Digitizing a Paper Workflow: Step by Step

Why do paper workflows survive so long?

Almost every small company has workflows that still run on paper: technicians’ job reports, vacation requests, delivery notes, incoming invoices with an approval stamp. They survive not out of nostalgia. Paper is robust, anyone can use it, and the workflow works as long as nobody looks closely at how long it takes and how often people have to follow up.

When such a workflow is digitized, something odd often happens. The paper form becomes a PDF form that is printed, filled in, and scanned. Effort rises, the effect fails to appear, and the team feels confirmed that paper was better. This leads to the guiding question of this guide: How does a small company digitize a paper workflow step by step so that it actually becomes faster and less error-prone, instead of just moving the form onto a screen?

To answer it, we first clarify the terms, then introduce Michael Hammer’s principle as a thinking model, review findings on the effect of digitization, and weigh whether to replicate a workflow one to one or rethink it. At the end there are seven steps using the job report as an example, how to handle retention rules, and a step for this week.

A scanned form is a digital photo of paper, not a digital workflow.

What does digitizing mean, and what is a media break?

A workflow is a recurring sequence of work steps with a clear beginning and end. Digitizing a workflow means that information in it is captured, passed on, and processed electronically without anyone retyping, printing, or forwarding it by email. This definition comes from the guide to digitalization in the right order, which also describes which workflow comes first.

A media break occurs when information changes its medium and is transferred by hand: from paper to keyboard, from email to spreadsheet, from phone to sticky note. Every media break costs time, and every one is a source of errors, because things are copied, shortened, or misread. A workflow is only digital when the information passes every handover without a media break.

A second distinction comes from the lean office. Processing time is the time someone spends working on a case. Waiting time is the time the case sits: in the van, in the inbox, on the pile. As the article on process before tool shows using a quoting workflow, waiting time in office workflows is usually a multiple of processing time. That is exactly where good digitization starts.

The thinking model: don’t automate, reorganize

Michael Hammer, a computer scientist and management consultant, published an essay in 1990 whose title became a formula: don’t automate, obliterate (Hammer 1990). His observation: companies invested heavily in computers and rarely achieved large improvements, because they sped up workflows based on yesterday’s technology, assumptions, and objectives. The strength of information technology lies precisely in breaking away from such routines.

His best-known example was Ford’s accounts payable department in North America. According to Hammer, more than 500 people worked there in the early 1980s, while Ford’s Japanese partner Mazda handled the same task with about five. Ford later achieved a substantial reduction, according to Hammer of 75 percent, not through faster invoice checking but by reorganizing how order, goods receipt, and invoice were matched. From such cases Hammer derived principles. The most important for small companies is: capture information once, and at the source.

Like a letter that is scanned, sent by email, and printed out again by the recipient, a workflow remains a paper workflow as long as someone handles the information anew at each station. If it is captured where it originates, by the technician on site, by the supplier in the invoice, by the customer in the form, the media breaks after that disappear by themselves.

What studies show about the effect of digitization

That technology alone achieves little is well documented. Brynjolfsson and Hitt (2000) showed that a large part of the value of information technology lies in complementary changes, that is, in new workflows, ways of working, and responsibilities. In a survey of about 400 large companies, those that made decisions in a more decentralized way got an effect from the same investment that was 13 percent higher on average.

How a concrete paper workflow changes is shown by a case study from a large university hospital in Turkey. Zekioğlu and colleagues (2025) followed the digitization of the patient records archive in three stages: scanning, cost and time comparison, and interviews with employees. By the end of the study, 222,461 files had been digitized. According to their cost-benefit calculation, the equipment for the digital archive was 2.37 times more advantageous than building an additional paper archive.

The third part is notable. Before the changeover, the archive staff described liking their work but suffering from dust, heavy lifting, and searching for misfiled records. After the changeover, they reported higher motivation, less work-related worry, fewer physical complaints, and a feeling of being valued. The study has clear limits: a single institution, only three employees interviewed, and an archiving task that differs from a job report. But it shows that digitization can be experienced as relief by those who carry the workflow daily, if it removes their actual burdens.

