Your Reorder Points Are a Guess Wearing a Number

Every distributor and every shop has one: a reorder point nobody can explain. The number sits in the system, it triggers a purchase order every few weeks, and when you ask where it came from the honest answer is that someone set it years ago and then left.

That number is a guess wearing the costume of a decision. It looks authoritative because it lives in software and prints on a report. Nobody has tested it against reality since the day it was typed.

Where the number actually came from

In most founder-operated businesses, reorder points get set one of three ways. Someone eyeballed a year of usage and rounded up. The software imported a default during implementation and nobody changed it. Or a stockout hurt badly enough that somebody doubled the number in frustration, and it never came back down.

None of those are wrong on purpose. They are what happens when a business grows faster than its definitions. The number stops getting reviewed the moment it works well enough to stop generating complaints.

Meanwhile the business changes underneath it. Lead time from that vendor went from two weeks to five. A customer who bought monthly now buys quarterly. You added a second location drawing from the same stock. The number knows none of that.

Three questions that expose it

You do not need new software to find out whether your reorder points are real. You need three answers per item, and you can start with your top 20 by dollar volume.

First, what is the actual lead time, measured from PO date to received date, across the last 12 months? Not what the vendor promises. What the vendor delivers.

Second, what is real demand during that lead time, including the spikes? Average usage hides the month that emptied the shelf.

Third, what does a stockout on this item actually cost? A stockout on a fast-moving consumable is an inconvenience. A stockout on the part that halts an install is a customer.

Those three answers turn a reorder point into a decision you can defend. They also show which items have earned safety stock and which are quietly tying up cash you could use somewhere else.

This is a definition problem, not a software problem

The reflex is to fix this with a purchase: better forecasting, an inventory module, a demand engine with AI in the name. Any of them can help. None of them will save you here, because every one calculates from the same inputs you have not defined.

Feed a forecasting tool an untested lead time and it returns a confident, precise, wrong answer faster than a person could. You can’t optimize what you haven’t defined. Automating an undefined reorder point means you now get the wrong purchase order automatically, on schedule, and nobody reads it because the system generated it.

What good looks like

A defined purchasing workflow is not complicated. It has a documented method for setting a reorder point, a named owner for reviewing it, a schedule for that review, and a short list of triggers that force an off-cycle look: a vendor lead time change, a lost account, a new location, a price break that changes the order quantity.

Write that down and two things happen. Buying stops depending on one person’s memory, which means that person can take a vacation. And the numbers become auditable, so when something goes wrong you can find out why instead of guessing again.

Once that exists, technology finally has something worth pointing at. Reporting can flag the items whose lead times drifted. Automation can draft the purchase orders. A model can propose adjustments for a human to approve. That sequence works. The reverse is how businesses end up with expensive software and the same stockouts.

Start with your top 20 items and the three questions. Most owners find at least one number that has not been true for years.