We have old saws like this because they are timeless in their validity.
Here are REFM’s recommendations for writing and testing formulas in Excel before deploying them in your models:
1. Does the formula possess all of the relevant elements that it should? (See our post here for the exhaustive generic version of all projection formulas.)
2. Does the formula work with zero dollar (or zero unit) values as inputs, or does it crash out?
3. Does it work as you would expect it would work with $1 (or 1 unit) values? What about with negative $1 or fractional values (this may or may not apply)?
4. Does it pass the common sense test? In other words, does the triggering, growth and termination of the resulting data series make sense in the context of how the other line items in the model do those things?
5. Is the line item formatted correctly?
Thoughts? Additions? Please contribute. The goal is to spend LESS time in Excel, not more.