We normally don’t reveal the names of our clients, but it’s public information that we were part of the “strike team” that helped assess and turn around California’s unemployment claim backlog in 2020. As more states seek to modernize their unemployment systems, we have some advice about what makes for an effective strike team:

  • No one needs another list of problems. When we show up, we try to solve as many problems as we can while we’re on site–ranging from inefficient SQL queries, to installing system instrumentation like New Relic, to building a workload management spreadsheet. (In fact, we commit to multiple hands-on-keyboard fixes in most of our contracts.) An assessment should understand where you are, where you need to go, propose realistic steps for how to get there, and implement as many solutions as possible before writing a report.

  • It’s critical to follow real claims through the process from start to finish in order to identify (and solve) issues. While many things can be done over Zoom today, this is not one of them. A map of the claims flow invariably exposes surprising issues in technology, policy, and human systems.

  • The mainframe is not the problem. The mainframe is never the problem. The mainframe is probably the only thing saving your organization from disaster, in fact. Stop focusing on replacing mainframes and COBOL. We have never (ever) (not once) seen a mainframe replacement work. Gradually replacing subsets of functionality with new systems can work, but it is a topic lengthy enough for its own series of write ups.

  • Focus on outcomes. In unemployment, you want to know how many claims you have; what your backlog is; and how you can optimize processing that backlog down the fastest. Looking forward, you want a plan that avoids future backlogs, or at least a framework for rapidly addressing them should they arise.

  • We have never had a project where we did not recommend system instrumentation: both across your IT systems (e.g. New Relic, Splunk APM) and across your websites (e.g. Google Analytics). You probably don’t need to invest in a mobile version of your website if 95% of your website visitors are using desktop computers; you probably really need a mobile version if 95% of your website visitors are using their phones, and calling upset when they can’t get through online.

  • Consider previous assessments and reports. If they recommended the same things you’re now recommending: why did they fail? Make sure you are clear on how your plan is different and more likely to succeed. Find the people in your organization who know why the last recommendations failed, and have actionable opinions about it (this is something we specialize in).

If you’d like help with your unemployment modernization plans, we’re available. We also suggest you read the California unemployment strike team report for ideas.