This slim volume is a refreshing look at something that I am not used to seeing. I am involved in one project at a time, occasionally more. A project overseer has to have a different view.
If there are several projects, these could potentially have demands on the same (limited) resources. Johanna Rothman has an aversion to context switching, and rightly so, as this produces unproductive time as a change between what was current and what is now current occurs. She makes a forceful argument in favour of allocating resources (machine, people, budget, etc.) to the highest priority item that is to be worked on without distraction, working down a list. It is therefore necessary to rank projects. It is only when ranking has been undertaken that it is known which is the highest priority item. The aim has to be to create release-able running tested features.
Of course, any ranking is not permanent. Rankings change over time, and a high priority item that is not completed may subsequently take a very low priority as the reason for its (early) completion has now disappeared. Rankings have to be reviewed, and Rothman uses the three-fold categorisation of projects, to decide what should be worked on. For each item under consideration, a decision needs to be made; commit, kill or transfer. She strongly advises against any other decision; never say maybe to a portfolio request. When you say maybe, your manager hears `yes'. Your peers and staff hear `no'. You are then placed in a no-win situation.
Part of the commit, kill transform decision can involve recognising projects that can and will not succeed - doomed projects. Killing your own (doomed) project is very different from killing a senior manager's pet project. There are good pointers throughout the book to helpful sections elsewhere, and these are both forward and backward looking, and deal with just this kind of knotty problem.
The resource allocation process that Rothman advises includes never committing more than 50% of your staff for 3 months or more to the same project. Can the project be broken down into something that can produce demonstrable results in under 3 months? If so, it is something that could be contemplated (and may go on to a second, third, fourth stage.) Something larger than 50% of your people and/or longer than 3 months could end up being a `bet the organisation' decision.
In amongst the words are some little gems. She talks of individuals manipulating scoring regimes and metrics - they will do this if there is any advantage in it. Therefore it is very important to watch your metrics, to see if they are telling you the right thing. She advocates Project Managers to start somewhere, but most of all to start. Finally, she advises all against using any form of ultimatum with a senior manager. Her wise words are that these rarely result in anything except a career-limiting conversation for you. That probably means there will be one less person on the payroll - and YOU will be the one leaving.
The view from 30,000 ft is different to that from 500 ft. This book is more about that high level aspect. Knowing about life from that height can actually make more sense of life at 500 ft.
Peter Morgan