Lessons in Flow and Value Creation from the US Midterm Elections

Early voting for the US midterms opened Monday in Texas. Politics aside, this year expects to see unprecedented turnout, based on what we’ve seen already in primaries and special elections as well as early polling.

Eager to perform our civic duties, my colleague Trevor and I hopped on an electric scooter and rode the ten blocks to the nearest polling station in Austin. Before we left the office, we checked an online map with wait time data to make sure the lines wouldn’t be too long. We got the ‘all clear’ from the map, but when we got there it was a different story. The line snaked around both sides of a long indoor corridor and it appeared to be moving quite slowly.

Not ones to be easily deterred (and to be honest I was afraid to get back on that scooter) we joined the line to cast our votes. It was painfully slow and when I finally got into the voting room itself, I could see why. They had clearly not optimized for flow.

Ten voting machines stood unutilized while a poll worker struggled to verify a single voter’s address. There was only one workstation for verifying identities and that bottleneck was holding up the line, increasing wait times for the one hundred or so people and starving workstations further downstream, including the actual voting machines.

Now, I know we’re dealing with a government operation here and they face some challenges when it comes to resources and scale (plus, in many ways, we are a captive audience). But this was an awfully bad end-user experience and I guarantee it is actually undermining the value the US Government is trying to create.

Let’s think about that for a minute. In a democratic society, citizen participation is a key value for both Federal and State governments. The government’s desired “business outcome” should be high voter participation in general elections (again, politics aside). The Secretary of State of each individual state is tasked with making that happen.

To achieve the desired outcome and maximize participation, you must get as many people as possible to vote. That requires low barriers to voter registration; easy and convenient access to polling locations; efficient, fast moving lines at the voting locations; and secure, easy-to-use and fully functional voting machines.

Long lines at a polling place are a deterrent and effectively lower participation. Therefore, polling places should be designed to maximize flow through the system.

So, how does one go about maximizing flow? You need to measure your baseline, implement improvements and measure it again. There are a set of flow metrics that help organizations learn how to optimize for speed in a system, in order to create more value:

  • Flow Velocity: In a given reporting period — let’s say 30 minutes — how many people complete their voting? This measures the system’s throughput.
  • Flow Time: How long does it take, from the moment a person arrives at the polling station till the moment they cast their ballot and leave? This measures time to value.
  • Flow Efficiency: Of the total time spent at the polling station, what percentage of time are people actively voting vs. waiting? Where are the bottlenecks causing those waits?

If the Texas Secretary of State measured these things, I’m pretty sure he would gain insight into how to architect the process for speed and efficiency and the ultimate creation of value. He would have a baseline of the end-to-end system performance. He would be able to get to the root cause of long wait times and relieve the constraints to flow. And then he would be able to measure again and see if he’s improved. And then he’d move on to bust the next bottleneck.

Moreover, imagine he displayed these real-time metrics in an online dashboard for all polling locations? That would be more reliable and more helpful than the red light/green light map Trevor and I relied on.

Taking it one step further, what if all states used Flow Metrics and shared their data? The Texas Secretary of State would be able to compare his polling stations’ metrics with those of other states and learn from those that are doing better. He might be able to build his case to get more funding, for example to equip each station with multiple computers to verify voter identity and not just one.

As I so often do in my role of helping organizations with their digital transformations, I asked myself — what can this experience teach us about building products at scale? What can it teach us of the challenges our customers — IT departments in XL enterprises and government agencies — face to deliver value to their customers faster?

Enterprises need to start using Flow Metrics for their digital products and services, because they are the only metrics that help understand how value flows and where it is constrained, and most importantly, how IT performance correlates with desired business outcomes.

Success, after all, is about getting the outcome you want — not about the process itself.