Your range is a portfolio. Manage it like one.

A big supermarket can stock forty thousand products and let statistics sort them out. You cannot. A small shop's range is a portfolio of a few hundred bets made with brutally finite shelf space — and unlike the supermarket, every yes you say to a product is a no to something you will never see fail, because it was never given the chance.

Ranges grow by accumulation

Ask how any product joined the range and the answers are archaeology: a rep was persuasive in 2023, a customer asked twice, it came in a bundle, it has always been there. Each entry was reasonable. Nobody ever ran the exit process, because there isn't one. The result is a range that reflects the shop's history far better than its customers' present.

The corrective is the review we described in judging products on contribution, not charm — but assortment optimisation adds the portfolio questions that product-by-product review misses:

Coverage, not count. Do the products overlap — four mid-priced versions of the same need — while whole price points or needs sit empty? Duplication feels like choice to the owner and looks like clutter to the customer.

The role test. Every product should have a job: traffic driver, margin earner, basket builder, credibility signal. A product with no job is shelf decoration with a carrying cost.

Space against contribution. Metres of shelf are your scarcest asset. If a category takes a fifth of the space and produces a twentieth of the margin, the range is arguing with the shelf plan, and one of them is wrong.

Add by swap, not by addition

The single most protective rule in range management: after the range is set, nothing comes in unless something goes out. It forces the comparison that accumulation avoids — is the new bet better than the weakest current one? — and it caps the silent growth in tied-up cash that we traced in The cash frozen on your shelves.

Kauzio keeps the portfolio view current — contribution, velocity, overlap, dead weight — so a range review is an hour with a clear page instead of a weekend with a spreadsheet. The shelf is finite. The range should be chosen, not accreted.