Retail Shelving Rollout: CRI 95 Strip Across 180 Stores in 11 Months

The starting condition

Each store has 26–32 m of shelf strip across four floors, and stores were being refitted in clusters of 12–15 per month. The previous supplier delivered on a 12-week lead time with no bin commitment, so stores refitted six months apart had visibly different whites in the same product category. The retailer had also specified CRI 90, which left red garments looking flat under the shelf light.

Constraints

Measured results

Colour-match complaints

9 in the previous 60-store rollout → 0 in 180 stores

First twelve months after installation

Fitting labour per store

3.5 hours → 2.1 hours

Pre-terminated reels, no site soldering

Replacement lead time

12 weeks → 2 days

From the central spare allocation

Colour rendering

CRI 90, R9 18 → CRI 95, R9 62

Verified on the integrating sphere per lot

What we changed, and why

1. Reserve the die bin before the first store The full 24,000 m was reserved against one CRI 95 die bin before the rollout started. Colour drift between stores was the single largest risk in the programme, and holding the bin was the only control that worked at this scale — re-binning mid-programme would have produced a visible difference between store clusters.
2. Move the specification from CRI 90 to CRI 95 We supplied CRI 90 and CRI 95 samples for a side-by-side test on the same rail of garments. The retailer changed the specification on the evidence: reds and navy read correctly under CRI 95 with R9 at 62, and the 12% cost increase was accepted against the cost of a second refit.
3. Pre-terminated reels per store Each store received its reels cut to the installed length with JST SM connectors fitted and pull-tested, labelled by floor and bay. Store fitters had been cutting and soldering on site, which produced inconsistent joints and about 40 minutes of extra labour per store.
4. Bin record and spare allocation per cluster Every store has a bin record filed against its cluster, and 2% spares from the same bin are held centrally. A failed section is now replaced from stock in two days rather than triggering a production order for a single reel.

“The rollout was boring in the best way. Every store looked the same, which is all we wanted and what the last supplier could not deliver.”

Store development manager, Fashion retail chain, Germany — Fashion retail chain, 180 stores

Where this approach stops working

A reserved bin works above roughly 5,000 m per programme. For a single-store refit the correct answer is stock strip from a current bin plus a labelled spare reel; reserving production for 32 m would multiply the unit price for no visible benefit.