How to stop your team retyping email orders into the system

Walk into the office of almost any wholesaler, distributor or trade supplier at 9am and you'll see the same scene: someone with two screens, reading an email on one and typing it into the system on the other. Order by order. All day.

Nobody planned this. When the business was doing ten orders a day it wasn't worth fixing. Now it's doing a hundred and fifty, and the fix never happened — the office just grew another desk at a time.

What retyping actually costs

Do the arithmetic for your own operation, because it's rarely done and always surprising:

• A straightforward emailed order takes 3–5 minutes to key in: open it, find the account, match the products, check the quantities, enter it, file the email.
• At 80 orders a day, that's 5–6 hours of someone's day — most of a full-time job doing nothing but moving information from one screen to another.
• Then add the errors. A mistyped quantity or a wrong product code doesn't cost minutes; it costs a wrong delivery, a credit note, a return journey and an annoyed customer. Even a 1-in-50 error rate at that volume is a problem every single day.

And there's a cost that never shows up in a spreadsheet: the person who does this is the only one who knows how. When they're on holiday, orders slow down. That's not a staffing issue — that's a process held together by one person's memory.

Why this never got fixed

Three reasons, and none of them are your fault.

The old technology was genuinely bad. OCR and "document capture" tools from the 2010s needed every supplier to send the same neat layout. Real orders don't look like that — they arrive as a forwarded email chain, a photo of a handwritten list, a PDF from someone else's system, three lines in a WhatsApp message. The old tools choked, someone spent more time fixing them than typing, and the project died. If you tried this five years ago and gave up: you were right to.

The "proper" fix asks your customers to change. ERP vendors will happily sell you a customer ordering portal. The problem is that your customers like sending emails. A tyre fitter ordering at 7:30am isn't going to log into your portal — he's going to text the same message he's sent for ten years. Any fix that depends on hundreds of customers changing their habits is not a fix.

It grew slowly enough that nobody noticed. Nobody signs off "let's spend six hours a day retyping." It just accumulates, one hire at a time.

What works now

The technology changed in the last couple of years, and it changed in exactly the way this problem needed: modern AI systems read documents the way a person does. Layout doesn't matter. Format doesn't matter. A forwarded email with the order buried at the bottom of a chain — fine.

A working setup looks like this:

1. Orders land where they already land — your existing sales inbox. Customers change nothing.
2. The system reads each one and extracts the account, products, quantities, delivery details — matched against your product list and your account names, not a generic guess.
3. A person checks instead of types. The order appears pre-filled next to the original email. Your team member glances, corrects anything odd, and confirms. Thirty seconds instead of four minutes.
4. Anything unclear gets flagged, not guessed. Unknown product, ambiguous quantity, new customer — those route to a human with a note saying exactly what's uncertain. This is the part that makes it trustworthy: a system that knows what it doesn't know.

The person doesn't disappear from the process — they move from typist to checker. That's the difference between this and the automation horror stories: nothing enters your system without human eyes on it, but the eyes now do fifty orders in the time typing did five.

What this doesn't fix

Honesty matters more than a sale, so: order entry automation will not fix pricing disputes, credit-hold decisions, or a product catalogue that's a mess in the system. If your product codes are chaos, fix that first — the best reading system in the world can't match against a list that doesn't make sense. And if you're doing fifteen orders a day, the arithmetic probably doesn't justify it yet; a good part-timer is the right answer at that volume.

How to tell if you're ready

Three questions:

• Volume: are you keying in more than ~30 orders a day? Below that, the payback is slow. Above it, the payback is quick and gets quicker as you grow.
• One system of record: do orders end up in one place (Sage, Xero, an ERP, even a well-run spreadsheet)? If yes, there's something concrete to connect to.
• A known bottleneck: does everyone in the office already know who the "order typing" person is? If the answer is instant, that's your sign.

What to do this week

Before talking to anyone (including us): measure it. Count one day's orders that arrive by email, message or PDF. Time three of them being keyed in, honestly — including finding the account and filing the email. Multiply. Then multiply by 250 working days.

That number is what the status quo costs per year. Whatever you do next — build something, buy something, or decide it's not worth fixing yet — you'll be deciding with the number in front of you instead of a feeling.

And if the number shocks you, that's usually the moment owners get in touch.