An internal tool to replace the spreadsheet
The job your business currently runs out of a spreadsheet that three people have a copy of. Built around the steps you already take, on your own accounts, with an export back to Excel on day one.
This is the build where the price varies most, because the scope is your process and nobody else has the same one. The estimate comes after the call, not before it.
What actually happens,
and when it is your turn.
Your rows are marked. Add them up and the whole thing costs you well under two hours, which is the number every other agency's process page leaves out.
From the call to the day it goes live. One figure for the whole build, agreed in writing before any of it starts.
What moves it: how many steps the process has, how many people touch it, and what it has to read from or write back to.
YouA longer call — an hour rather than twenty minutes. We watch the spreadsheet being used rather than being described.
The process is written down as steps, with the exceptions listed separately. That document is the whole build and it is yours whatever happens next.
YouRead it with whoever actually does the work, not just whoever owns it. They will find the step nobody mentioned.
The fixed price, against that written scope. If the honest answer is that you need something bigger than this, you get told here.
A working version on a private link, with real data imported from the spreadsheet so it is your work you are looking at.
YouUse it for one real day in parallel with the spreadsheet. Nothing else finds the missing step this fast.
Roles, exports, the awkward exception, and the reports somebody currently rebuilds by hand every month.
YouEveryone moves over on a day you pick, with the spreadsheet kept read-only for a month in case.
Six choices
only you can make.
None of these are technical and all of them change the result. They are listed here so the call has an agenda and so you can think about them before it rather than during it.
Most spreadsheets have grown a second job on the side. Building both costs more than building one, and it is worth being deliberate about which you are paying for.
Roles, set by you, changeable by you. Not everybody needs to see margins, and finding that out afterwards is unpleasant.
If a number has to be right to the franken, it needs a stricter level of checking than a week buys. Naming those fields early is what keeps the price honest.
Customers or suppliers with accounts is a different security problem and a different price. It is a yes-or-no at the start, not a later addition.
The accounting system, the warehouse, the shop. Each connection is quoted on its own because each one is a different amount of work.
Your database, your storage, your logins. If you stop paying us it keeps running, and that is only true if it was set up that way from the first day.
What to have ready.
None of it is required. All of it makes the price firmer and the first draft closer.
Three questions
this one gets and the others do not.
Why not just buy something off the shelf?
Often you should, and we will say so. A product that fits eighty per cent of your process and is a hundred francs a month beats a build. This is worth doing when the last twenty per cent is the part that makes you money, or when fitting your work to somebody's product is exactly how the spreadsheet won in the first place.
What if we outgrow it?
The data is in a normal database on your own account and the export is there from day one, so outgrowing it is a migration rather than a hostage situation. That is deliberate: a tool you cannot get your data out of is a worse spreadsheet.
Three weeks does not sound like enough.
For an ERP it is not, and we would not take that job. Three weeks buys the tool your team uses every day for one process, built properly. If the call reveals something bigger, you get told before there is an invoice rather than in week two.
If the work is repetitive rather than structural — the same job every week by hand — a role may be cheaper than a tool.