Can you trust your shelf planning software?
Shelf planning software only produces a range-review scenario worth approving when it is fed accurate product dimensions, real fixture geometry, store-format data, ranging and replenishment limits, commercial measures, and current execution evidence. When any of those inputs is wrong, stale, or unowned, a plan that looks finished on screen can still be commercially unreliable.
A range review can reach the approval meeting with clean facings, neat blocks, and a convincing category story. None of that proves the numbers behind it are current, or that the shelf can be built and held in a real store. A single wrong pack dimension, an idealised fixture file, or a replenishment assumption nobody owns is enough to turn a polished plan into a costly reset.
The useful test is traceability. Every material input should carry an owner, a verification date, a store-format scope, and a status. A simple three-status register, marking each input as verified, assumed, or decision-pending, keeps a working estimate from being read as an approved fact. Most range-review problems are not software faults. They are decision-control gaps that only become visible once the model is built.
What data does shelf planning software need before modelling begins?
Shelf planning software depends on three kinds of data: physical facts about what fits, operational facts about what can be replenished, and governance facts about which rules the model must follow. Product imagery on its own is not enough to model a shelf.
Product records need unit dimensions, pack orientation, and the version of the pack actually being ranged. Case dimensions often differ from unit dimensions. Display-ready packaging changes usable depth, height, and facing capacity the moment it is opened on shelf.
Fixture data needs the usable shelf width and depth, shelf thickness, uprights, dividers, lips, hooks, and any blocked space. A nominal bay width flatters the plan when the real fixture loses room to elements that never hold stock.
The model also needs facings, shelf position, capacity, adjacency rules, pack-out limits, availability, and replenishment frequency. A shelf that technically holds the range can still fail in trading if the fast sellers cannot be refilled within the store’s service model.
Every source should carry an owner and a last-verified date. A visually complete model becomes unreliable the moment its dimensions, fixture files, or availability assumptions fall out of date.
| Input group | Minimum record | Common weakness |
|---|---|---|
| Product | Unit and case dimensions, orientation, pack version, display-ready format | Old artwork or packaging measurements left in the file |
| Fixture | Usable dimensions, shelf positions, dividers, hooks, blocked space | A nominal bay size used instead of the real fixture geometry |
| Operations | Facings, capacity, replenishment frequency, pack-out limits | The model assumes more stock or labour than the store can support |
| Governance | Owner, source, verification date, format scope, status | An assumption enters the model with no named decision owner |
What commercial assumptions belong in a range review?
A range review needs the commercial logic that explains why each product earns space, holds its position, moves, or leaves the fixture. That logic should sit beside the physical plan, not behind it.
Sales and rate-of-sale figures are a starting point, not a verdict. Margin, category role, strategic and destination lines, range architecture, supplier terms, and availability risk all change how the raw numbers should be read.
Delisting and substitution decisions need the most care. Cutting one line can push demand onto another SKU, open a gap in a price tier, weaken a brand block, or strip choice from a specific shopper mission. The model should record the expected effect and name the person accountable for approving it.
The three-status register keeps evidence and judgement apart:
- Verified input: a current fact with a named source, owner, date, and format scope.
- Working assumption: a reasonable estimate used to build the scenario, with its uncertainty stated.
- Decision-pending: a commercial choice still waiting on an accountable owner to approve or reject.
An approved decision should never be buried inside an input file. It needs a visible record of what was chosen, why, and what was rejected. That record earns its keep at the next review, when someone asks why the shelf is built the way it is.
Why is one average store not enough?
One average store is rarely enough, because a store network holds different formats, fixture conditions, service models, and range depths. An idealised store hides the exact places a plan will break during execution.
Take a retailer running flagship, standard, convenience, and regional formats. The category strategy can stay constant while fixture width, aisle approach, shelf count, stockroom space, and replenishment frequency all shift by format.
A plan built for the standard store can overload the convenience format, under-range the flagship, or pile stock pressure onto a regional site. Local fixture quirks can also block a sightline or remove a shelf position the plan assumed was there.
Store clusters are better modelled against rules that separate what stays fixed from what can vary. Fixed elements might be the category role, core range, price architecture, and required adjacencies. Variable elements might be range depth, facing count, shelf count, promotional space, and local lines.
A dependable range review shows how the recommendation holds across the formats you actually operate. It also flags the exceptions that need their own planogram or an operational call before rollout.
What can shopper evidence add to a shelf plan?
Shopper evidence tests whether an efficient shelf plan is visible, understandable, and usable from the aisle. It adds observed behaviour to a model built mostly from product, fixture, and commercial data.
Simulated shops and eye tracking can show approach direction, product visibility, gaze sequence, attention, and selection. Storelab Research uses these methods to expose a weak category cue, a blocked view, or a product that is present in the plan but hard to find in the aisle. That kind of evidence earns its place when two scenarios look commercially alike: one may guide shoppers through the category cleanly, while the other sends them scanning back and forth or draws the eye to the wrong block.
The method has limits worth stating plainly. Eye tracking shows where attention went. It does not settle margin, ranging, supply, pricing, or strategy. A virtual study is only as sound as the realism of its products, fixtures, task, sample, and design. Shopper evidence should challenge the plan where it adds something real, then sit alongside live trading data and professional judgement rather than override them.
When are the inputs too weak to approve the plan?
Inputs are too weak to approve when the plan cannot be traced back to current facts, owned assumptions, and store conditions that can actually be executed. Approval should stop the moment a material weakness could change the range, space, or rollout decision. The common signals:
- Missing or stale dimensions: product or fixture data no longer matches the physical shelf.
- Unclear format assumptions: reviewers cannot tell which stores the scenario represents.
- Unowned commercial inputs: margin, range role, substitution, or supplier assumptions have no decision owner.
- Conflicting sources: teams are working from different sales, product, or fixture files.
- Unrealistic replenishment: planned capacity relies on delivery frequency or labour the store does not have.
- No decision record: the chosen scenario has no stated reason, owner, or rejected alternative.
- Known execution failure: the plan cannot be built or maintained in one or more important store variants.
The rule is simple. No scenario should move forward unless every material input traces back to an owner, a date, a format, and a status. Where one of those is missing, fix the weak input before the shelf plan is approved, not after the reset has gone to stores.
Before the next range review, it is worth checking the product data, store assumptions, and approval workflow together, in that order. Storelab can help retail and FMCG teams model the options, test the shopper questions worth answering, and carry the approved decision through to execution.