Task automation in a print shop: the ten minutes you never get back

Automation does not save you an hour. It saves you ten minutes, forty times a week, from the people who should be selling. That is the whole argument, and it is a better one.

Automation gets sold as transformation. In a print shop it is almost never transformation. It is ten minutes, removed from something that happens forty times a week, done by somebody who should have been on the phone to a client.

That is a smaller story and a much more reliable one, because you can actually count it.

How to rank what to do first

One calculation, and it settles most arguments:

• Minutes the task takes, honestly measured rather than remembered
• Times it happens in a week
• Multiplied, then priced at the loaded rate of the person doing it
• Minus how long it would take to automate, amortised over a year

Do that for ten tasks and the order becomes obvious. It will also surprise you. The task everyone complains about is usually not the one costing the most, because the expensive ones are the small ones nobody notices.

The tasks worth automating in almost every shop

• Order acknowledgement. Every order, every time, immediately. It removes a whole category of inbound question.
• Artwork checking at upload. Format, resolution, colour mode, transparency, checked before a human ever opens the file.
• Proof reminders. The client who has not approved is the reason the job is late. Chase automatically, on a schedule, and stop doing it by memory.
• Status updates. Pushed to the client when the status changes, not written when they ask.
• Reorder prompts. The client who bought polos in March is a real prospect in March next year, and almost nobody contacts them.
• Order to invoice. Generated from the order that shipped, not retyped.
• Supplier purchase orders. Built from the order rather than assembled by hand.
• Low stock and backorder alerts. Known before the client asks, not after.

The three that should stay human

• Pricing anything custom. Rules can calculate. People decide.
• Anything after a complaint. The moment a client is unhappy, automation makes it worse. Route it to a person immediately, with the full history attached.
• The first conversation with a new client. You are establishing whether you can work together. Nothing about that is a repeatable task.

Fix the process first, always

Automating a process nobody has written down does not remove the confusion, it scales it. If two people in your shop do the same task differently, automating it means picking one of them and making everyone else wrong, which is a decision worth making deliberately rather than by accident.

The cheap version of this: before automating anything, write the steps on one page. Most of the time that page removes two of the steps on its own, and you automate a shorter task.

How to know it worked

• Hours returned per week, per task, measured before and after
• Rework: how often the automated task has to be redone by a human
• Inbound questions: the right automations make the phone ring less
• What the freed hours were actually spent on, which is the only one that turns into revenue

That last measure is the one people skip, and it is the one that decides whether automation made you money or just made you less busy.