Policy & Practice | Fall 2026

There is a national pattern of extended timelines, rising costs, and continued reliance on legacy systems. Projects are taking longer than expected, and most states have not reached operational status after many years of work. At the same time, there is a small number of states that have completed implementations. Idaho and Arizona are notable examples, com pleting their systems in approximately three and six years. West Virginia, while not highlighted here, also demonstrates that implementation is possible under similar conditions that other states face today. These states stand out, not because they necessarily had fewer con straints. They faced pressures around staffing, procurement, competing pri orities, and legacy system demands. But they were able to move through those challenges and reach implemen tation. What’s to be understood from their success? Why CCWIS Is So Difficult CCWIS was never intended to be a simple system replacement. It requires states to change how systems support child welfare practice. This includes modernizing data structures, improving interoperability across programs, meeting federal reporting and data quality expectations, and sup porting more timely decision making for caseworkers. The technology choices are also more complex than ever. Back in the 1990s, when many states still operated their child welfare programs largely on paper, the options to digitize were simpler—generally, you

CASE STUDY, ARIZONA: MAINTAINING CONSISTENT DIRECTION

Arizona’s CCWIS (Guardian) implementation is one of the two examples that the Children’s Bureau identified as successful. The state completed its CCWIS in approximately six years, faster than the national average and ahead of many states still in development. Our firm, Public Consulting Group (PCG), acted as the planning and project management office (PMO)vendor, and we witnessed how Arizona maintained a consistent direction over time. As requirements evolved, procurement decisions were made and resistance emerged, the program stayed coordinated across business leadership, technical teams, and vendor partners. That coordination depended on governance that supported timely decisions. It required a shared understanding of priorities and a consistent approach to resolving issues as they surfaced. These are the points where many large public-sector programs slow down. In Arizona, they were actively managed. The result was not just a system that was built, but a program that reached implementation.

could build a new IT system largely from scratch or you could transfer from another state that had already made such an effort. Since transfer ring was generally cheaper and less risky, some of those systems were transferred many times, passing on enhancements along the way. For instance, the Connecticut LINK system, implemented in 1996, was transferred and re-transferred to other states at least eight times, and several of those implementations have been the focus of CCWIS replace ment projects. Even in those simpler times, IT implementation failures occurred, but today’s technology options are much more complex. The federal Children’s Bureau’s August 2023 “Technical Bulletin #10: Low Code Solutions,” does well recognizing the complexity that “low code” solutions introduced, particularly customer relationship management (CRM) solutions, such as Salesforce and Dynamics. 2 As the bulletin notes, “low code solutions enable the faster delivery of business applications through minimal manual software coding,” but this isn’t guar anteed to result in lower costs and complexity. The configurability of these systems is not infinite and code changes can come with major complications—or, in the words of the bulletin, “the additional cost and complexity of modifying the

OOTB [out-of-the-box] functionality negates the benefits.” State adminis trators can’t be reasonably expected to know the exact functionality and configurability of low code solutions and know exactly where changes may be needed, but failure to understand these factors has been a major cause of CCWIS project failure and delay. Another option is to select one of several tailor-built CCWIS products available (including CRMs that have been heavily pre-customized for child welfare administration), but that still requires a detailed, and even unrea sonable, understanding of how well each product will meet business needs, which is also very difficult to do with limited information—so many options, and ever-changing technology and program needs. Ultimately, a successful CCWIS project will rely on some degree of accepting that the new system will not meet every requirement or wish. Knowing which requirements are truly non-negotiable, and then having readi ness to adapt and accept, are essential. Other Completed States Show Similar Patterns Idaho and West Virginia’s efforts reinforce a similar set of lessons. While each state approached implementa tion differently, they shared an ability to make decisions, maintain forward progress, and adjust when needed.

Nathan Grossman is a Manager/ Portfolio Lead at Public Consulting Group (PCG).

Policy & Practice Fall 2026 34