Stockouts, overstock and stock in the wrong place: what an inventory planner actually solves

Part 1 of two on the reorder problem. Part 2, why good inventory planners still miss, is about what remains unsolved after the tool is bought.

Take a typical multichannel product brand, invented here to make the problem concrete. It sells its best-selling dutch oven on its own website, on Amazon, and through a handful of retailers. In August the dutch oven was unbuyable on Amazon for nine days. At the same time, the outside warehouse that serves the website was holding six weeks of it. A promotion email had gone out on the third. The container from the supplier, promised for the first, cleared customs on the twelfth. Nobody did anything wrong on any single day. Demand moved, the supplier ran late, and the stock sat in the wrong place, all at once, while the person deciding what to buy next worked from a Monday spreadsheet.

That is the reorder problem as it actually presents itself to a brand selling on more than one channel. There are two ways to get it wrong, and a third that only multichannel brands get to experience.

• Out of stock on a winner. Sales stop the day it happens. On Amazon the product also loses ranking, and since April 2024 a product that runs thin for too long attracts a low-inventory-level fee on every unit it does sell. The cost is not the missing week. It is the weeks it takes to earn the position back.
• Buried in a loser. Cash that could have bought the next container is sitting on a shelf as last season’s colour. Amazon charges a monthly aged-inventory surcharge on stock that has sat in its warehouses for six months, and the outside warehouse charges storage regardless. The way out is usually a markdown, which turns the cash into less cash.
• The right amount in the wrong place. Six weeks of cover at the warehouse while Amazon runs dry is a stockout and an overstock on the same product at the same time.

Both worries tend to get buried under everything else a growing business demands, until one of them lands on a Monday morning as a fact. Owners are not careless about this. They are running six functions at once, and the reorder is the one that only makes noise when it is already too late.

What a planner is actually solving

Strip away the vendor language and an inventory planner is doing one job: turning four uncertain things into a purchase order with a quantity and a date. It helps to know what each of the four is protecting against, not how to calculate it.

The forecast is what the brand thinks will sell, per product and per channel, over the coming weeks. It protects against buying for last year’s business instead of this year’s. It is built from history, which is also its limit: it cannot see a promotion that has not happened yet.

The lead time is how long replenishment really takes, from placing the order to the goods being sellable. Not the supplier’s quote. The real figure, including freight, customs, and the warehouse getting it onto a shelf. It decides how early the brand must order, and every week it is wrong is a week of stock the brand did not plan for.

Safety stock is how much uncertainty the brand chooses to absorb. It is the extra held on top of what the forecast says will sell during the lead time, sized to how much sales wobble and how late the supplier tends to be. Too little and the best seller runs out on a normal bad week. Too much and the cash is on a shelf. Most planners size it with a version of the same textbook formula; the choice the brand makes is how much risk of running out it will accept.

Allocation is where the stock needs to sit so that the channel can actually sell it. The warehouse serves the website. Amazon serves Amazon from its own buildings. Retailers take pallets. A brand can hold exactly the right total and still be out of stock where the customer is looking.

That is the whole machine. Forecast, lead time, safety stock, allocation, and out comes a recommendation: order this many, by this date, and send this share to Amazon.

This is not an unsolved software category

A brand that goes looking for help with the arithmetic will find plenty, in four rough classes. Amazon-native: Amazon’s own restock recommendations, minimum inventory level per product, and automatic top-ups from its bulk storage, for Amazon’s stock only. Channel-specific planners: SoStocked for Amazon sellers, Prediko for Shopify brands. Multichannel planners such as Inventory Planner and Flieber, which forecast across channels, recommend buys with supplier minimums and safety stock, and recommend transfers between locations. And inventory platforms a brand may already run, Cin7 and Linnworks among them, with forecasting and reorder points built in.

On their own terms these tools are good. The core forecasting and replenishment arithmetic is well understood, and mature software already handles it. None of it is magic.

So the honest advice to a brand still planning from a spreadsheet or from gut feel is short: get a planner. The category is mature, and it is the right first step, provided the tool does not arrive with a second job attached, a month of configuration and training for whoever has to run it.

One thing to understand before signing up, because it is where Part 2 begins. Every recommendation a planner makes depends on the business feeding it current information. The lead time per supplier is true on the day it was typed. The promotion is in the tool only if someone entered it. The product being wound down keeps its full target until someone lowers it. The stock count is only as good as the last time someone corrected it.

The tool knows what it has been told. Keeping it told turns out to be the hard part.