasks generated in the same run, treating their quantities as if they were already placed there. In other words, the Replenishment Algorithm's conditions (empty location, same SKU, capacity, and so on) are evaluated against committed inventory plus these pending virtual placements — a more complete, up-to-date view of the warehouse than committed inventory alone provides.


When to Use It

Enable Soft Allocation when:

  • You want target-location suggestion to account for inventory that's about to exist (from other open replenishment tasks), not just inventory that already exists.

  • Your Dynamic Replenishment wave rule regularly generates more than one task per run for the same product, and later tasks should "see" what earlier tasks in the same run already claimed.

  • Your Replenishment Algorithm's conditions — empty location, same SKU, capacity, and similar — should be evaluated against this fuller view rather than committed inventory alone.

  • You want the Suggest Another Location flow during the put phase to be evaluated against that same enriched view.

Prerequisites

Before enabling Soft Allocation,

  • Dynamic Replenishment itself must already be set up —

    • Replenishment levels on your products,

    • Source/target locations

    • Dynamic Replenishment job type

    • An assigned Replenishment Algorithm.

Soft Allocation is controlled at the wave rule level, so the wave rule governing your replenishment job type must already exist.


Step-by-Step Instructions

  1. Navigate to the Wave Rules screen.

  2. Open the wave rule that governs your Dynamic Replenishment job type or create one.

  3. Enable Soft Allocation for Dynamic Replenishment on the wave rule.

  4. Click Save.

  5. Run the wave rule (or wait for its schedule) to generate replenishment tasks.

  6. Confirm: for multiple tasks generated in the same run, each task's suggestion reflects the locations/quantities already suggested to earlier tasks in that run — not just committed inventory.

If Soft Allocation is left off, target-location suggestion behaves exactly as it does today — only committed inventory is considered.


How Replenishment Algorithm Rule Order Shapes the Outcome

Soft Allocation enriches what the Replenishment Algorithm sees as existing inventory — it doesn't change what the Replenishment Algorithm looks for. So the same enriched view can produce very different outcomes depending on which rule your Replenishment Algorithm checks first.

Example 1— "Suggest an empty location" as the 1st priority rule
With no same-SKU rule ahead of it, the Replenishment Algorithm evaluates "is this location empty?" against committed inventory plus other open tasks' virtual placements.

  • Task 1 (Lot A): no location holds any inventory (real or virtual) yet, so Location 101 qualifies as empty and is suggested.

  • Task 2 (Lot B, same SKU, different expiry): Location 101 no longer evaluates as empty, because Soft Allocation now counts Task 1's virtual placement there as existing inventory. Location 102 is the next location that evaluates as empty, and is suggested instead.

  • Task 3 (Lot C, different receiving date): by the same evaluation, Location 103 is suggested.

The Replenishment Algorithm was only ever asked "is this location empty," so it correctly keeps finding the next location that evaluates as empty — spreading the same product's inventory across several target locations whenever lot, expiry, production date, receiving date, or PO number differs.

Example 2— "Check Assigned SKUs" (or "Suggest a location that has same SKU") as the 1st priority rule, "Suggest an empty location" as the 2nd


Here the Replenishment Algorithm first evaluates "does any location already hold this SKU" against the same enriched view:

  • Task 1 (Lot A): no location holds this SKU (real or virtual) yet, so priority 1 finds nothing and the rule falls through to priority 2, which suggests empty Location 201.

  • Task 2 (Lot B, same SKU, different expiry): priority 1 is evaluated again — and this time it does find a match, because Soft Allocation counts Task 1's virtual placement at Location 201 as existing SKU inventory. Task 2 is suggested Location 201 too, even though its lot/expiry/production date/receiving date/PO number differs from Task 1's.

  • Task 3 (Lot C): the same evaluation applies, and Location 201 is suggested again as long as capacity remains.

The two examples run the identical enriched-inventory evaluation — the difference in outcome comes entirely from which condition the Replenishment Algorithm checks first.

⚠️ In order to acquire more efficient inventory distribution, the use case in Example 2 is highly recommended compared to use case in Example 1.

Only seeking empty locations may result very fast depletion of available locations or over-distrubution of identical products.

Instead, using an algorithm action that seeks for a match based on existing inventory as a prior rule (Check same SKU/Client/Product Group etc.) and then applying empty location search will maximize the efficiency of the soft allocation system.

Example — "Suggest Another Location" during the put phase
When a picker requests an alternate location mid-put, that request is evaluated against the same committed-plus-virtual view as the original suggestion, so it stays consistent with everything else suggested in that run.


⚠️ What to Expect When Soft Allocation Is Enabled

Soft Allocation changes the pool of information the Replenishment Algorithm evaluates against, which can surface a few behavioral shifts worth planning for:

  • Suitable locations may deplete faster than usual.

    • Because locations are counted as claimed as soon as they're virtually placed — not just once they're actually filled — you may see "no eligible location" outcomes sooner than you would without Soft Allocation.

  • Task generation may take longer than usual.

    • Each suggestion now has to be evaluated against a larger, enriched view (committed inventory plus other open tasks' virtual placements) instead of committed inventory alone.

  • A task that would have completed as a full LP pick may no longer be one.

    • If an LP contains the same product across different attributes (lot, expiry, production date, receiving date, or PO number), Soft Allocation can spread those units across multiple virtually-claimed locations instead of consolidating them onto one — the exact outcome depends on which Replenishment Algorithm actions are in use.

  • "Suggest Another Location" attempts may offer fewer alternatives.

    • Since Soft Allocation has already virtually claimed some of the locations that would otherwise be available as alternates, pickers may see a smaller set of options to choose from.

⚠️ Known Limitation: Volumetric Putaway Actions

Soft Allocation does not extend to using volumetric putaway action in the Replenishment Algorithm. Occupied volume at a location is only updated when physical inventory actually moves in or out of that location — a non-completed replenishment task's suggested inventory is not physically there yet, so it cannot be reflected in a location's occupied volume.

This means that if your Replenishment Algorithm uses a volumetric putaway action, that condition is evaluated against real, physical occupied volume only — it does not account for the virtual placements from other open replenishment tasks in the same run. Volume-based actions can therefore still suggest a location that Soft Allocation would otherwise treat as constrained for non-volumetric conditions (empty location, same SKU, and so on).

Support for volumetric conditions under Soft Allocation is planned for a future release.

Common Questions & Edge Cases

Does Soft Allocation change what's shown on the Dynamic Replenishment Report or other inventory screens?
No. The enriched view is used only for suggestion evaluation. Reported on-hand and available quantities always reflect real, committed inventory only.

What happens to a virtual placement once its task is completed?
Once the put is executed, the target location's real inventory reflects the placement — the location now genuinely holds that inventory, rather than holding it virtually.

What happens if a task with a suggested location is cancelled or skipped?
Its virtual placement is removed from the evaluation view, so later tasks in the run no longer see that location or SKU quantity as existing there.

What happens if no eligible location remains?
If the Replenishment Algorithm's conditions can't be satisfied against the enriched view, the task fails to generate a target location and hence task creation would be failed for the inventory that is subject to the replenishment.

Does Soft Allocation work the same way across multiple warehouses?
Yes. Each warehouse's wave rule has its own independent Soft Allocation setting, so the enriched inventory view in one warehouse never includes another warehouse's tasks.

Did this answer your question?