Imperfect estimates

A little more on estimating the value of automation:

No estimation is perfect, by definition.

• You'll have some uncertainty about some of the numbers.

• Human experiences such as stress and low morale are harder to quantify.

Regardless of these limitations, your estimate is still very useful, if you make it with reasonable care.

For example, say your estimate comes out to $100K. It's possible that someone else in your organization might estimate it at $90K or $110K.

But is it likely they'd come up with an estimate of just $10K, or of $1 million?

An estimate of "$10K to $1 million" is too broad to be very useful — the top end is 100 times the lower end.

But an estimate of "$90K to $110K" doesn't have that problem. Yes, there's some uncertainty in the range, but that top number is only about 1.2 times the lower number.

Here's the thing:

Every estimate, like every measurement, carries some imprecision.

If I tell you I'm 5 feet 10 inches tall, the actual truth may be that I'm 5 feet 10 and one-eighth inches tall, or 5 feet 9 and fifteen-sixteenths inches tall. And even those numbers might not be precise enough to reflect my actual height down to the 1000th of an inch.

But we don't let that stop us from taking such measurements, and then using them to describe the world around us.

It's the same for your estimates. There will always be uncertainty and imprecision.

Don't let that stop you from making careful estimates, and then using them wisely to apply your resources toward your goals.

All the best,
A.