Retail previs offers a rare loop back to real sales data

The picture

Two virtual-reality retail vendors publish, as retrieved on 16 September 2026, materials describing a previs application unlike most others in this manifest: a virtual store a shopper can walk through before a real reset is built, so a prediction made in the headset can, in principle, later be checked against actual sell-through once the physical store opens. OneDoor's own resource page on VR in retail and ScienceSoft's own page on virtual reality in shopping both describe this shopper-testing method: a simulated store replica paired with eye-tracking heatmaps, used to study what draws attention before a physical reset is committed.

What the documents show

Both pages are vendor materials, and both name figures without naming the study behind them. OneDoor's page cites its own '2025 Report: The Cost of Poor Visual Merchandising' for a claim that shopper-expectation misalignment costs retailers '$125 billion'; that is OneDoor's own figure, from OneDoor's own report, not an outcome from OneDoor's VR testing itself. ScienceSoft's page states a return-rate reduction of '60-80 percent' and a project cost 'starting at $150,000', with no baseline, no named client and no comparison to a control group. Neither page cites, or links to, a single published instance of a retailer comparing a virtual-store prediction against the sell-through the retailer actually recorded after building the store. The validation loop the topic describes is structurally possible; nothing in either cited document shows it has been run and published.

What it is allowed to decide

As it stands, documented only by vendor pages, this use of previs can support Class 2 Creative approval, informing merchandising and layout decisions ahead of a build, since eye-tracking data can plausibly guide a designer's judgement. It cannot support a Class 3 Commercial claim of proven return-on-investment, because no cited source discloses the substantiation such a claim requires. It holds Dimensional authority to the extent a store's geometry is modelled to scale, which neither page confirms in specifics. It holds no independently checkable Auditability, since neither the '$125 billion' nor the '60-80 percent' figure comes with a reproducible method a reader could inspect.

The disclosure label

Labelled honestly: a vendor's own shopper-testing product, plausibly Class 2 Creative approval; Dimensional authority held to whatever degree the store model is built to scale; Auditability not held, since the headline efficiency figures are vendor claims without disclosed baseline or independent verification. Dated 16 September 2026, this label would be asserted by the vendor, not by a retailer that has run the comparison.

• Has any retailer using this method published the actual sell-through recorded after the physical reset shipped?
• What population and time window does the vendor's own efficiency percentage cover?
• Would the '$125 billion' misalignment cost hold up if traced back past the vendor's own report?

Retail previs is unusual because reality eventually checks its prediction; until a retailer publishes that check, the loop remains a promising structure rather than demonstrated evidence.