Nearly every first meeting opens the same way: "We want to digitise — which software should we use?" The question sounds reasonable but sits in the wrong place. Choosing software is step three. The two steps before it decide whether the project lives or dies, and neither costs a thing.
Step 1: Find where you are bleeding, not where it sounds modern
Digital transformation is not digitising everything. It is moving the steps that cause measurable losses onto a machine. To find which ones, spend a session with the person doing the work and write down three numbers:
- How many staff hours does this step eat each week? (Multiply by hourly pay to get money.)
- How many times a month does it go wrong, and what does each mistake cost to fix?
- What opportunities are lost because it is slow? (Late orders, customers walking away, stock past its date.)
Those three numbers turn "this feels inconvenient" into a specific sum per month. That sum is what you compare against the price of software — and it is also what lets you say no to proposals that sound good but rescue nothing.
Step 2: Write down the current process, ugly parts included
Before discussing software, map how the work actually runs: who types what, into what, who they pass it to, who approves, where it is stored. Pen and paper is fine. What matters is mapping the real process, not the one in the company handbook.
The real one always has ugly corners: a personal spreadsheet the accountant keeps, a group chat where orders get confirmed, a notebook in the warehouse. Do not hide them. Those corners are exactly what the software has to attach to, and exactly where projects fall apart when the developers did not know they existed.
Software cannot replace a process nobody understands yet. It only makes the misunderstanding run faster.
Step 3: Choose the approach, then the tool
Only now does software enter, and there are usually just three roads:
| Approach | Fits when | Trade-off |
|---|---|---|
| Use a finished product | Your process matches most of the market: sales, stock, basic cash | Cheap and immediate, but you adjust some habits to fit the software |
| Finished product plus custom extensions | Mostly standard operations with a few unusual points | Middling cost, depends on how extensible the base platform is |
| Custom build from scratch | The process is a competitive advantage, or no product can cope | Costlier and slower, in return the software follows how you work |
The common mistake is jumping straight to the third row because "our business is unusual". Most businesses are not unusual at the till or in the cash book — they are unusual in one or two places, and only those one or two places are worth building.
Work in slices, not in one block
A "complete digitisation in six months" project is almost always late, and late in the worst way: six months gone with nothing usable. The safer route is slices of 2–4 weeks, each of which has to stand on its own:
- 1Slice 1 — Stock and goods: receiving, issuing, stocktakes, low-stock alerts. Usable the moment it lands, no waiting on anything else.
- 2Slice 2 — Sales: orders draw on the stock data from slice 1 and deduct automatically.
- 3Slice 3 — Receivables and cash: attached to the orders from slice 2.
- 4Slice 4 — Management reporting: by now the data is clean enough for reports to mean something.
The order is not arbitrary: each slice feeds on the one before. Building reports before the data is clean is the fastest way to produce a beautiful dashboard nobody believes.
Three costs that get forgotten
The software price is only part of it. When budgeting, add:
- Migrating old data. Export from the old system, clean it, load it, reconcile. For a business with a few years of history this is heavy work, not a quick step.
- Your own staff’s time. Surveys, pilot runs, learning — all real hours, and during the parallel run they are doing the job twice.
- Running cost after handover. Servers, domains, backups, support. Small but constant, and without it the system decays.
Signs you are on the right track
After the first four to six weeks, the project is healthy if you can see three things: at least one team has abandoned the old method entirely; there is one report management genuinely opens rather than one built for show; and the figures in the software match what you count in reality. If none of the three is true, stop and review rather than pushing on to finish.
SealCore runs this first survey at your premises, including when the conclusion is that you do not need custom software and a finished product will do. To talk specifics, see the solution areas or call the number in the contact section.
Want to talk specifics?
SealCore surveys at your premises and sends a fixed quote after the first session — including when the conclusion is that you do not need custom software.
