Reading Sell-Through: Reordering Lamps Before the Shelf Goes Empty

The most expensive unit in lighting retail is the one a customer wanted on the day you had none. Lamps are bulky, visible and disproportionately requested once a shopper has imagined one in the room, which makes the cost of a stockout higher than the unit margin suggests — the customer who leaves empty-handed often buys the sofa elsewhere too. The discipline that prevents this is unglamorous: a weekly sell-through read per SKU, a reorder point set before you need it, and a supplier whose domestic stock makes the last mile of the plan fast.

What to measure, and how often

Sell-through rate is units sold divided by units received, read over a fixed window. For lamps, a weekly read per SKU is the useful cadence: monthly is too slow to catch a fast mover before it gaps, and daily produces noise. Track it in units, not dollars — a slow high-ticket lamp and a fast cheap one must not be averaged into one dashboard number.

• Units on hand per SKU, counted the same way every week
• Units sold in the trailing week and trailing four weeks
• Weeks of cover: on-hand divided by average weekly sales
• Reorder point: expected demand over the replenishment lead time, plus buffer
• A flag on any SKU whose weeks of cover drops below the reorder lead time

Setting the reorder point honestly

The reorder point is demand during lead time plus safety buffer. With an overseas factory order, lead time is measured in months and the buffer has to absorb production and freight variance — the import calendar maps where those weeks actually go — which is why lamp programs built on factory-direct replenishment chronically over-order just to feel safe. With a domestic stocking supplier shipping within a business day and delivering in days, lead time collapses to a week or two, and the reorder point falls with it. The same service level can be held with materially less capital tied up on the floor and in the stockroom.

Making the weekly read a habit instead of a project

The spreadsheet fails when it lives on one person's desktop and depends on their memory. Fix the ritual instead: same day each week, same fifteen minutes, one owner and one named backup. Read the numbers out loud in a standing huddle — on hand, weeks of cover, action — because spoken numbers get argued with and argued-with numbers get corrected. The meeting is not about analysis; it is about catching the two or three SKUs whose cover has quietly dropped below the line and turning each one into a dated action before the weekend traffic arrives.

Keep the actions visible. A reorder that was 'flagged' three weeks running is a stockout that has been scheduled with extra steps. The read earns its keep only when every red line leaves the meeting with an owner and a date, and when the same SKU does not come back red twice without a different decision behind it — deeper buffer, faster channel, or a conversation with the supplier about why the last replenishment landed late.

The patterns worth acting on

Three patterns recur in lamp assortments. A model that sells steadily for weeks and then spikes is usually featured somewhere — find out what before reordering, because the spike may or may not persist. A model with strong traffic and weak conversion is a presentation problem: check the bulb in the display, the position of the switch, and whether the price ticket survived cleaning. A model whose returns cluster on one defect is a supplier conversation, not a reorder decision — freeze the reorder until the fault is understood.

Sell-through data is also your negotiating position at tier reviews: a steady four-week curve per SKU is the strongest evidence for a better price band.

Frequently asked questions

What weeks-of-cover number should trigger a reorder?

One week more than your realistic replenishment cycle, plus a buffer week for demand variance. If your supplier ships in a day and you receive in under a week, a reorder point around two to three weeks of cover is defensible; stretch it only for seasonal peaks.

Do I need software for this?

No. A shared spreadsheet with one row per SKU and five columns — on hand, sold this week, sold four weeks, weeks of cover, action — catches most problems. The discipline of the weekly read matters far more than the tooling.

How do seasons change the plan?

Lamp demand lifts around move-in seasons and the gifting window, so raise the buffer ahead of those peaks rather than holding elevated stock year-round. Your trailing-four-week figure will tell you when the lift actually starts in your region, which is usually earlier than instinct suggests.