A position says what would prove it wrong. Test each on your own store.

1. A stockout costs future orders, not only today’s.Wrong if customers who hit a stockout reorder at the same rate over the next six months as similar customers who didn’t.
   What a Stockout Costs
2. Rank stockout risk by who buys the product, not by how many units it sells.Wrong if your top ten products by units are the same ten as your top ten by repeat customers.
   Who a Stockout Hits
3. When something is out, give the customer a reason to wait or a good alternative. Don’t pay them to wait.Wrong if customers you gave a discount to wait go on to buy more over the next year than customers you gave an honest explanation.
   Back in Stock, or Something Close
4. The cheapest safety stock is information from customers who already told you what they’ll need.Wrong if your subscription and replenishment data predict next month’s demand for your top products no better than last month’s sales do.
   Information for Inventory
5. Clearing overstock with sitewide discounts moves the problem from the warehouse to your customers’ expectations.Wrong if customers acquired or reactivated during clearance sales buy at full price as often as everyone else.
   The Markdown Trap
6. Before any product changes, name the customers who depend on it.Wrong if your last discontinued product’s heaviest buyers reordered something else at the usual rate.
   Find the Dependents
7. A discontinued product needs notice, a last chance to stock up, and a replacement you’ve tested on the people who’ll lose it.Wrong if the dependents of a product retired with that playbook churn at the same rate as the dependents of one retired without it.
   The Discontinuation Playbook
8. A reformulation is a discontinuation with the same name on the label. Handle it like one.Wrong if repeat buyers of a changed formula reorder at the same rate as before the change.
   Reformulation, and New Coke
9. Cutting duplicates can raise sales; cutting what someone relies on loses the customer, not just the sale.Wrong if customers who lost their usual product in a cut order as often afterwards as customers who didn’t.
   Fewer Products, For and Against
10. Prune by a product’s role in repeat baskets, not by its margin alone.Wrong if the low-margin products you cut last year took no other orders with them.
    Prune by Basket Role