Someone rekeys the same order twice
You already guessed the answer.
Most briefs arrive as a solution — “we need a new website”, “we need a dashboard”. Sometimes that is right. Often it is the symptom talking. We start by finding the constraint, then build the smallest thing that removes it.
What this looks like in practice
Orders arrived by email and were typed into two systems.
What we built
A parser that reads the order, and a single write path into the ERP.
What changed
Order entry went from 40 minutes to under a minute.
The weekly numbers were a spreadsheet one person maintained.
What we built
A read-only dashboard sourced straight from the database.
What changed
The meeting starts with the numbers instead of waiting for them.
An overnight sync failed silently about once a month.
What we built
Health checks with an escalation path that names a person.
What changed
Failures are known within four minutes, not four days.
1 / 3
04 / Range
Most engagements are two or three of these together. Which ones is an output of the diagnosis, not something you pick from a menu — and if none of them fix your problem, we will say so.
Usually the smallest part of the job and rarely the constraint. When a site is the answer, it is because the current one is costing you enquiries you never hear about.
Almost always where the money is. The work is not clever: find the step a person does because two systems will not talk, and remove that step.
A dashboard is worth building once the underlying number is trustworthy. Before that it is an expensive way to display a disagreement.
The cheapest thing on this list and the one most often missing. Something already runs overnight in your business. Nobody is watching whether it worked.
05 / Method
This is the part that makes the range credible. We do not start by picking a tool, and we do not quote until we know what is actually wrong.
A short paid engagement. We watch how work actually moves — not how the org chart says it does — and talk to the people doing it.
A written account of where time and money are going.
One page. The constraint, what we would build, what it costs, and what changes if it works. If the honest answer is that you do not need us, that is what the page says.
A fixed price and a date.
Small releases you can use while they are still being built. You see it working in your business before it is finished, which is when changing our minds is still cheap.
Something in use, not a demo.
Documented, running in your accounts, under your control. You can keep us on for changes, or not. Both are normal.
A system you own outright.
06 / Diagnosis
None of these are technology questions. The answers tell us where the money is going, and they are the same five questions whatever you make or sell.
Duplicate entry is the most common finding and the easiest to price. It is also the one nobody reports, because it has been the job for years.
Usually a process that only exists in someone's head, or a spreadsheet only they know how to repair.
A dashboard will not fix this. Something upstream is producing two answers, and that is the thing to find.
Anything on this list is a monitoring problem. It is nearly always the cheapest fix on the table.
Systems outlive their reasons. Some of the work is switching things off.
07 / Proof
In each case the client arrived asking for something specific. In each case we built something smaller, for less, after finding what was actually wrong.
Tell us what is costing you time. We will tell you what we would build, and what it would take.