Document digitisation: when it pays off and how to do it well

Anyone who does document digitisation for a living should say it first: you do not always need it. Sometimes a scanner and a shared folder are enough. For a handful of contracts to keep, or for everyday mail, for example, they are usually the right choice.

Yet in some cases a folder of scans costs more than it seems. The cost is not in the scanner. Instead, it is in the time people spend looking for a document and retyping its data. In this article I explain how I tell the two cases apart.

Why the folder of scans starts ahead

A folder of scans costs little, is ready in an afternoon and everyone in the office can use it. It needs no decisions about fields, names or formats. Its cost is also predictable: the time of whoever does the scanning.

That is why my starting point is always the simplest option. The question is not "why digitise?". The question is "what does it cost us, today, not to have the data in these documents?".

When document digitisation pays off

In general, the signs are fairly clear. Usually, moreover, more than one shows up at once.

• Side spreadsheets. The data in the documents is retyped by hand into a spreadsheet, because a scan cannot be filtered or added up.
• Steps outside the system. Data copied by hand from a document into the business system, email used as an archive, checks left to memory.
• Documents arriving in different formats. The same kind of information comes on paper, as a PDF, as a photo or by email, and each format is handled differently.
• Archives nobody can consult any more. Binders or folders of scans with no consistent names: finding a file from three years ago takes half a day.

In these cases document digitisation is not a luxury. In fact, it is often how you stop paying twice: for the time spent filing and for the time spent retyping.

When document digitisation does not pay off

By contrast, I advise against digitising when documents are rarely consulted and nobody uses their data. A good scan with a clear name is enough in that case. Invoices, accounting records and payslips, moreover, follow their own rules and stay with whoever keeps the books: they are not the work I describe here.

I also advise against starting when it is not yet clear what needs extracting. After all, document digitisation fixes a way of reading: which fields, which names, which classifications. If those choices change every month, it is better to settle them first on a few sample documents. Only then does it make sense to handle the whole archive.

The middle path: file where that is enough, extract where it is needed

In reality the choice is rarely all or nothing. Often the most sensible answer is a mix. Part of the documents stays as it is, scanned and filed under a clear name. Only the documents that matter become data: the ones behind reports, comparisons or frequent searches.

Here is a hypothetical example. An installation company runs a common business system for orders and customers. However, its technicians fill in job reports on paper, and someone retypes them. So the whole archive does not need digitising. What is needed is to digitise the reports: extract the data from the sheets, check it and load it into the existing business system, flagging doubtful cases. Nothing else changes.

This way document digitisation stays small and focused. As a result, costs stay small too.

The costs of document digitisation to plan for

In document digitisation, however, the scanning cost is only the first line. An honest quote considers others:

1. scan quality: resolution, skewed pages, handwritten sheets that a person has to read;
2. sample checks: part of the extracted data is always compared with the original, and that time must be planned;
3. delivery formats: a spreadsheet, a CSV file or a direct load into the business system are different jobs;
4. retention: where the originals and the scans are kept, for how long and who can see them.

There is also security, which is often overlooked in document digitisation. A company's documents often hold personal data and confidential information. So, on the documents I receive, I apply least-privilege access, tested backups and deletion of the copies at the end of the job, as agreed. These cost less when they are decided before starting.

Finally, I always ask for two points to be clear in writing. First, who owns the data and in which format it is delivered. Second, what happens to the originals and the copies when the job ends. Of course, the same questions apply when the work is done in-house.

Data and lock-in: what the EU Data Act changes

On lock-in there is news from Europe. The Data Act has applied since 12 September 2025. Among other things, it requires cloud service providers to make switching to another provider easier.

Also, according to the European Commission's explainer, from 12 January 2027 providers may no longer charge switching fees, including data egress fees. Until then they may still charge the costs they incur for the switch.

What does it mean in practice for document digitisation? Digitised data must stay yours and portable. An archive you can only browse inside a subscription service, with no full export, is a new form of lock-in. Still, taking the files out is not enough if names and classifications exist only in that service. That is why I deliver data in open formats, with the list of fields and the rules used.

The questions I ask before starting

Before I propose anything, I ask the company to answer a few questions:

• which documents are retyped by hand today, and why?
• how much time do searching and retyping cost each week?
• which data is really needed, and which can stay in the document?
• who will use the digitised data, and in which system?
• what should happen to the originals when the job ends?

Often the answers are enough to decide. Sometimes they lead to a better-organised folder of scans. Other times, instead, they lead to a targeted digitisation of only the documents that matter.

On the page about digitisation of documents and archives I describe how I set up the work: which documents, which data, which checks.

If your company has an archive nobody consults any more, or a spreadsheet filled by retyping documents, write to me. I will tell you whether it is worth digitising, and how.