Financial Reports

The File Everyone’s Afraid to Touch: Key Person Risk in Finance

The File Everyone’s Afraid to Touch: Key Person Risk in Finance
Key Takeaways
  • Most finance teams have one file nobody wants to change: It usually started as a quick fix, grew for years, and now holds logic only one person fully understands.
  • Key person risk is as real as spreadsheet risk: When the logic lives in one head, a vacation or a resignation becomes a reporting problem.
  • Newer tools can create the same problem faster: A script or app built in a hurry can become just as fragile as the spreadsheet it replaced.
  • You can keep Excel and lose the fragility: Moving the data plumbing and the logic into a governed system leaves the spreadsheet as the place people read and work.

On almost every first call, someone eventually says a version of the same sentence: there’s this one spreadsheet, and if we could automate it, that would be fantastic.

Then they tell me who maintains it, and it’s almost always one person. Finance teams call that key man risk, or key person risk: a process only one person can run or fix.

This piece is based on three conversations I had in recent months, with any identifying details removed.

Why do finance teams end up with one spreadsheet nobody wants to touch?

The first team I’ll describe runs budgeting, forecasting and month-end out of a collection of spreadsheets stored in a shared folder. Their finance lead described it as a bunch of Excel files everywhere, with numbers crunched by hand and passed back and forth between them.

One colleague maintains most of it. Every update means touching several files at once, and the more files involved, the muddier things get. Their year-end report runs on logic complicated enough that the finance lead said automating it would be fantastic, and then added that it wasn’t even the first thing on the list. One of their summary sheets didn’t quite add up to the reported total, which is exactly the kind of thing you find when a file has been patched for years.

Each fix made sense at the time, and the fixes piled up. That’s how spreadsheet risk builds. It happens at large banks too. A JPMorgan task force report on its 2012 London Whale trading losses found that the value-at-risk model for the portfolio ran on Excel spreadsheets completed by hand, with data copied and pasted from one sheet to the next.

What does key person risk look like in a finance team?

The second conversation was the most striking I had all season, because the fragile file wasn’t a spreadsheet at all.

The company’s CEO had built his own tool to model customer contracts and handle billing. It pulls data from the accounting system, the CRM and the billing platform, then applies their contract rules, credits and exceptions. He built it quickly, with AI doing a lot of the coding, and it worked.

He was also the first to name the problem. He called it a single point of failure, which finance teams also know as key person risk, or key man risk. He was the only one who really understood it, there was no proper governance around it, and he worried about hidden gotchas: calculations that might drift outside tolerance without anyone noticing.

That’s the same problem as the spreadsheet, just built faster. Speed made it easy to create, and it’s still one person’s head holding it together.

When we showed how that logic could live in a system the whole team could see and work with, he called it super valuable and said, “I love that.”

Can you keep Excel and still get rid of the fragile parts?

The third team does everything in Excel, and they like it that way. The pain is everything that happens before the spreadsheet: manual exports from the accounting system, reports pulled by hand, and a project tracker for capital spending and fixed assets that someone updates line by line. On top of that sits a multi-entity consolidation and a rolling budget, all assembled manually.

What clicked for them was keeping Excel as the window they look through, while the data collection and consolidation happen underneath it automatically. They’d still write in the formulas they know. The data would come in on its own. The part that got the biggest reaction was being able to serve themselves and pull a report the moment they need it. It also gets the logic out of a single head: when the mappings and consolidation steps live in a system with version history and an audit trail, someone else on the team can trace how any number was built.

What does replacing a fragile spreadsheet look like in practice?

The closest example I can point to is a construction consultancy in the UK with several offices. Their management accounts ran on a single spreadsheet that had grown to 30 to 40 megabytes and was riddled with errors.

They replaced it with one version of the truth, built accurate headcount and subcontractor forecasting across every office, and gained enough confidence in their numbers to rely on them for the bank’s due diligence on a loan.

That last part is the one I come back to. A file you’re afraid to touch is also a file you can’t stand behind when someone outside the business asks hard questions.

Where should you start if your team has a file like this?

Ask who would fix it if it broke tomorrow. If the answer is one name, write down what that file does, step by step, while that person is still around to tell you. You’ll need that list whatever you decide to do next, and it’s usually shorter than people fear. Then check three things about the file: whether anyone else can change it safely, whether changes are tracked, and whether its totals still tie back to the source systems.

If you’ve got a spreadsheet, or a script, that everyone tiptoes around, I’d like to see it. For your Datarails demo, bring the file nobody wants to touch, and we’ll walk through what it would look like with the logic in a system your whole team can see. The person who built it usually ends up the most relieved.

Related Articles

Become a Partner

Drive Business Performance With Datarails

Drive Business Performance With Datarails

Drive Business Performance With Datarails

Drive Business Performance With Datarails

Drive Business Performance With Datarails

Drive Business Performance With Datarails