Prior to giving your estimate, do you understand the boundaries of what it is that you are being asked to do? Have you created a scope diagram? Have you done stakeholder analysis? Do you understand where all of your stakeholders are? Do you understand the type of project (maintenance ticket vs. new software development project)? Depending on your answer to these (and others), you will be able to determine some estimate to your part. The BABOK lists eight ways that you can perform estimation:
1. Analogous Estimation, which is basically a Range-of-Magnitude, sometimes called a ROM.
2. Parametric Estimation. If you have enough history, you can multiply the effort by a factor to arrive at an estimate. So, if history shows that eliciting requirements from a specific functional group takes 10 hours, and this project has a factor of 10, then your estimate is (10×10) = 100.
3. Bottom-Up Estimation. The BA collects the tasks from all the functional groups and rolls them up for a total.
4. Rolling Wave, which is an estimation refinement technique. It involves basing estimates for the next iteration (or phase) on the current iteration’s work.
5. Three-Point Estimation, when you estimate the best-case, worst-case, and most-likely scenarios.
6. Historic Analysis. Go back and look at your previous projects for data to see how much time a similar effort took you.
7. Expert Judgement. Talk to those who have done similar work in the past.
8. Dephi Estimation. This is a good technique if you are getting everyone together and have them vote on how long something will take. The participants vote, discuss, vote, discuss, etc., through several rounds until you reach a consensus. The “Planning Poker” technique is similar to this.

Given your question about a fixed-price contract, you still have to go through estimation, but since your price is fixed, you can only bring a limited amount into scope that fits within that contract. You do the best estimation that you can, and then see based on your price, how much work that you can fit into the deliverable.

Finally, learn from your past estimates. If you gave someone a fixed-price contract and it would up being a lot more work than you thought, keep that in the back of your mind when you start your next project estimation.

Cheers!

Thank you in advance!

Interesting perspective on which type of time to track. The PM may care about effort, but only so far as you can fiot the work in to available time.

Once they start scheduling various workstream and committing to deadlnes they then get more focused on when you can complete the task, right?

I guess cost management and capability are the priorities at first but once you are underway you have to make deadlines so subsequent tasks can get started.

I like to estimate work time. Elapse time requires you to know all of the other things that may get in your way and is much less reliable. The important thing about work time is for everyone to understand that it is the amount of time needed if you weren’t working on ANYTHING else! Most project managers ask for work time and then schedule their project resources for 6 work hours per 8 hour day. This allows time for non-project meetings, breaks, other ongoing assignments, etc.

I looked at your discussion on time tracking and enjoyed it. Keep up the good work!

Barb