Fastenal's third-quarter net sales rose 15.5% to $2.46 billion, according to its October 14 earnings release. Its operating margin climbed to 21.5%, from 20.7%. Its gross margin slipped to 44.8%, from 45.3%.

Two margins, two directions, one quarter. Fastenal's own release explains why, in basis points.

Before a recent session with an industrial distributor association, we surveyed its members (n=15). 13 said someone at their company has used AI for work. 10 said only a few people on their team have tried it. 13 put customer follow-ups in their top four pain points. And 7 asked for real examples, step by step.

That last number is the reason for this article. Instead of a framework in the abstract, here's one example, start to finish.

A few years ago, I asked an owner what he thought would happen when we rolled out new pricing at his company. He didn't hesitate.

"Nelson, we have a good portion of our salespeople who are old school. They're not going to like this. They've been setting margins at 23% for the last ten years, so 23.5% is going to feel like a lot to them. But our newer people don't want to memorize pricing. They want the system to suggest it and they'll use it."

That one sentence explains more about why pricing technology fails than any feature comparison I've ever seen.

A margin number can be accurate and still tell you nothing useful. That is not a contradiction. It is what happens every time an average does its job.

Margin moved last quarter, and when someone asked why, the honest answer was a guess. That gap between seeing a number move and knowing what caused it is what six specific forces, price, cost, volume, mix, product, and customer, are built to close. At this month's STAFDA workshop, we explored the root causes of that margin variance gap and the steps you can take to close it.

Three numbers from the room first. At the P21WWUG pricing session on September 10, among attendees who responded to the live polls (n=30–32 per question):

• 57% said they have no reason-code field or Alert Maintenance rule tracking pricing overrides.

• 59% either hadn't reviewed their product-group pricing in over a year or don't price at that level at all.

• 33%, one in three, could name the owner of every price library they run.

None of that is a P21 problem. P21 gives you near-total freedom to build pricing however you want. That freedom is why, a few years in, one person ends up holding a structure of libraries, books and pages that nobody else fully understands, and why a single vendor cost increase turns into two days of manual work and hoping nothing gets missed. It's a structure problem, and structure is fixable.

Here are three of the six patterns we walked through live, and what to check in your own system.

Last month's decisions have a way of showing up in this month's numbers. Dead stock bought during a busy stretch ties up capital that lands on the next balance sheet. Orders shipped at outdated prices, while a vendor's cost increase was still working its way through the system, landed in the following period's revenue at a margin nobody flagged at the moment.

Your P&L does not lie about this. It also does not explain it. It shows the number, not the decision that drove it. That is the margin reporting gap.