Friday, 30 March, 2012 11:29 PM
A project manager's key skills are knowing what to focus on, and managing uncertainty, yet our project management tools provide little help. MS Project can show us "Here are the 200 tasks due to complete this month", it cannot tell us "Here are the 5 that you should focus on, because these are the most likely to derail your project. It can tell us that the project is scheduled to finish at 3:15PM on Wednesday the 14th of August 2013, but not whether this prediction is likely to be overshot or undershot, nor by how much.
There is an attempt to manage uncertainty in MS-Project through PERT - Project Evaluation and Review Technique - where you give three durations: optimistic, expected, and pessimistic. However not only is this well hidden, it is unrelated to the rest of Project and doesn't appear to have even rudimentary validation. You can, for example, specify a task as having a duration of 5 days, but an optimistic duration of 10 days, an expected duration of 20 days, and a pessimistic duration of 2 days. Go figure! Even if you do find the PERT features and use it properly, it's hardly worth the effort for all the practical help it gives you in running a project. I haven't used PERT since the days when state-of-the-art was project management software with punched cards.
In 1997 Eliyahu Goldratt published "Critical Chain", claiming to be a new way of managing projects that improves project performance through explicit management of buffers. See http://www.robertb.co.nz/project.htm for a number of papers on Critical Chain Project Management (CCPM). CCPM generated a flurry of interest in the project management community, but it is simplistic and overhyped even though some organizations have reported great successes. Also, it is not well supported by mainstream project management software such as MS-Project, making it difficult to use. Project managers like me who see the value of explicit buffers but want to use this concept within conventional project management use an informal combination of Project, Excel, Word and gut feel to understand what is actually happening within our projects.
MS Project could do so much better with a relatively simple enhancement. This paper, first published in 2001, shows how: http://www.robertb.co.nz/GeneralResources/Buffered%20Gantt.doc It defines a very simple way of properly handling uncertainty properly, representing it on the Gantt chart and through to the final project completion.
I am told that the best way of having this idea noticed and picked up by Microsoft is to post it to a forum: I hope I've chosen the correct one. I'm happy to continue with discussion of this idea if anybody thinks its worthwhile.
- Edited by Robert Barnes Friday, 30 March, 2012 11:30 PM
Saturday, 31 March, 2012 1:45 PMModerator
I'm not clear what the goal of your posting is -- but I can assure you Microsoft will likely not "pick up" things made in a post to a forum. If you think your idea has legs, I believe the good folks at Microsoft would encourage you to develop and market an add-in or other value add to Project.
As an aside, Project 2010 no longer includes the PERT feature, so you may wish to update your report by removing references to the feature.
- Marked As Answer by Mike GlenModerator Saturday, 30 June, 2012 9:47 PM
Sunday, 1 April, 2012 8:56 PMModerator
I read your document. I agree that CCPM is simplistic and on the whole seems to be a way of covering for poor scheduling. I personally schedule as tight a series of dates as possible then add Risk buffers calculated from remaining risks. Risks have % chance of happening, a $ impact, but also a schedule impact. So if a particular stream of work (EG house foundations) has a risk with a 25% chance of happening and a 10d schedule impact if it does, I insert a 2.5d Risk contingency in the schedule (no resources unless extra work is involved). If the risk does not happen, I use Project 2010's new inactive feature to remove it. An estimated schedule date can be calculated using earned value metrics, but I use durations and work rather than cost to do this.
For Microsoft to include a new feature, there has to be a widespread demand for it. Microsoft does have a wish list site (try searching for it), but only ideas that are most popular and are also deemed important by key partners and customers are likely to get the go ahead. I know because I've asked for a number of features I deem important for professional schedulers but am still waiting many versions later ;-( .
So thank you for your ideas, but if you passionately want to see it implemented, you need to start a campaign!
Rod GillThe one and only Project VBA Book Rod Gill Project Management
- Marked As Answer by Mike GlenModerator Saturday, 30 June, 2012 9:47 PM
Tuesday, 17 April, 2012 9:28 PM
Thank you for your feedback Julie. I posted here because local MS staff told me that this was where I should post my idea if I wanted it picked up.
Back in 2001 when I had the idea for "Buffered Gantt Charts" I tried to develop an add-in, but was unable to because the task and gantt objects lacked sufficient properties or methods to express what was needed. Or perhaps my VBA skill wasn't enough. I figured that it wouldn't be too difficult to implement for a MS programmer with access to the source code, but it seemed to be impossible with VBA for Project. There is already a Critical Chain add-in called "Pro-Chain", but this seems to make the same simplistic mistakes as Critical Chain Project Management and I was not inclined to purchase it to investigate it in detail.
Thanks for letting me know that PERT is no longer in Project. It's not surprising that it's been dropped, it was so well hidden that only those of us with long memories would have known to search for it, and then when you found it it the implementation was so poor that you'd only try using it once.
Tuesday, 17 April, 2012 10:06 PM
Rod, thank you for your feedback. It seems that you handle buffers the same way as I do, by manually adding buffer tasks into MS Project. My proposal provides a simpler and better way. The main benefit is that, by making buffers visible and an accepted part of the project - in other words, accepting reality - we change the dynamics of task scheduling.
Normally the focus is on "THE estimate". When the project leader asks a project worker for the estimate, the worker's incentives are strongly around "Not exceeding the estimate", he will get little credit for coming in early but major discredit for being late. Therefore the worker's estimate will be as large as possible. However if he is asked for two estimates: "What is your most optimistic estimate" and "What is your guaranteed estimate" the nature of the conversation changes and the leader and worker communicate more honestly, and then can work together to achieve a better project outcome. If you are interested I'll send you some PPT from my project management course showing how this works - this is invariably rated the best part of this two-day course.
I googled "Microsoft Project Wish List" and found this: -
Is this where I should post the idea? This seems to be unofficial, not a MS forum.
BTW, the product "Spider" is, I think, project management software that already incorporates a similar idea. But I don't want to consider using an obscure software product from goodness-knows-where, I want to use the market leader that everybody is familiar with.
Monday, 21 May, 2012 10:42 AMModerator
Please re-post if this is still an issue, or mark it as answered.