Estimation is the ability to see project tasks

I have been a software professional all my life.

Over the years have done tonnes of projects.

Each of them have required me to propose the solution and then propose an effort estimate meaning how many man days of work and what kind of man days i.e. architect days, administrator days etc are expected.

I have been able to take this one step further....

I have been able to judge with 90% accuracy the effort that could be required based on a small description of the business outcome for which software must be developed.

How do I know I am that accurate !!

Well, I have been diligent enough to track the effort post my estimation of a project effort.

But I thought my project estimation skill would be restricted to the software projects related to my area of business.

I was wrong ...

Consider these two true stories.

Setting my MBA marketing class. We are discussing a very famous case study.

The case study says the CTO of this project company thought he had 90% chance of making the new model of the projector available. Yes, you read that right. Projector as in overhead projector we use to make presentations.

But some events leading to that pronouncement of the CTO make me think. ...

I try to imagine the complexity of processes required to make a new model of the projector.

I tell the professor based on what I am hearing the company does not have 90% chance that the new projector model will be ready in time so the company must not pursue the strategy of launching this model.

Professor says facts of the case must be considered as provided and the discussions moves on.

I am a little disappointed. I had staked my word in front of 90 other senior executives in the class.

Case discussion finishes.

Professor again asks me in front of the class what I thought was the likely probably of delivery of the new projector.

I said it was 60 to 70%.

To me and my classmate's surprise the CTO of the company had actually given that same exact figure to the board. But the figure had been changed for the benefit of the case discussion to 90%.

I was vindicated !

But most importantly, I learnt that my project estimation was not restricted to the kind of projects that I work on.I could use it for projects related to other businesses.

So why was I right ?

I think it is because project estimation is about being able to

A. Judge the various components of a project then

B. Estimating the complexity that will be encountered for each of those components. Followed by

C. Estimating the time required to obtain buy-in from the various stake-holders. Remember time required is a result of the number of meetings, presentations and approval processes required. And then to

D. Judge and estimate a ball park figure of the effort in days for the above.

E. This ball park figure now needs to be validated with your gut experience developed over delivery of many projects that you have been part of. Hey ! I did not say it was easy.

Voila! you have done your project estimate. With practice they should be in the 80% - 90% accuracy range.

Wait .... I did say I had two stories.

My second story is when I estimated for a shipping software after a 5 minute speech by a classmate who was a ship fleet owner about what the software did.

Well. I did give him an estimate of the profit percentage the software vendor was making based on the my estimate of the work required to make the software and then customize it for his needs.

You guessed it right. The ship fleet owner dutifully strong armed the owner of the vendor company to confirm that my estimate of the profit margin was accurate!

Do you have such experiences? What have they taught you about project estimations?