For a new solution to be used, Davis’s technology acceptance model (1989) points to two perceptions. Perceived usefulness is the belief that the tool improves one’s own work. Perceived ease of use is the belief that using it takes little effort. For a technician, this means: digital capture on site must be faster than the paper form, or the paper form stays.

Replicate one to one or rethink? Two perspectives

The first view says: first replicate the workflow exactly as it is. The team knows the fields and the sequence, the changeover is easy, and improvements come later. This view has a true core, because a digitization that changes everything at once overwhelms a small team.

The second view follows Hammer: whoever replicates the old workflow digitizes its detours as well. Fields that are on the form only because someone needed them twenty years ago, signatures nobody checks, copies for files nobody looks at all migrate into the new system and become harder to remove there.

I consider a combination right, and it matches the MSS sequence from the article on process and tools: map, simplify, select. Before anything becomes digital, whatever nobody needs is cut. Then the simplified workflow is replicated in a familiar way, with one big change: the information is captured at the source. Everything else, such as automations, follows only once the new workflow runs stably.

Seven steps using the job report as an example

A building services company with 14 technicians records jobs on paper. One real case is traced from the job to the invoice, and each station is noted with working time, waiting time, and medium.

1. 1. Choose the workflow

   Take a workflow that occurs frequently, on which money depends, and that generates follow-up questions. The job report meets all three conditions because it directly delays the invoice.

2. 2. Trace a real case

   Follow an actual case from start to finish and note station, medium, working time, and waiting time. Count the media breaks.

3. 3. Simplify

   Cut fields nobody uses and steps that check nothing. In the example, sorting is dropped, and the owner’s approval applies only above a value limit.

4. 4. Capture at the source

   Decide that the information is entered where it originates. The technician records times and materials on a tablet on site, with required fields for whatever otherwise triggers questions.

5. 5. Choose the tool

   First check whether existing software can handle the workflow, such as mobile capture in the invoicing or trade software. Only if not does a new tool come into question.

6. 6. Abolish paper

   After a short trial with two technicians, the switch happens on a fixed day and the paper forms disappear. A lasting parallel operation doubles the work.

7. 7. Measure

   After four weeks, the same case is traced again. Waiting time, follow-up questions, and the time from job to invoice are measured.

Two articles help with the fifth step. The guide to choosing software describes requirements, test phase, and decision. The overview of no-code automation explains when automation within existing software is enough and when an integration platform is needed. So that the new version of the workflow stays findable for everyone, it is recorded as a process page using the one-page method.

Retention and records: what to consider when abolishing paper

Whoever abolishes paper must clarify what happens to the records. Retention periods apply to tax-relevant documents, and the German tax authorities have set out in the GoBD, the principles for the proper keeping and storage of books, records, and documents in electronic form, how electronic records must be retained. Under certain conditions, paper records may be destroyed after scanning if the procedure is documented in a traceable way and no special rules stand against it.

For invoices, the situation has also changed. Since January 1, 2025, all domestic companies in Germany must be able to receive e-invoices from other companies, and a PDF by email is not an e-invoice. Whoever digitizes incoming invoices should think about both together, as the article on the e-invoicing mandate describes. Which documents in your own company must be kept for how long and in what form is something to clarify with your tax advisor. This guide does not replace that advice.

Conclusion: how does a paper workflow become a digital one?

The guiding question was how a small company digitizes a paper workflow so that it actually becomes faster and less error-prone. The answer follows Hammer’s principle: capture information once, and where it originates. To do so, a real case is traced, media breaks are counted, whatever is superfluous is cut, and only then is a tool chosen, preferably one already in the company.

As the findings show, technology works mainly together with changed workflows, and it is adopted when it genuinely saves those involved effort. The largest gain almost never lies in working time but in waiting time. The sober observation is: a workflow is not digital when there is no more paper form, but when nobody has to retype anything anymore.

The next step

1. Choose a paper workflow that occurs frequently and on which money depends.
2. Trace a real case from beginning to end and note medium, working time, and waiting time at each station.
3. Count the media breaks and mark the longest waiting time.
4. Together with those involved, cut all fields and steps nobody needs.
5. Check whether your existing software can already capture data at the source before you look for a new tool.