The Work Nobody Funds
Internal tools are consistently the last thing a business builds and frequently the thing with the best return per pound spent.
The reason is visibility. A customer-facing feature is easy to justify because somebody can point at it. A tool that saves four people two hours a week each has a larger measurable return and no obvious champion, so it stays on the list while the spreadsheet gets worse.
Meanwhile the cost compounds. The workaround becomes the process, the process becomes institutional knowledge, and the knowledge lives with whoever built the spreadsheet.
Why These Are Cheaper Than They Look
Internal tools carry a genuinely lower bar than customer-facing software, and it is worth exploiting rather than ignoring.
Users are known, trainable and reachable. Edge cases can be handled by a person rather than by code. The interface has to be clear, not beautiful. Onboarding is a conversation rather than a designed experience. Failure is embarrassing rather than commercial.
That combination means a tool that removes a real cost can often be built in weeks. Applying customer-facing standards to internal work is how these projects become unaffordable and therefore never happen.
What To Build First
Follow the workarounds. They are a map of what is missing.
Where two systems are reconciled by hand. The most reliable candidate.
Where somebody rebuilds the same report every week.
Where a spreadsheet has become the source of truth for something that matters.
Where one person is a bottleneck because only they can do a thing.
Where the same question gets asked repeatedly because the answer is hard to look up.
What Moves The Price
How many roles need different access. Internal does not mean everyone sees everything, particularly around pay, margins and personal data.
Integration with existing systems, and their API quality.
Whether it replaces a spreadsheet or a process. Replacing a file is quick. Replacing a way of working needs the process defined first.
Whether it needs an audit trail. For anything touching money or personal data, yes.
Data migration from whatever exists now.
Where These Go Wrong
Building what the manager describes rather than what the team does. The manager describes the intended process; the team runs the real one with three workarounds. Build the first and it will be routed around.
Making it too rigid. Real work has exceptions. A tool that forbids them absolutely gets abandoned in favour of the spreadsheet, which never says no.
Treating it as disposable. Internal tools become load-bearing surprisingly fast. Skipping backups and access control because “it is only internal” is how a Tuesday becomes expensive.
No owner after launch. The same fate as the spreadsheet, one layer up.
Questions To Ask Whoever Builds It
- Have you watched somebody do this job? Not been told about it.
- What are the exceptions, and what happens to them?
- Who can see what? Especially pay, margin and personal data.
- What happens if this breaks on a Monday morning? Internal tools become critical quietly.
- What is the smallest version that removes the worst of the pain?
A Brief You Can Send Anyone
The job: [what somebody does, step by step, including the workarounds]
How long it takes: [and how often]
Who does it: [and what else they would be doing]
Where the data lives now: [including the spreadsheets]
Who must not see what:
The exceptions: [how often, and what happens]
What breaking would cost: [honestly]
How We Approach It
Watch the work first, then build the smallest thing that removes the worst of it, with proper access control and backups because internal tools become critical whether or not anybody planned for that.
Where several of these accumulate into a coherent operating layer rather than separate tools, an internal operations platform is the better shape.
Get in touch with the spreadsheet everybody complains about.