Last time I talked about “cheating” at the PMO prioritization game. Ok, it wasn’t cheating, more like a trick to keep your projects small, fast – and easy to fit into a crowded schedule. Totally above board, but some might call it manipulative. Remember the old management hack … how do eat an elephant?
If that’s too cliche for you, let’s talk about ways to consistently win at the prioritization game. How can you make sure your projects are legitimately higher priority than the next one? Two areas to look at … packaging and parity.
Packaging: Build a Better Mousetrap
There’s a lot to be said for project “marketing”. It’s a fair analogy – you are competing for resources (customers) among multiple alternatives (projects). If you can package the project correctly, you can rise above the din and get the attention you need.
I’m not talking about inflating the business value or padding the project charter with buzzwords and altruism. It’s more about making sure all your i’s are dotted and t’s are crossed, with a concise description, a clear cost-benefit, and a workable action plan. Rise above your competition, don’t take any shortcuts – remember, most people in IT …
<warning> I’m going into some over-generalizations here – just to make a point. Don’t take this personally …
Make your projects “the cream that rises to the top” by focusing on the following:
- Description: Succinct and to the point, you should be able to describe what, why, how, and how much in less than 200 characters each. Remember, this “list of projects” is probably going to appear as a SQL query or Excel spreadsheet with 20 to 50 other projects along for the ride.
- Business Value: This, too, needs to be a simple statement of real business value. Stay away from soft terms like productivity and cost reduction unless you can back them up with examples. Best to align it with named corporate strategic initiatives, too.
- Scope: Hey, that cheating stuff actually works. Keep the project size down to something that can actually get done in the foreseeable future; in a time of scarce resources, people don’t like to commit long-term (and >90 days is really long term). Deliver incremental results – think Rod Carew (singles and a high batting average), not Dave Kingman (lots of home runs, but lots of strikeouts).
Parity: Portfolio Management
There is a fundamental flaw in all of this “project marketing” stuff – we should be prioritizing those efforts that will bring the business the most value. It’s really not about win or lose for my projects – what is the right thing to be working on right now?
A good prioritization process will be facilitated with a tool like the one shown below
The list includes many attributes for each proposed project …
- Which core business objective(s) does the project support?
- What business unit is being supported?
- What are the costs involved? Includes capital + expense, one-time and recurring.
- What are the benefits delivered? This version expresses benefits in terms of payback / ROI.
A fair and equitable mix of projects will make sure that all business areas and strategic objectives will get their proper share of the project resource pool. Thus it behooves the responsible relationship manager / IT liaison to make sure they understand all projects being proposed, so we can quickly define how the scarce resources are being shared among competing business units – and make sure
the right projects get started.
Note: On my source code page there is any copy of the spreadsheet shown above. I have hacked up the code to genericize / obfuscate the original, which featured some nifty stuff. The buttons at the top left, for example, allowed the user to pull in all cost and benefit metrics from standardized project description templates (other spreadsheets). All totaling was automatic; set your budget and then check off projects that will get funded, until you run out of budget.
Lots of neat features that (probably) no longer work – much of the code is intact, however, so feel free to download and hack away. If you are interested in a fully working copy, let me know, we can talk ….