The toolroom supervisor's guide to stockout-free weeks

The Monday-morning pattern every NZ toolroom knows

You walk in at 6:30am. The pot list for the week is on the whiteboard. Somewhere around 7:15am an operator says the words that define the next three hours of your day: "we're out of the 10 mm three-flute." You phone the supplier, they're out too, you phone the next one, a parts-held-up email goes to production. By 10am the machine's running but you've lost the morning.

This post is about how to stop that from being a weekly event. Not with heroics. With a few operational defaults tuned against the data your shop already keeps.

Stockouts have three causes — name yours

There are really only three patterns underneath most toolroom stockouts:

1. Your thresholds are wrong. The minimum on that SKU was set in 2019; demand has changed; nothing triggered a reorder because the min is stale.
2. Your lead time is wrong. The supplier used to be 3 days and is now 11. Your reorder point still assumes 3.
3. Your reorder cadence is wrong. The PO got raised, sat in DRAFT for a week, then got APPROVED on Friday afternoon — which meant a Monday delivery window that doesn't exist.

Each of these has a different fix. Lump them together and you'll chase the wrong one. Track them separately and the pattern resolves in a month.

Playbook — the five things that matter

1. Review minimums against rolling 90-day demand, not memory

The most common anti-pattern: a shop sets minimums once, when the ERP goes in, and never revisits. Demand mix moves under them. Niyamis TMS's crib view shows the 90-day rolling take-rate on every SKU next to the configured minimum; the gap between those two numbers is where stockouts breed.

Cadence: quarterly review. An hour per location, once per quarter, pays for itself in the first week.

2. Record lead time from the supplier's promise, not the PO date

Track two dates on every receive event: when the PO was sent, and when the items actually arrived. The difference is the supplier's real lead time, which you should feed back into the reorder point. Suppliers drift. Measure the drift.

TMS does this automatically. Every RECEIVED PO timestamp updates the rolling supplier lead-time estimate. If your primary vendor slows from 3 days to 7 over six months, the reorder points for that vendor's SKUs shift with it.

3. Kill the Friday-afternoon PO-approval bottleneck

The single biggest avoidable failure mode we've seen on NZ floors: POs sitting in DRAFT or APPROVED state over a weekend. Nothing moves on Saturday. Nothing ships on Sunday. Monday starts already two days backwards.

Fix: automate PO approvals below a threshold (say, $500) for preferred vendors. Every approval above that threshold gets an escalation email if it sits for more than four business hours. TMS's PO workflow has this built in; other systems support it with a bit of scripting.

4. Build a "mission-critical" watchlist

Not every SKU deserves equal attention. Identify the twenty SKUs that, if they stock out, halt production within one shift. Those get:

• Tighter minimums (30% buffer instead of 10%).
• Dual-sourced wherever possible — primary plus backup vendor, so one supplier's late delivery doesn't become your stockout.
• Daily check by the opening shift supervisor, not weekly by the crib.

5. Treat scrap alerts as demand signals, not just quality signals

A spike in scrap on a specific SKU is about to become a spike in demand. The tool that's failing will get replaced. Your reorder math assumes normal-wear replacement; abnormal-wear replacement comes faster.

When TMS flags a scrap anomaly on a SKU, it also bumps the reorder forecast for that SKU until the anomaly clears. Supervisors see both the quality issue and the supply consequence in one alert.

A week in the life of a TMS-running toolroom

Monday 6:30am: open the alert inbox. Three alerts.

• One is a routine reorder-point trigger for a common end-mill. Click approve. PO goes out. Done in under a minute.
• One is a scrap anomaly on a 12 mm drill — three scraps on the same machine in the last 48 hours. You walk over, check the machine, find a coolant issue. Real problem, real fix, before the machine produced its fourth scrapped part.
• One is an amber confidence on a life-prediction for a medical-part drill that's been in service four weeks. TMS thinks it's two shifts from end-of-life. You stage the replacement.

Twenty-five minutes total. The rest of the day is actual toolroom work.

What happens when you do this

Shops that adopt this playbook in the first six weeks of a Niyamis TMS engagement typically see:

• Stockouts per month drop by 60–80% (subject to shop volume).
• Average PO cycle time from DRAFT to APPROVED drop from days to hours.
• The supervisor's Monday-morning reactive time drop from 3 hours to under 30 minutes.

Numbers are ranges because every shop is different. The shape of the improvement is remarkably consistent.

Want this for your shop?

Book a thirty-minute call with Philippe. We'll look at your crib data and tell you which of the three stockout causes is biggest on your floor.

— Philippe, Niyamis