ExcelNation #2: how expensive is your spreadsheet habit?

ExcelNation #2: how expensive is your spreadsheet habit?

Would you advocate allocating 30% of your wage bill to undocumented, untested software?

Author: Circadian – circadian-capital.com

Labour is free, right??

The software costs nothing, everyone knows that, but what about the time your staff spend in Excel? Have you a got a good handle on this?

What fraction of your labour cost is spent on spreadsheets? Are you higher or lower than competitors? Is the software worth it?

Maybe you have a rough idea of the answers, but we’ve all been working with Excel so long we stopped asking questions years ago.

So, let’s guesstimate some costs to test your intuition. Armed with this, you will at least be able to benchmark and evaluate other approaches.

How much does it really cost?

Nobody knows exactly, but we can probably come pretty close – and the answer is fairly scary.

Let’s go with a model like this –

Workforce A | dvanced

On the one side we have sophisticated high-intensity users. They’re the authors/builders who design spreadsheets that often go on to become permanent assets of the firm.

These spreadsheets:

  • Take weeks or months of committed time to build (pre-maintenance) and are called upon every week/month when somebody “runs the numbers”.
  • Will have data inputs separated from formulae, multiple dedicated tabs (heck, coloured tabs even :), and hundreds of thousands of cells.
  • Some use VBA;
  • Others take hours to run.
  • They’ll definitely crash from time to time.
  • The more complex models even have their own dedicated pilot.

These models will be used for analysing data (e.g. financial, operational), determining trends, making projections, evaluating alternatives etc.

Typically we find these spreadsheets in Operations/Manufacturing, Engineering and Research, Finance.

Errors in these models represent risk to the organisation, a fact users and managers are well aware of. There will be no documentation, user testing or version control.

Workforce B | asic

On the other hand many users will be maintaining listicles, comparing names, addresses, responses, tracking data (e.g. budgets, sales, inventories) etc.

These files use fewer functions, are smaller, and less long-lived. In many cases these spreadsheets act as ad-hoc editable databases maintained across users. Sharing and maintaining consistency can be a problem.

We find the prime creators and distributors of these files live in the Marketing, Sales & Distribution, and Human Resources departments.

These users spend fewer hours per week in Excel and don’t much enjoy the experience.

There will be no documentation, user testing or version control. No change there then.

Okay, so how much?

Dartmouth has done some research so we don’t have to (left hand side). We can use the time buckets to figure out with some estimates for your workforce –



  • 34% of all employees’ time is spent in Excel.

That initially feels high – but not massively so, and certainly not in finance (as I look around I see 2 of 5 screens with Excel open) – and other research points to 38% of office workers’ time being spent on Excel.

  • Advanced users spend nearly 60% of their day in Excel.

If anything that might be slightly low.

  • Advanced users make up nearly half of your employees.

This rings true – most people end up becoming fairly useful on Excel.

Copy paste that somewhere and play around with those numbers to see if you can adjust for your workforce.

Here’s how those numbers look for a firm with a higher engineering/technical quotient –

In other words, some 30% of your employees’ time is spent on advanced spreadsheet work, most of which is directed toward advanced prototypes built without a safety net and where the firm might be running active positions and risk.

So where does it leave us?

Spreadsheets: there is no better alternative

I would argue that Group B isn’t the problem – consider this a residual cost. There is no faster way of building and managing mutating ‘databases’ that are fairly short-lived in any case.

Yes they’re a bit awful, but they’re damned cheap and get the job done.

Advanced Users: alternative approaches instead?

Group A though feels awfully spicy.

I can’t think many employers would actively choose to allocate 30% of the labour bill for staff to operate undocumented, untested software.

And yet we all do it. No wonder the Bank of England gets a bit miffed at us from time to time.

About Circadian

Transform your office procedures into services. Fund ranking, Model Portfolios, Client Docs, Client Comms. These are just some of the ways we build your business.





Make possibility reality

Become an IA FinTech Member
and see where it takes you.

Login to your account