The message format that makes every wholesale order come in the same way

The format that makes a wholesale order come in the same way every time is a five-line message you ask the customer for: the word ORDER, their account number, the delivery date, one line per product with the structure code x quantity unit, and the word END followed by how many lines the order has. It is a writing agreement, not software: you announce it today and it works on Monday. What makes it work are three details almost nobody includes: that the unit is mandatory on every line, because that is where the money is lost and not in the quantity; that the number written after END works as a check that no line is missing; and that you are the one who sends the template already filled in with what they bought last time.

The problem with taking orders over WhatsApp in a wholesale distributor is not the volume of messages. It is that every customer invented their own language: one sends a photo of a handwritten note, another a two-minute voice note, another writes "send me 10 of the litre-and-a-half one and the usual on biscuits". All three end up with the same person trying to guess, and guessing is what gets paid for later.

The format: five lines and one word at the end

This is the message you ask the customer for:

ORDER
Account: 1042 - Autoservicio San Martín
Delivery: Thursday Aug 7
1104 x 6 BOX
2210 x 2 BULK
3087 x 12 UNIT
END 3

What each line does:

• ORDER on the first line. It is what separates an order from the other forty things that come into that number: price enquiries, complaints, photos of delivery notes, greetings. It is also what lets you search for them in the chat when somebody disputes what they ordered.
• Account. The WhatsApp writing to you is almost never the account holder: it is the manager, the stocker, the son, the owner's new phone. Without an account number the order gets entered under the contact's name and not the business, and that comes back a month later as an argument over the open account.
• Delivery with an explicit day. "For tomorrow" sent at 11:40 p.m. is ambiguous, and "urgent" is not a date.
• One line per product, always with the code first. Never two products on the same line.
• END plus the number of lines.

If the customer needs to clarify something, a single NOTE line goes after END. And the hard rule: a note cannot contain quantities. "NOTE: add 2 boxes of 1104" is not a note, it is a line, and if it comes in as a note it gets lost.

Asking for codes forces you to hand them out, and that is where half the attempts fail

The code goes first because commercial names have variants the customer does not distinguish and that in your system are different products with different prices. But asking for codes has a flip side most people skip: the customer has to have them at hand.

Do not send the 600-item catalogue. Send each customer the list of the ones they bought in the last three months, which in a small grocery store is usually between 20 and 40. And if your system has duplicate codes or codes that changed last year, that gets sorted out before you ask the customer for anything: you cannot demand precision about data you keep messily yourself.

The unit is where the money is lost, not the quantity

"10 of the litre-and-a-half one" can be 10 bottles, 10 six-packs or 10 boxes. All three are possible orders from the same business and the difference between the first and the third is an order of magnitude.

That is why the unit goes on every line, in capitals, and from a closed vocabulary of four words that you define and that are not up for negotiation: UNIT, PACK, BOX, BULK. Four, not seven. No "sack", "trolley" or "the big one".

And the rule that holds all this up: a line without a unit does not get entered, it gets asked about. Entering it by assuming the most likely unit is what comes back on the truck later. The cost of that mistake is not the three minutes of the person who entered it: it is a delivery run you paid for twice on a line whose margin does not cover even half that trip.

END plus the number of lines prevents the order that arrives in four messages

The pattern is familiar: the order comes in at 4:10 p.m., at 4:40 p.m. "oh, and two of 2210" arrives, and at 6:05 p.m. a third message shows up when you have already invoiced.

END with the total number of lines tells you two things instantly. If you count fewer lines than the number declares, the order is incomplete and you ask for it again before touching it. And everything that arrives after END stops being an ambiguous continuation: it is a modification.

That treatment is worth spelling out from day one: it is the hole the mess comes back through. An addition is requested in a separate message that starts with CHANGE, the account and the complete line with its unit, and it does not exist until you confirm it in writing. And the cut-off for modifications is earlier than the one for new orders: a modification touches something that is already picked.

Without a cut-off time, the perfect format is useless

The format sets what comes in. The cut-off time defines until when. A flawless order that comes in at 7:40 p.m. when the route sheet closed at 7 p.m. is as useless as a two-minute voice note.

The cut-off goes in the same message where you announce the format, and it goes repeated in every confirmation: "received, it goes out on Thursday's route; anything arriving after 6 p.m. moves to the next route". Repeating it in every confirmation is what makes it stick; announcing it once is not enough.

The exact message to announce the change

Ready to copy and adapt:

Hi [Name]. From Monday we are changing how we take orders
so we can confirm faster and without mistakes.

Copy this message, fill it in and send it to this same number:

ORDER
Account: [your account number]
Delivery: [day]
[code] x [quantity] [UNIT / PACK / BOX / BULK]
END [number of lines]

Below is your list of usual codes.
Orders that arrive this way get confirmed with stock
and price the same day. Cut-off for next-day
delivery: 6 p.m.

Two things about how to send it. First: do not announce it to 200 customers on the same day. Start with the 20 who place the most orders, who are the ones who save you the most time and the ones least likely to argue about the change.

