Digitising an SME: where to actually start

“We need to go digital.” The sentence comes up in every board meeting, and it is almost always followed by the same move: buying software. Six months later the licence is paid, nobody uses it, and the handwritten notebook is still on the counter.

This is not a tooling problem. It is an ordering problem. Here is the order that works, in the companies we support.

Step 1 — Write down what actually happens

Before buying anything, spend a week recording the journey of one order, from the customer's request to the payment. Who receives it, on what medium, who passes it on, where it waits, where it gets lost.

The exercise looks trivial. It almost always reveals two or three places where information is re-keyed identically, and one place where it regularly disappears. Those are your priorities. They are not the ones you would have guessed.

Step 2 — Remove paper where it costs the most

Not everywhere: where it costs. Paper is not a problem in itself. It becomes one when it has to be searched for, re-typed, or when losing it stops the work.

Start with a single form — the delivery note, the service report, the leave request — and make it live digitally end to end. One. Finished, used daily, and adopted by the team. That first visible success is what makes the second project possible.

Step 3 — Be findable, and reachable

Your customers look for you on a phone. Three things matter, in this order:

1. An up-to-date business listing — hours, address, phone. It is free, it shows first, and it is the most frequently out of date.
2. A site that loads fast on an average connection. A beautiful site that takes eight seconds is a site nobody sees.
3. A channel someone actually watches. An address displayed but never read does more damage than no address at all.

Step 4 — Connect what already exists

Most SMEs already have the pieces: a stock spreadsheet, an order book, invoicing software, a till. The immediate gain does not come from replacing them, but from wiring them together.

When an order placed on the site decrements stock, notifies the workshop and feeds invoicing, you have removed two re-keyings and one source of error — without changing a single one of your tools.

Replacing a tool everybody masters with a “better” one nobody knows is the most reliable way to make a digitisation project fail.

Step 5 — Measure, then automate

Automation comes last, and only on what is already stable and measured. Automating an ill-defined process means making the same mistakes faster and at greater scale.

Once your data is clean, the same questions finally get numeric answers: which product sells, what delivery time you actually hold, when in the week your team is overloaded.

What to refuse

• The all-in-one solution. It does everything by halves, costs a lot, and migrating off it is a project in itself.
• The vendor who keeps the credentials. Domain, hosting, accounts: all in your name, from day one.
• The eighteen-month project. Cut it up. Shipping every two months teaches you more than a hundred-page specification.
• One-shot training. A three-hour demo does not replace three one-hour visits during the first month of real use.

An order of march, not a catalogue

Quarter	Goal	Sign it worked
1	Map it, pick one process	Three people describe the same procedure the same way
2	Digitise that process, become findable	The paper form is no longer printed
3	Connect the existing tools	No data is entered twice any more
4	Measure, then automate one repetitive task	You steer on figures, not impressions

How long, how much

Less than feared for the first three steps — they rest mostly on decisions and internal time. Budget arrives at step four, when the links between your tools have to be built. That is also when return becomes measurable, which makes the spend defensible to a partner or a lender.

The IT Advisory & Digital Transformation page describes how we run this scoping. If you are at step one and mapping feels insurmountable, let's talk: it is usually a half-day workshop.