Dispatch October 27, 2026

Running Out Mid-Job: Reorders That Remind Themselves

You're halfway through the job, hands full, and you reach into the bin for the part you were sure you had three of. There's one. Then there's none.

Both options now cost you. You can stop the job, drive to the nearest supply house, pay retail for the thing your standing order gets cheaper, and lose the back half of the afternoon. Or you can improvise and hope the customer doesn't clock the corner you cut.

Running out mid-job is the kind of small disaster that never shows up as its own line on the books, so it's easy to pretend it isn't costing you. It is. A couple of emergency runs a month, retail markup on every one, plus the jobs that slid a day because you were standing in an aisle instead of finishing. Nobody budgets for it, and that's exactly why it keeps happening.

What running dry really costs

Part cost is the small one here. Say a fitting runs you four dollars on your supplier account. The same fitting at the counter down the road is seven, and you're buying one instead of a box, so the per-piece price stings worse. Annoying, not painful.

Afternoons are where it actually hurts. An emergency run is rarely fifteen minutes. It's drive time, the wait at the will-call desk, the drive back, and the twenty minutes it takes to get your head back into the work. Figure most of an hour gone, sometimes two, and that's an hour you can't bill or a job that now bleeds into tomorrow.

Quietest of all is the trust cost. The customer watched you pack up and leave in the middle of their job. Maybe you made do with the wrong size because a second trip wasn't an option. Neither one earns you the next call.

The one thing the robot can't do

Here's the honest limit, and it's a big one. A robot can't count what you don't track. If the only record of how many filters you've got is a rough picture in your head, no automation on earth can watch it for you. Software reads numbers, not shelves.

So the whole idea rests on one lightweight habit: a single running count of the handful of things that stop a job cold when they hit zero. Not every screw and zip tie. Just the stuff that sends you racing to the supply house.

That count can live anywhere you'll actually keep it current. A shared note, a plain spreadsheet, the inventory field in the job software you already use. The tool matters far less than whether the number stays close to true. If the count drifts, the reminders drift with it, and you're back to guessing.

How a low-stock trigger works

Once there's a number to watch, the automation part is simple. You set a floor for each item, a point low enough that you've got time to reorder but not so low you're cutting it close. When the count drops to that floor, a reminder fires, or better, a draft purchase order shows up ready for you to look over.

In plain terms, it runs like this:

  • You set the floor. For each item you're watching, pick the count where a reminder should fire. Enough runway to reorder, plus a little buffer for a busy week.
  • The count ticks down as you work. Pull three filters for a job, the number drops by three. This is the part that leans on a human habit, which we'll come back to.
  • The trigger watches the number. When it touches your floor, the automation notices on its own. No calendar, no memory, no sticky note on the dash.
  • A draft order lands in your lap. The good version doesn't just ping you, it drafts the purchase order: right supplier, right quantity, ready to send.
  • You approve it, or you don't. One look, one tap. Slow month or switching suppliers, you hold off.

Nothing gets ordered and no money moves until you say so. The robot does the watching and the typing. You do the deciding.

Watch the few things that stop a job

The temptation is to track everything in the van. Resist it. A count you can't keep current is worse than no count, because it fires reminders you quickly learn to ignore. Pick the short list that genuinely wrecks an afternoon, and leave the rest alone.

Run each candidate through a few questions:

  • Does running out actually stop the job, or just slow it down a little?
  • Is it slow or expensive to replace in a hurry, like a special-order part or something priced ugly at retail?
  • Do you go through it often enough that the count actually moves?
  • Is the lead time long enough that a panic order the morning of won't save you?

Most trades land on a surprisingly short list. A plumber might watch a few common fittings and a water heater or two. A cleaning crew watches the supplies you can't grab at the gas station on the way. A clinic watches the one or two consumables that cancel appointments when they run dry. Five to fifteen items, usually, not five hundred.

Keep your hand on the order

One more guardrail, and it's the one that matters most: you okay the order and the spend, every single time. This isn't the robot being timid. It's the split that keeps the whole thing safe, machines do the relay, you do the judgement.

Prices move, suppliers run into backorders, and sometimes you know something the bare count doesn't. A big job next week means you should order double. A slow stretch means you'd rather not tie up the cash. A draft you approve respects all of that; an order that fires on its own respects none of it.

At TellRobots, that approval step is the first thing we wire in, not the last. A reorder tool you can't trust with your money isn't worth the afternoon it's meant to save.

Worth doing this week

None of this needs a big system to get going. You can start on a notepad and tighten it up later.

  1. Write down the last two or three times you ran out mid-job. That's your real list, not a guess.
  2. Pick five to ten items worth watching using the questions above. Start small; you can always add more once it's working.
  3. Start one running count somewhere you'll keep current: a shared note, a spreadsheet, or the job software that's already open.
  4. Set a floor for each item, a number that leaves you a few days of runway before zero.
  5. Wire up one reminder or draft order for your most painful item, and keep the approval in your hands.

Get one item watching itself this week. Once you've approved a reorder from the truck instead of losing an afternoon to the supply house, you'll know exactly which ones to add next. And the shelf stops surprising you.

More dispatches →