Order Management

You promised a date. What did you base it on?

In this article

Nori.mx6 min read

Orders come in byEmailEDI · portalPhoneWhatsAppSales repTHIS WEEKMTWTFTue: you promisedThu: their line stopsWhat was it based on?Inventorywhen is it committed?On confirmingAt picking

The customer called because their line goes down Thursday and your shipment never arrived. You promised Tuesday. Between those two sentences sits a week of work by several people, and odds are not one of them can tell you exactly where it broke.

It did not break at the end. It broke the day somebody said “yes, Tuesday” without anything to back it up.

How orders actually reach you

Do this on paper: write down every path an order took to reach you last month. At a company selling into maquiladoras, five usually come out.

The buyer’s email. EDI or the customer portal, where releases get published and nobody tells you. The phone, which ends as a note on a desk. WhatsApp, which ends on one person’s phone. And the rep, who carries the order in his head out of the meeting.

Five paths, five separate inboxes, none of them talking to the others. The number is not the point. The point is that there is no single queue. Nobody can open one screen and see everything committed and not yet shipped. If that screen does not exist, the delivery date is not calculated. It is guessed.

The promise lives in somebody’s head

Ask who owns the date. At a company this size the answer is usually a person’s name, and usually the one who has been there longest.

That person is good at the job. They know the tooling for that part runs long, that this customer always asks to pull in, that there are no carriers in November. All of that is real knowledge and it is worth money. The problem is not that it exists. It is that it is written down nowhere, so the company cannot promise a date without that person. If they are out sick, if they quit, if they are in a meeting when the rush order lands, the date gets promised by somebody who does not know.

It is the same exposure faced by anyone who inherited software nobody documented, except here what is undocumented is commercial judgment.

The exact moment it breaks

Here is the most common cause, and it is almost never the one people chase.

Does inventory get committed when the order is confirmed, or later?

If it gets committed later — at production scheduling, at shipment prep, at picking — then between confirmation and that moment the same piece is still available to every other order. Two reps promise the same box. Both are right according to what the screen showed them. One of those two customers finds out on Thursday.

When this happens it does not look like an order problem. It looks like an inventory problem: “the system says we have it and we don’t.” So the warehouse gets the blame. But the warehouse did nothing wrong. It gave what it had to whoever got there first. What failed is that nothing was ever set aside.

If your counts do not match the screen, the problem may sit one step earlier. We covered that in when a maquiladora needs a WMS.

The order that gets keyed in three times

Follow one order from arrival to invoice and count how many times the same information gets typed.

Three is normal. Once when it arrives and gets moved into your own format. Once when it goes into the system to be produced or picked. Once to invoice. Sometimes a fourth for the shipping notice.

Every re-key is a chance for a quantity, a part number or a date to change. And because each copy lives somewhere different, when they disagree nobody knows which one governs. The question that exposes this: when the customer changes the schedule, how many places have to be touched for the change to stick? If the answer is more than one, sooner or later one of them does not get touched.

Four numbers that measure this without buying anything

These four can be pulled in a week with what you already have, and they are worth more than any demo:

  • Out of ten orders, how many ship complete and on the promised date. If you do not know, that is the first finding.
  • How many people touch an order between arrival and confirmation. More than three means the path is not defined, only habitual.
  • How many hours pass between an order arriving and somebody putting a date on it. Slow answers cost you orders you will never know you lost.
  • How many orders changed after being confirmed, and why. If most of the changes are yours rather than the customer’s, the problem is the promise, not execution.

What has to be agreed before you buy anything

An order management system makes a rule enforce itself. It cannot invent the rule. Three of them have to exist first, on paper, with names attached:

Which order gets the inventory when there is not enough. First in, largest customer, the one with a line-down penalty. All three are defensible. None is automatic. If it is not written down, every shortage gets settled by argument and the loudest voice wins.

Who owns the date and what they calculate it from. Not the person’s name: the criteria. What gets checked before committing — stock, capacity, supplier lead time, carrier — and what buffer is left.

What happens when the customer changes the schedule. Who finds out, how fast, and who decides which orders get pushed.

If those three are not settled, an OMS will automate the argument instead of preventing it. It is the same sequencing we described for companies with several fronts open at once: process first, then system.

What a system actually changes

With the rules written down, an order management system does three things that do not hold up by hand.

It pulls all five paths into one queue, so the order that came in over WhatsApp sits next to the one that came in over EDI. It commits inventory at the instant of confirmation, which makes the double promise impossible. And it calculates the date against what is actually on hand and what actually fits, not against somebody’s memory.

The rest follows: your on-time number stops being something anyone has to assemble by hand, because the system already knows it.

And if a customer is pushing you onto EDI, these are the same subject. EDI is one of the five paths, and connecting it before the queue is in order only makes the mismatch travel faster.

Where to start this week

Do not get a quote yet. Take the last ten orders that shipped late and, for each one, write down where the time went. Ten cases is enough: eight of them usually break in the same place.

If that place is confirmation, it is a promise problem. If it is the warehouse, it is an inventory problem. If it is the supplier, it is not yours. Three different roads, and confusing them is expensive.

A side note that tends to surprise people: a good share of the calls your team takes are not orders, they are people asking whether it shipped. Once the queue genuinely exists, that answer can give itself — and that is where a customer portal pays for itself.

If you want a read on your own case before talking to anyone, Chira asks these same questions and gives you back, in writing, what to fix first.

Related service

Order Management

Every sales channel, one order flow without friction.

See the service

Articles on this service

Order 4471Deliveredon FridayPaid by cardRefund:same card5 business days to withdrawfrom deliveryFSSM1T2W3T4F5If it arrived wrong2 monthsto claimYou respond in15 daysat most

Order Management

Your customer has five days to change their mind

After Buen Fin come the returns. The deadlines Mexico's consumer law sets to withdraw, claim and refund, and what your order system needs to record.

6 min read

Online store55" TV · $8,999Marketplace55" TV · $8,999Counter55" TV · $8,999In stock1sold 3 timesNOVEMBER13 — 17Buen FinArt. 43: if it says"available," you have it

Order Management

Buen Fin: you sold online what you didn't have

Your online store, the marketplace and the counter all sell from the same shelf. What Mexican law says about listing items as available, and what to fix by Friday.

5 min read