AI automation for your print shop: where it pays off

AI automation for a print shop is not about your press. That is where your craft sits, that is what you invested in, and no clever algorithm needs to touch it. It is about the layer in front of it: the inbox where quote requests arrive with half the specifications, the customer calling to ask whether their print job has shipped, and the reorder that says 'same as last year, but 500 this time'. That work earns you nothing extra, but it does cost the hours of exactly the people you need most.

A shrinking market with 25 to 30 percent overcapacity

Every year the Dutch graphic media industry fund has research agency GOC calculate the sector's numbers. The December 2025 report contains a sentence you probably already feel: despite years of decline in the number of companies and in employment, the industry is still stuck with overcapacity of between 25 and 30 percent. In other words, there is more press capacity in the country than there is work.

That has one direct consequence for your quotes: price competition. The same report shows the industry splits cleanly in two, half the companies are shrinking in revenue, the other half is still growing. The difference between those halves is rarely the machine. It is who wins the work, and how smoothly that goes.

The average print shop is smaller than it was ten years ago

On 1 January 2025 the Netherlands counted 2,620 graphic companies, 21 percent fewer than ten years earlier. It is mainly the larger firms that disappear: the 20 to 49 employee class halved in ten years and the 50 to 99 class shrank by 43 percent. Average company size dropped from 9.5 people in 2010 to 7.3 in 2024 (figures: Statistics Netherlands, processed by GOC, December 2025).

In the band this article is about, 5 to 50 people, roughly 380 companies remain. Those are the printers too big to just do everything themselves and too small for an in-house automation department. What happens in practice is simple: the same amount of customer contact gets spread over fewer people. You will not notice it at the press. Everyone notices it in the inbox.

Your press is current, your front office is not

In 2024 the industry invested 107 million euros in fixed assets. That money went mainly to printing presses, printing equipment and finishing. About software the same report has one dry line: investment in computer hardware and software happens at a far lower amount. And of the companies that do have innovation plans for 2026 and 2027, a little over half, nearly one in five does not yet know what those plans should be.

That is not a reproach, it is logical. A press standing still costs money immediately and you see it the same day. A jammed inbox costs money that never shows up on an invoice: the request you answered too late, the quote nobody followed up, the customer who had it printed elsewhere this time. You only miss it once you add it up.

Where the hours really go: the quote request

Ask a printer about their busiest work and you rarely hear about the press. You hear about this morning's email: '500 flyers, what does that cost?' No format, no paper stock, no weight, no word on single or double sided, no finishing, no delivery date, and no clue whether the customer supplies print-ready files or still needs design work.

• It takes two or three messages before anyone can even start calculating.
• Your estimator, often your most experienced employee, prices requests you already know will mostly never become orders.
• The customer who gets a complete answer first usually wins the order. Not the cheapest one, the first one.
• Quotes stay open because nobody has time to chase them.
• 'Same as last year' means someone goes digging through the archive to find out what last year actually was.
• Questions about file delivery, bleed, resolution, colour space, converting fonts, get explained for the hundredth time.

None of that is hard work. It is work that arrives in fragments, always in between other things, and always to someone who had something better to do at that moment.

What AI automation actually does in a print shop

The layer you automate sits between your inbox and your estimating software. Not inside it, and not after it. In concrete terms it looks like this:

• Completing requests. An assistant reads the incoming email, spots what is missing and asks for it straight away. Your estimator gets a complete request instead of half of one.
• Recognising reorders. 'Same as last year' gets looked up in your own order history and prepared as a draft, so someone only has to check it.
• Answering order status. Where is the job, has it been printed, when does it ship, 24 hours a day, without pulling anyone off the floor.
• Explaining file requirements. Bleed, resolution, colour space and file format, in plain language and in your customer's language.
• Following up quotes. A message after three days and another after ten. In a market with overcapacity that is often the difference between the order and silence.
• Sorting requests. Standard work your press handles easily, separated from custom work that genuinely needs a human look.

What you should leave alone

There is plenty in a print shop you should not touch, and we would rather say that now than afterwards.

• The calculation itself. If you run MultiPress, Prindustry or something similar, let it do the maths. Paper prices, setup times and machine speeds are properly modelled there. A language model inventing a price alongside it is the most dangerous thing you could build.
• File preflight. That has been solved for decades by specialist software that knows exactly what it is doing. There is nothing for AI to improve there.
• Press planning. Sequence, setup times and which jobs you gang on one sheet: that is experience, not pattern recognition.
• Judgement on colour and quality. If your customer calls about that, they want a human on the line, and they are right.

Do the maths: what is one extra order per week worth?

An average graphic company buys in 43 percent of its net revenue as paper, ink, plates and outsourced work. The remaining 57 percent is value you add yourself (GOC, December 2025). So on a 1,500 euro order, roughly 855 euros is left to pay your people, your machines and your building.

Now suppose that by answering faster and more completely you win one extra order per week. Over 45 working weeks that is 45 × 855 = 38,475 euros of extra added value per year. Put the alternative next to it: hiring someone. In this industry an extra full-time employee has to generate 134,500 euros in added value on average, and in 2024 employer costs added another 39.6 percent on top of gross salary (both: GOC, December 2025). So that one extra order per week covers about 29 percent of an employee: too little to hire for, plenty for automation that costs a few thousand euros once.

Do fill in your own average order value and your own number of requests, because those numbers are yours and not ours. If the result leaves you cold, that is a perfectly good outcome too. Now you know.

The estimator who retires takes your knowledge with him

The same report names what the industry itself sees as its biggest staffing challenge for the coming years: the accumulated graphic expertise has to be handed over from departing older craftspeople to newcomers with little or no graphic training. Sixteen percent of companies expected bottlenecks in 2025 in retaining craft skills and control of the production process.

That knowledge usually lives in one head. Which customer always supplies low resolution, which paper you recommend for which job, which finisher you call when it is urgent, why you block half an extra day for that one recurring order. As long as it is written down nowhere, it leaves with the person. So before you automate anything, this is the first win: writing it down and making it searchable. After that an assistant can do something with it. The other way round does not work.

What it costs

I start with a free introductory conversation. We discuss what takes time, which systems you use and where automation could help. You then receive a proposal with automation opportunities, integrations, a schedule and costs.

What you can start with this week

• Count your quote requests for a week. How many were complete enough to price immediately?
• Check how many quotes are open right now that nobody has chased.
• Ask your estimator how many hours a week go into chasing details instead of calculating.
• Find the ten questions customers ask most about supplying files. Is the answer somewhere they can find it themselves?
• Run the one-extra-order-per-week sum with your own average order value and your own margin.

If you end up at a couple of hours a week, leave it be. If you end up at a day a week or more, it is about real money and it deserves a conversation.