Project schedule reviews

Question

As a PMO project coach, I am often asked to review the time schedule of projects. Some are detailed schedules in MS Project while some are high level (capacity) planning in a spreadsheet. There are waterfallish types of projects, as well as projects using sprint planning, like (Scrum) or Kanban.

Even though few projects are alike, I try to be objective by reviewing the following items:

• Exhaustive WBS and scope: Is there a big chance on surprises or big changes?
• How reliable are the estimates?
• Resource availability checked.
• Any external dependencies?
• Have other risks and risk mitigation plans been taken into account?

The overall aim of course is to improve predictability and coach junior or new PMs with best practices. But all this doesn't come cheap, and I need to be able to scale this process (although the number of projects is growing, current staffing of the PMO needs to remain constant). So short of rebuilding the schedules myself,

• What would be a standardized, minimal and methodical check for a schedule review?
• Would it be possible to assign a score or metric, some kind of quality evaluation of the schedule that could serve as an early warning sign?
• And finally, is there an easier way to compare the original plan against what actually happened (analysing the five bullets above, taking into account the impact of changes and management decisions), without getting lost in the finer details of project life but good enough to be used as Lessons Learned?

Hmm, now that I have written out this question, I am even more skeptical :-).

Answer 1

A standard approach to conduct audits will quickly turn into a useless control that everyone will be scared to stop using but no one will use the data from it. If I were you, I'd develop my audit approach consistent with the goal I am after as well as the project and its unique set of issues and risks.

I can see a checklist to sort of bang any project against at the highest level, but any depth of analysis would require a unique approach, in my opinion. And it is NOT cheap; therefore, you would want a unique approch in order to get the biggest bang for every audit dollar spent.

Answer 2

I don't think you can assign a metric that would hold or be useful against all of the project types, but I think you're list here is a good start for the review.

I would use a checklist with your five points, delving a little deeper into each one -

WBS - show how the WBS was arrived at, who had input, why was it structured that way, etc? This will give the PM's some insight into the process and may point to gaps.

Estimates - how were they estimated, do they include O,ML,P ranges, is this work that's been done before or new tech, etc.

And so on. I think that if you build a simple checklist, and then have the PMO run through it with the PM providing answers and supporting info as to how the schedule was built you'll get a good read on the reliability.

As for Lessons Learned - best practice there is to do it at the end of each phase and compare actual to planned (both schedule and plan), and document as you go. That way decisions are fresh in the mind, and lessons learned on hase 1 may be able to be implemented on Phase 3 to avoid similar issues. At the end of the project it's really too late for LL. Everyone's spent, trying to move on, and if it didn't go well they're looking for where to pass the buck. Get the lessons before there's that protective need.

Answer 3

For a high level review I would compare the amount of uncertainty against the amount of "buffer" in the schedule. More specifically, how much room is built in before scope x needs to be delivered by date y.

This relies heavily on experience and a general feel for the range of what could go wrong and a feel for stakeholder tolerance/parameters to accept change in the schedule, scope, cost...

Answer 4

Budget versus actuals isn't too hard if you have a formal estimation model, but can be difficult otherwise. For instance, I've seen organizations do analysis of:

Model Estimate (Original Scope) -> 200 mandays Actuals = 300 mandays Model Estimate (Actual Scope) -> 275 mandays Mis-estimate = 25 mandays

Without a formal model, this is hard to do. Even with a model, sometimes PMs get lazy.

In a prior life at a large consultancy, the general experience was estimates were over-run because of unplanned work, rather than mis-estimated work.

For early checking of plans, I currently use dual checklists - one that contains a bunch of questions related to how known is the scope, the other related to how complex it is. The goal of these is teasing out how much you don't know going in, or what surprises you might have.