Effective Analytics Outreach to Departments in New Orleans

By Zachary Markin • April 13, 2017

The City of New Orleans’ Office of Performance and Accountability (OPA), the city’s home for data-driven improvements, has had notable successes in using analytics to reduce fire risk and address blight. These programs originated through OPA’s performance management work and with ad-hoc requests from departments. As the city’s data capacity matured and OPA completed more successful projects, director Oliver Wise sought systematic ways that the OPA team could better engage with other city departments at different levels of data literacy. Although the OPA team has strong data skills, he knew that the most successful projects would require the expertise and ideas of departments.

Wise said, “In order to scale the NOLAlytics program, we knew that we had to harness the creativity and capabilities of public employees to identify opportunities where analytics could provide value.”

With the help of Richard Todd, a graduate student from Princeton’s Woodrow Wilson School of Public and International Affairs, OPA developed a toolkit for departmental engagement around data. The portfolio includes presentation tools, an analytics typology, a website, a form for departments to express interest, and sustained programming of departmental engagement. According to Todd, the motivation of the work was “to generate analytics projects with insights from central OPA, but also to empower those people who work in the field and will be bursting with insights if you give them the tools to engage in the analytics agenda.”

The overarching framework for the strategy is the NOLAlytics Use Case Typology, which describes six possible opportunities for analytics to be used in government. Todd noted, “the purpose of the typology is not to be analytically pure, but to bridge the [knowledge] gap.” The categories are:

Nolalytics Civic Analytics Typology

1. Finding the needle in a haystack

This approach uses predictive modeling to identify opportunities for impact when specific targets are difficult to identify within a large group. For example, a regulatory agency may only have capacity to conduct a limited number of inspections on of thousands of potential violators. Using historical data on previous violations, analytics can be used to predict where these high-risk "needles" are to greatly increase the agency’s efficiency.

2. Prioritizing work for impact

When an agency has a backlog of cases to address, analytics can help prioritize them based on potential impact or complexity instead of in the order they are received. For example, calls for pest control assistance can be reordered based on the danger of serious outbreaks, and the highest-risk ones can be addressed first.

3. Early warning tools

For certain categories of events, a small proactive intervention can result in huge savings in the prevention of an adverse outcome. Often, however, it can be difficult to identify where among many options investing limited resources is worth it. A predictive model can use early warning signals to flag areas for investment of preventative resources. One example is predicting homes where lead poisoning is most likely to enable targeted inspection and faster remediation.

4. Better, quicker decisions

Surfacing important information and building automated systems for providing real-time information to public servants can enhance civic operations. Operational awareness tools such as situational awareness tools for public safety teams facilitate data-driven decisions.

5. Optimizing resource allocation

When city resources are allocated randomly or based on historical precedent, data can ensure that the allocation is effective and informed by current circumstances. One example of this work is standby locations of ambulances, which, with the help of analytics, can now be driven by data about where an incident is most likely to occur.

6. Experimenting for what works

Testing alternative options for services, or “A/B testing,” can inform strategy. For example, if a city is attempting to maximize citizen response to outreach, it can test both SMS and email communications, or even multiple messages within a single channel, to determine which is most effective.


A set of slides describes this typology and focuses on the possibilities of improved services and outcomes to stimulate insights in departments and spur practical discussions. When OPA presents the typology to departments, each category includes an introduction, examples, and guidance on identifying potential projects. The presentation concludes with next steps for departments interested in working with OPA, including how they assess project feasibility and required resources. It also includes a project checklist that departments need to complete with OPA before a project begins.

After finalizing this typology and accompanying toolkit in the summer of 2016, OPA issued an enterprise-wide call for proposals, supported by an “office hours” session open to any public employee. This outreach has led to over 20 new project proposals and a new structure for OPA’s long-term engagement throughout the city.

OPA Director Oliver Wise said, “By cultivating a cache of project ideas, we can strategically prioritize the projects that will likely have the greatest impact.”

New Orleans’ presentation materials and typology are available for other cities to adapt and use in their own departmental outreach.

Top photo credit iStock.com/SeanPavonePhoto

About the Author

Zachary Markin

Zachary Markin manages the Civic Analytics Network. Prior to joining Ash, he worked as independent software consultant and data infrastructure engineer building enterprise scale data integrations, warehouses, pipelines, and analytics tools. He discovered first hand the power of data infrastructure when building custom invoicing, record keeping, and analytics tools to run the Fresh Corner Cafe, a healthy food distribution company that he co-founded in Detroit, Michigan. He finds particular joy in developing technology that solves real world problems. Zach holds a BSE in chemical engineering from the University of Michigan.