Somewhere in your business there is a spreadsheet that started as a convenience and became infrastructure. It tracks the orders, or the bookings, or the stock, or which client is owed what. Three people edit it. One of them has a colour-coding system that only they fully understand. Nobody would describe it as a system, and yet if it were deleted this afternoon the week would be a write-off.
Spreadsheets are excellent tools and there is nothing embarrassing about relying on one. The question is not whether spreadsheets are good — it is whether this particular one has quietly taken on a job it was never designed to do.
The tipping point is rarely about size. It is about how many people depend on it and how badly it fails when something goes wrong.
Two yeses is normal. Four means the spreadsheet is now a liability rather than a shortcut, and the fourth question is the one that occasionally turns into a genuine compliance problem rather than an inconvenience.
The instinct is to replace the whole thing, which is exactly how a small useful project becomes an eighteen-month rebuild that nobody adopts. Replace the workflow, not the file.
Pick the single process that costs the most time or causes the most errors — usually data entry, an approval step, or a handoff between two people — and build only that. A first internal tool is typically four things: a form with real validation, a list you can filter, permissions so people only see what they should, and an export back out to a spreadsheet.
That last one matters more than it sounds. Keep the spreadsheet for analysis. What you are removing is shared editing of critical data, not the ability to open something in a grid and look at it. Tools that cut off that escape route get quietly abandoned.
Validation that refuses impossible entries. An audit trail showing who changed what and when. Roles, so the part-time assistant cannot accidentally overwrite the pricing column. And a single source of truth that does not exist in four versions across three inboxes.
Building is the right answer less often than agencies like to admit. Run the three-way test first.
Buy it if the process is standard. Invoicing, CRM, bookings, payroll, helpdesk, stock control for a common retail model — these are solved problems, and a subscription will beat anything custom on cost, reliability and support. Building your own version of an off-the-shelf product is one of the most expensive mistakes a growing business can make.
Automate it if it is a handoff. If the pain is really "this data lives in tool A and someone retypes it into tool B", the fix is an integration or an automation, not an application. Days of work, not months.
Build it if the process is genuinely yours. When the way you quote, schedule, allocate or price is a real part of how you compete — and no product matches it without heavy compromise — a small custom tool pays for itself, because the process is the advantage and you should not be bending it to fit somebody else's software.
Build the part that is uniquely yours. Buy everything that is the same in every business on your street.
Whichever route you take, decide before you start who maintains it, where it is hosted, how people log in, and how it is backed up. The build is rarely the expensive part; an unowned internal tool three years later usually is.
Look at the spreadsheet your team would panic about losing. If several people edit it, if data gets retyped out of it, and if a new hire needs a tour to understand it, it has outgrown the format. Replace the riskiest workflow first, keep the export, and buy or automate anything that is not specific to how you work. The goal is not to be impressive — it is to stop a Tuesday afternoon from being ruined by a broken formula nobody spotted.
We build small, focused internal tools that replace the fragile parts without disrupting how your team works.
Talk to us