Project management and scheduling with shifting priorities and project juggling

Question

I'm new to any kind of formal PM and structure, but have been tasked with coming up with schedules task lists for a 'mom and pop' construction business that mainly operates by contract (small project, < 30 days completion time). We are a small business with < 30 employees.

My issue is that priorities and project focuses are either ambiguous, uncertain, or shift from day to day. Nailing down anything with management/ownership is very difficult; project and task juggling is the norm, and nothing is on a schedule until the day of (which may change by 10AM, who knows?). We also have a lot of ambiguity coming from a construction methodology standpoint, new technology, and vendor uncertainty. The norm is "fly by the seat of your pants" and "this is what we are going to do until we're not". Like it or not, this is what I am dealing with.

My question is this: how best can I handle trying to schedule out projects (using a Gantt chart, I guess?) or any kind of project or resource management (I've got the PMBOK and some other books) when management/ownership shifts tasks from day to day or morning to afternoon, new priorities or focuses arise overnight, and projects can be killed or put on hold instantly? I know my situation can't be unique.

(My second question is more pointed: does MS project have anything in place for such a situation, besides manually refiguring tasks?)

Answer 1

Traditional project tracking techniques are not going to work here. You're in a dynamic and shifting environment, which does not lend to traditional, "waterfall", style planning. Set the PMBoK to the side and reach for different tools.

Based on what you've described, a simple Kanban system is probably your better approach. The key reasons are:

• It is a pull-based model, where work is pulled in when someone can work it.
• It handles interrupts well. If something is a higher priority, it moves to the top of the "ready" column and the next person who can work it will pull it when they are free.
• Highly visible. If you set this up on a physical wall, with post-it notes and blue tape, everyone can immediately see where work is at.
• Easier to set and enforce Work in Progress limits. You'll know really fast if someone is working on multiple things at once.

You can learn more about Kanban from either of these two books: Kanban - David J. Anderson Start Stopping, Start Finishing- Arne Roock

Answer 2

Joel's Kanban suggestion is a fantastic one that would enable your company to deploy employees on whatever job on demand. However, I don't think that would relieve you of your requirement to measure performance on a specific job in order for you to credibly forecast future finish dates and report to your customer. And you need that forecasting ability in order to ensure you will have resources on site when you need them. If you have a special piece of equipment originally scheduled to be on some site three weeks from now, you will need as early as possible indicators to know if that date is coming in or slipping; else, you will have a piece of equipment on site on that date with possibly nothing to do and that is an expensive error.

I think you still need to develop your schedule per construction job as you would under more normal circumstances but building it, choosing proper planning values for duration and start and finish dates, based on the chaos of your environment. And then you need to invest in the talent and the processes to continually update and modify (not rebaseline necessarily) each schedule based on what occurred that day. This means daily updates, likely several times a day, that will push and pull your milestones on the schedule in a very active and nauseating way. But if your schedule is constructed properly and maintained properly, it should give you reliable indicators from which you can forecast future finish dates.

If Kanban enables one to do that, that's great, but I am not aware of that possibility.