Win a copy of Re-engineering Legacy Software this week in the Refactoring forum
or Docker in Action in the Cloud/Virtualization forum!

# Hey can you give me some suggestions on how to estimate time...

Ivan Jouikov
Ranch Hand
Posts: 269
I know it sounds kind of weird, but my manager wants me to figure out how much time I can spend on how many things and how many clients I can handle at once.

What it comes down to is being able to give her a good estimate of how much time it would take me to do a certain task.

I�ve tried estimating in the past and I usually end up underestimating significantly, like I�d say it would take me about 6 hours to set up a payment processor with paypal, but it would normally take me maybe 10 hours because I�d have to read up some stuff, correct bugs, test it.

So I wanted to know if you can give me some advice on how you estimate how much time it takes you to do something (and therefore how much \$\$\$ to charge). Do you have some sort of a system developer for that or what?

Jeffrey Spaulding
Ranch Hand
Posts: 149
Depends on you efficieny

I am good with

- First Estimate
- Think it over (will give +10% to +20%)
- Multiply by 1.3

Voila the result

J.

Ivan Jouikov
Ranch Hand
Posts: 269
Yeah but just today a client wanted me to implement an email notification system (a reminder type of thing) for some of his online services, and I estimated that to be at about 2 hours of work at most... I mean just do a daily cron and a simple check against the whole DB and send emails accordingly...

According to your ways I should amp that up to 3 hours? Making the client pay \$150 instead of \$100?

Ben Souther
Sheriff
Posts: 13411
I believe the algorithm from "The Mythical Man-Month" was:
(Initial estimate x 2) + 10%.

This is based on the observation that at least 50% of the hours that go into a software project are spent debugging. Knowing that, it makes sense to put that 50% in the estimate right up front. The 10% is to provide a margin of error for the initial estimate.

This conversation would probably do better in either "General Computing" or one of the jobs forums. It's definately not Servlet specific.

Mark Spritzler
ranger
Sheriff
Posts: 17278
6
Yes, usually a developer will always underestimate. What you need to do is to keep data on your estimate and what it actually took you and then find the difference, so that the difference can be your personal modifier. So that in the future you make an estimate then multiply it by you modifier.

Also, this is more a Jobs Discussion topic than a Servlets specific question. Actually it has nothing to do with Servlets.

Moving this to Job Discussion Forum.

Mark