Skip to content
allanninal.dev/build
Collaborate

Part 5 of 7 · Supplier lead time watcher series ~5 min read

How it turns into a conversation

Everything up to here is measurement, and measurement is the easy half. What a business does with an accurate lead time number determines whether the supplier relationship gets better or gets formal, and those are very different outcomes.

Key takeaways

  • Nothing about a supplier changes automatically. The system produces evidence, not decisions.
  • Show your own internal delay first and the conversation changes shape.
  • A scorecard with a grade produces defensiveness; a distribution produces explanations.
  • Ask what changed before asserting that something did.
  • The best output is often a changed reorder point, not a changed supplier.

Nothing is automatic

The three supplier actions this system deliberately does not automateA horizontal row of five boxes. Drift detected, with evidence. No automatic reorder point change: a person decides. No automatic supplier switch, obviously. No rating: no A to F grade. A conversation, with numbers in it. A note says three refusals and one output, and the refusals are what make the output credible.WHAT THE SYSTEM DOES NOT DODrift detectedwith evidenceNo auto-reorder changea person decidesNo auto-switchobviouslyNo ratingno A to F gradeA conversationwith numbers in itThree refusals and one output. The refusals are what make the output credible.
Fig 1. The deliberate limits. Each of the three middle boxes is a feature somebody will request and each would make the system worse.
  • Security & identity
  • Analytics
  • People

Why not change the reorder point automatically

It looks like the obvious win and it is the one to resist hardest, because a reorder point that moves on its own means the amount of cash tied up in stock moves on its own. A quiet three-day increase across forty products is a significant working capital change that nobody approved.

The system proposes it — “this should move from 8 days to 13” — and somebody accepts it in one click. That is nearly as fast and it leaves a decision with a name on it.

Show your own delay first

A thirteen day lead time shown as one number and as its two partsA bar chart with two bars measured in days. Two series: our approval queue in orange, and the supplier's sent-to-received period in purple. The first bar, as usually presented, shows thirteen days all attributed to the supplier. The second bar, as measured, shows four days of our approval queue and nine days of supplier time. A note says the left bar is the number that gets quoted in meetings, and it is not true.05101520~13As usually presented~13As measuredOur approval queue, daysSupplier, sent to receivedThe left bar is the number that gets quoted in meetings. It is not true.
Fig 2. The same thirteen days, presented two ways. Walking into a supplier review with the left bar is how a review becomes an argument about facts.

Opening a supplier review with “our approval process is adding four days and we are fixing that” does two things. It establishes that the numbers are honest, which makes the rest of them harder to dispute, and it changes the register of the meeting from complaint to shared problem.

It is also frequently the more valuable finding. Four days of internal queue on every order is entirely within your control and is often larger than the supplier drift that prompted the analysis.

The scorecard problem

Supplier scorecards with letter grades are common and they reliably produce the same response: the supplier optimises for the grade. Lead time is easy to improve by quoting longer and delivering to the quote, which improves the score and makes the actual situation slightly worse.

A distribution does not have that property, because there is no target to hit. “Your typical is nine days and your worst quarter was thirteen” is a description, and the natural response to a description is an explanation rather than a countermeasure.

Ask what changed

How the conversation opens

  • “Our numbers show your lead times moved from about six days to about eleven over this year.” State it, with the count of orders behind it.
  • “Four days of the total is our approval queue, and we are dealing with that.” Before anything else.
  • “Has something changed at your end?” A question, and a genuine one.
  • Then listen. The answer is frequently a specific thing — a supplier of theirs, a machine, a person who left — and frequently temporary.
  • Agree what to plan for, not what to promise. A supplier who says eleven days honestly is more useful than one who says six and means it.

That last point is the one worth carrying away. The goal of the conversation is not to get the lead time back down; it is to find out what number to plan against. A supplier who tells you the truth about eleven days lets you set a reorder point that works, which is worth more than a promise of six that fails a third of the time.

The annual review this makes possible

Once a couple of years of this exists, the yearly supplier conversation becomes a different thing: here is every order, here is the distribution, here is how it moved, here is what we plan against. Most businesses cannot have that conversation because nobody wrote the dates down.

It also makes the occasional decision to change supplier defensible, which is the only context in which that decision should ever be made from this data — slowly, with two years of evidence, by a person who knows what else the relationship is worth.

Next: what all of this costs to run.

All posts