“if all you have is a hammer, everything looks like a nail”. – Abraham Maslow
Excel is great, but it is not the answer to every problem. Excel has its limitations and they should be considered before using it in a business-critical application. The decision to use Excel should be an active one, rather than the default option.
Over the years, I have learned through painful experience where and where not to use Excel.
Where to use Excel
Where the situation demands, Excel is often the most pragmatic choice of analytical tool.
However, the user should proceed with caution and be mindful of its limitations. In particular, in relation to the risk of error. Appropriate review processes must be put in place.
Where analysis is bespoke
For a one-off transaction, where the analysis is tailored to a very specific situation, not to be repeated, Excel is often the best tool for the job. Its ease of implementation and cost-effectiveness makes it the perfect tool for bespoke analysis.
Where analysis needs to be flexible
Where the calculation logic supporting your analysis is subject to frequent change, the flexibility of Excel is your friend. With sufficient training and a well-structured model, it is relatively easy to make structural changes on the fly.
Where analysis needs to be completed quickly
The clock is ticking, you need to get some analysis completed and on your boss’s desk by the end of the week. For most of us, Excel is the only sensible option.
Analysis needs to be freely shared internally and externally
Excel is ubiquitous. It is on pretty much everyone’s laptop, and most people have at least a basic grasp of how to use it.
Where you need to share analysis widely either within an organisation or externally, Excel can often be the only realistic option without running into issues with the other parties not having the right software to open/amend your analysis.
Where not to use Excel
As much as I love Excel, with the luxury of time and budget, it is not always the right answer. Here are some situations where I would avoid using Excel.
Very large data sets
Excel is not a database. It not only has a limitation on the number of records per sheet, but also suffers performance issues as data grows.
High volume repeatable transactions
When you are working with a high volume of repeatable tasks, i.e. most operational finance transactions, order processing, invoicing, accounts payable, receivable etc. There are generally much better, purpose-built applications available.
Excel does not handle collaboration well. When there is a requirement for multiple users to be accessing/amending data at the same time, then it’s better to look elsewhere.
The open structure of Excel leaves it prone to misuse by unqualified users. Where your users need simplicity and do not need / or want to see under the hood, then Excel is probably not the right tool.
Have you ever tried to shoe-horn Excel into an application it wasn’t suited for? Or succeeded in using Excel for a novel purpose?
Read more from our modelling risk series.