10 easy steps to make a perfect software estimate

1. Split everything into small chunks. The smaller the chunk, the higher the chance it will roll off from the table and under the sofa so that it’s gone and you don’t have to worry about it.

2. Make it unreasonably accurate. If you estimate a feature to take 3.35 hours, surely you know what you’re talking about.

3. Actually, do not use time, but points. It’s the same thing but we call it something else because it brings clarity.

4. Do it just before lunch. You’ll just want to get it done quickly, and intuition is what we’re going for here. Hungry people are scientifically proven to be more accurate in predicting the future. I’d link relevant research here, but I cannot be bothered. I’m writing this just before lunch.

5. Double the estimate. Then decrease it by 30% because you felt stupid staring at that big number.

6. Refer to #5 as your formula. People with formulas are people who are respected.

7. Believe in yourself. Your experience is no match for your confidence. Yesterday’s you is but a shadow of your future self.

8. Stupid ideas take much longer to build. That’s just physics.

9. Make lots of unhappy faces when you’re putting it together. Everyone knows unhappy people know what they’re doing and only an idiot would be smiling in this economy.

10. The only way to win is not to play the game. Quit your job.

This took a bit longer to write then I expected.