Predicting the Future by Looking Back
tip
Data cannot predict the future.
When you have projections in a spreadsheet it looks black and white, but there are a bunch of assumptions built into any model. Think of how many times you've made an estimate and been way off? Have you ever, even once, been perfectly accurate? I doubt it.
When you're too conservative and you come in under budget, you feel great.
When you're too optimistic and you come in over budget, you don't feel great.
So when making predictions, especially complicated predictions, like cashflow for a large company (or a small company for a long period of time) and you're combining lots of data, you need to review and look at your predictions and see how they compare with actuals on a consistent basis.
Here's where the review comes in, a construction company could stand to make a lot of money by not incessantly looking forward. Look back through a project, figure out where your actual wins and losses were. Have an actual, no blame, talk about what went well and what didn't.
"Did you you lose your shorts on material over runs? Oh but you found a way to get it for half the price, so it actually looks like you're a genius."
You got lucky.
"Oh, you actually estimated materials amazingly well but didn't account for the price escalation that was in the vendor's contract?"
I bet you won't ever make that mistake again.
The other thing that's important here, and a call forward to 003 is how you should use color if you're looking at these predictions. There should be something that's marking that the data source has switched from actuals to projections.
project highlight
none this time, sorry guys