Prioritizing public sector IT projects

On November 1, Barry O’Reilly – author of Lean Enterprise – joined our Thursday night IKNS discussion. He had a ton of interesting things to say.

I asked him about using the cost of delay divided by duration (CD3) in government. He describes the value of CD3 as an economic-decision-making-framework for projects. The idea is that if you calculate the revenue lost or not earned by waiting to complete each project, you can compare the relative value of projects and use that to prioritize the projects.

Unfortunately, the public sector isn’t driven by revenue, and so a strictly economic model isn’t going to work.

I suggested using time-saved and constituent-impact as public-sector-analogs for revenue.

I wish I had a more defined solution, but that’s all I have right now. The truth is that even estimating time-saved and constituent-impact might be unattainable. For two reasons:

  1. Public-sector-baseline-maturity
  2. Constituent-impact-complexity

First, people say “good enough for government work” for a reason. And it’s not because there aren’t any hard-working, butt-kicking government-workers. There are. But, despite those workers, there is a legacy of inefficiency (and some inefficient workers who take advantage of the legacy of genuine bureaucracy). Part of the legacy is a lack of baseline awareness. How much should we get done per day? Per week? Per year? Government (at least local government, where I work) is famously squeaky-wheel. That is, the squeaky wheel gets the grease. The loudest voice gets the attention.  The guns and gavels carry the day. The legacy of a pre-data world is very much alive.

I won’t say anything about the fact that lots and lots of government organizations are making the shift to measuring outputs and targeting outcomes.

Second, how do you measure constituent impact? The number of people? The kind of impact? The kind of person impacted? Is an hour of an impoverished-I-work-two-jobs-to-raise-my-two-children-as-a-single-mother’s time worth more or less than the real-estate developer’s time? Technically, on an hourly basis, the developer’s time is worth more. But saving an hour of a busy single-mother’s time seems pretty impactful to me.

Even if we decide the mother’s time is more valuable, how much more valuable? And that’s to say nothing about estimating the potential number of mothers who will be impacted by a project. Let alone the actual number of mothers who will be impacted by a project.

You get the idea.

The reason I loved Barry’s CD3 recommendation is that it takes me one step closer to creating an I-can-abide-by-that portfolio-management methodology in the land of a thousand cooks. And, as they say, one step at a time.

Leave a Reply

Your email address will not be published. Required fields are marked *