I Have to Ask You: Data-Driven Decision Making

In this series, guest columnists respond to one of three topics selected by ELGL co-founder Kent Wyatt. This week MeghanMarie Fowler-Finn, Performance Manager at District Department of Transportation, highlights three examples of data driven decision-making. 


A few hours before writing this I had just returned home from my first 12-hour shift of the year in Snow Command. Analysts and managers from multiple city departments gather in a large technology rich bunker type room whenever snow is forecast to ensure that no matter what snow or ice comes our way, safe travel is ensured as soon as possible. Not only do snow and ice present deeply personal and hyper-local challenges for residents and visitors to our capital, they present high profile, vulnerable, uncertain, and expensive challenges to city leaders.  Providing efficient and effective services is viewed by city leaders as an important measure of leadership. Our combined Performance Team of the Department of Public Works and the District Department of Transportation helps leaders make deployment adjustments in real time based on automated data.  

Snow Deployment

During the deployment we look at data from multiple systems collected into one centralized and transparent view available to all in Snow Command.  Using this data, we found that we needed more drivers because routes were short-staffed. We could also see that we had an excellent shift change with personnel rapidly checked out and in within an hour of the shift end. Despite not receiving much snow on this first snowfall, we gained confidence that our data systems were working and know that in addition to regular deployment activity we can monitor parking ticketing and towing as well as 311 service requests as needed if a big snowstorm should occur.

Parking Meters

Our team responsible for parking meters had enacted a performance-based contract system earlier in the year and needed to be able to measure performance of contractors. Their strong analytics team began cutting and dicing the data to understand meter up time and service request completion. Our Performance Team chipped in recently when there was a software and contractor system change to understand how this affected contractor performance. We set up an automated morning email to inform the parking team of the work their contractors had completed the day before. By analyzing and understanding their contractor’s work as opposed to taking their word for it, the parking team used data to hold their contractors accountable.  The parking team wants to be as effective as possible and is using their data to make better decisions about who deserves to receive contracts.  

311 Service Requests

Delivering quality and quick service is the goal. Zeroing in on bottlenecks helps identify the highest value process changes as well as possible technology changes to automate the process.  Snow and ice present special challenges in potentially emergence circumstances, but every day our teams complete hundreds of service requests from citizens. Requests range from traffic safety to sidewalk and bicycle issues. One target question we constantly address is how rapidly are we completing these requests? And we can use data to identify answers to sub-questions such as – is one team able to complete steps more quickly than others? And if so why? Is one path to completion more successful? And for which types or requests? We use the data generated by our work order management system to answer, which is generated by the work of our people. The goal is to continue to increase our efficiency and effectiveness of services through the use of data.

A performance team has the ability and responsibility to improve city services by alleviating external reporting and administrative pressures on teams who are doing the work on the ground. We can help them devote more time to doing their best work. I find that once our Performance Team proves our ability to help alleviate pressures, those teams (I see them as our clients) seek our help with operational improvements. And then we can work with those teams to help them use data to find additional opportunities for operational improvements.


Supplemental Reading