Second, and this is the real adoption lever: send each of them the template already filled in with their previous purchase, not blank. Nobody writes twelve lines from scratch. With the previous order in front of them, the customer deletes two lines, changes three quantities and sends it in forty seconds.

Part of your customer base will keep sending voice notes

Of every ten customers you send the template already filled in with their previous purchase, what we usually see is that six or seven use it within the month. The other three or four carry on with voice notes, photos of handwritten lists and "the usual", and they do not move through insistence. That proportion is the shape the curve usually has, not a promise about your customer base: you will read yours in week four, counting how many orders carried END.

Do not fight that remainder. The play is the other way round: when a voice note comes in, you listen to it once and reply with the complete template based on what you understood, asking for an OK or a correction. You turn the voice note into an order confirmed in writing, which is exactly what you need when somebody disputes two boxes. And along the way, a good share of those customers starts using the template simply because they discover that replying OK is faster than recording.

Time ten orders tomorrow and multiply by 22

That is all the measuring work needed here, and it is done with the stopwatch on your phone. In small-scale distribution the usual range is between 4 and 7 minutes per free-text order, counting reading it, deciphering it, looking up the code in the system, asking over WhatsApp what they meant and waiting for the answer. An order in the format takes between 1 and 2 minutes.

Then, one line:

(minutes it takes you today − 2) × orders per day × 22 = minutes per month.

With 35 orders a day and 5 minutes on average: (5 − 2) × 35 × 22 = 2,310 minutes, about 38 hours a month.

That number is the ceiling, not what you are going to save: it holds if 100% of the orders come in the format, and that is not going to happen. Multiply it by the share of your customer base that actually adopts it —at 70%, the 38 hours come down to 27— and only then compare it against what that person costs you. Even penalised like that, the maths still comes out at more than half a working week a month spent interpreting messages.

What the maths does not capture, and it is the most expensive part, is the order entered wrong. A wrong line gets paid for in redelivery, credit notes and goods coming back, and that does not show up on any stopwatch.

Where the format stops being enough

The format solves intake. It solves nothing of what comes afterwards, which is most of the work.

A perfectly formatted order still does not tell you whether there is stock for the three lines, what that particular customer's list price is, whether they are up to date on their account, whether that bulk unit has a minimum of two, or whether they ordered 6 boxes when their average over the last six months is 1 and they almost certainly mistyped. All of that is still done by a person looking at the system and answering by hand.

The three points where it breaks:

1. Confirmation with stock and price. The customer wants to know within minutes what is coming and how much it costs. Doing it by hand means opening the management system order by order. Up to about 15 orders a day it is perfectly viable. Above 25 you start confirming nine o'clock orders at three in the afternoon, and there you have lost the advantage the format gave you.
2. Out-of-stock items and substitutions. When 3 out of 12 lines have no stock, a back and forth starts that easily runs to ten or fifteen messages per order. It is where the time you gained evaporates, and no message format prevents it: what is needed is for the system to propose the replacement using the criteria you defined.
3. Price lists and volume discounts. With three lists and four tiers, the person entering the order has to remember which one applies in each case. These are the expensive mistakes, the ones only discovered in the month-end reconciliation.

The criterion, rounded off: with fewer than 15 orders a day and a single price list, the format plus one person entering orders is enough, and automating is money spent. Between 15 and 40 orders a day it is worth automating the confirmation —stock, price and order number returned in the same thread— and leaving data entry and exceptions in a person's hands. Above 40 a day, or with several price lists, or with delivery zones that have different cut-offs, the bottleneck stops being understanding the order and becomes validating it. That is no longer fixed by writing better.

The two-column week that settles all of this

The stopwatch above tells you what free text costs you today. This other one tells you whether that gets fixed by buying something, and it also takes a week: next to each order write down two times, the one it came in and the one you confirmed it with stock and price. Nothing else, in a notebook or in the last column of the spreadsheet you already use. By Friday you have twenty-five or thirty pairs of times.

If the distance between the two times stays even from Monday to Friday, your problem was order intake and the format above solves it entirely: there is nothing to automate yet. If the gap grows over the course of the day —the morning ones confirmed in twenty minutes, the five o'clock ones confirmed the next day—, the format has already done its part and what filled up is the validation queue: stock, the customer's price list, minimums per bulk unit, the open account. There what gets automated is the confirmation, not the intake, and that distinction is the one usually bought wrong.

And there is a task before all of this that cannot be skipped, because if it goes wrong neither of the two measurements is any use: open your catalogue and check whether the codes are stable. Duplicates of the same item, last year's renumbering living alongside the new one, three formats of the same product with the same code. Nothing gets automated on top of a catalogue like that, because the perfect format will hand you back a code your system does not recognise, and you are back where you started with one more step.

Before hiring anyone, then: move the cut-off time and send twenty pre-filled templates. That gives you back more time this week than any development starting next month, and it costs nothing. If after that the gap keeps growing in the afternoon, write to us at info@striqtech.com with the worst pair of times from the week, how many price lists you handle and which management system you use.