- Problem Definition
In previous experience I have worked on a FEED Project, the project was undertaken by a ring fenced task force where I was assigned as the project engineer. The project duration was given as 11 months by the Client (my original client from hell). Unfortunately the Project over-run its schedule by an additional 2 months. This analysis will study if there was a better way to estimate the schedule duration.
According to the PMBOK the following are the different tools and techniques used to Estimate activity durations:
- Expert judgment:
- Analogues estimating
- Parametric estimating
- Three point estimating (PERT)
- Group decision making
- Reserve analysis
While these methods are all viable, their precision and accuracies differ. If we take a closer look at the three point estimating method we could tell that the precision of these types estimates are much higher and therefore more practical to be used at this stage of the Project.
- Feasible Alternatives
In this analysis we will be looking at a comparison between two alternative methods:
- Continue using a combination of Expert judgment and Analogues estimating or
- Use the more accurate duration estimate using the Three point estimating (PERT)
- Develop the outcomes for each alternative
- A combination of Expert judgment and Analogues estimating – based on the time it took to do similar tasks on other projects and based on the experience of Client in similar projects
- Three point estimating (PERT) – duration based on preferable % probability (usually 80 – 90% probability)
- Acceptable Criteria
The acceptable criteria for this analysis will be a duration estimate accuracy >90% based on historical data (actual duration).
- Analysis and comparison of the alternatives
To perform PERT analysis, I contacted a few of my experienced colleagues to obtain their input on the time required to complete a similar kind of project with a ring fenced task. A PERT analysis carried out gave us the following results
Comparison between alternatives:
|Client Estimate||11 months||85%|
|PERT (P90)||13 months||100%|
|PERT (P80)||12 months||92%|
The result shows that using the PERT (P90 or P80) will give us a better estimate that fits within our acceptable criteria of > 90% accuracy.
- Select the preferred alternative
Based on analysis result, PERT P90 shows that the duration estimate is exactly the same with actual duration. The P80 analysis also gives us an accuracy of greater than 90% which would also be acceptable. All in all, the Client estimate can be seen as less accurate and more of a time driven schedule.
- Performance Monitoring & Post Evaluation of Result
It is advised that the schedule be analyzed a number of times as time goes by and as more information is available to get more accurate results. It is also advised that the schedule is analyzed in a more detailed manner for example looking into the duration of each task and analyzing each task on its own.
- Estimating Time Accurately: Calculating Realistic Project Timelines. (n.d.). Estimating Time Accurately. Retrieved June 12, 2014, from http://www.mindtools.com/pages/article/newPPM_01.htm
- Six Methods for the Estimation of Activity Duration in Project Management. (n.d.). Small Business. Retrieved June 12, 2014, from http://smallbusiness.chron.com/six-methods-estimation-activity-duration-project-management-41782.html
- A guide to the project management body of knowledge (PMBOK guide) (5th ed., ). (2013). Project Time Management. Newtown Square, Pa.: Project Management Institute.
- Martin, D. (2014, March 25). Estimation Part 2 – Accuracy vs Precision. . Retrieved June 12, 2014, from http://www.dontpanicitsolutions.com.au/lean-thinking/53-estimation-part-2-accuracy-vs-precision/53-estimation-part-2-accuracy-vs-precision
- Spent, D. (n.d.). Estimate Activity Durations — PMP Primer. PMP Primer. Retrieved June 12, 2014, from http://www.pm-primer.com/estimate-activity-durations/