This two-year grant responded to 2003 Priority 4: "Expanding Use of Automation Tools and Best Practices." The Project was designed to utilize Web site technology to increase customer services by providing parents, employers and case workers the ability to access case data and input data on-line. Colorado CSE intended to launch an interactive, e-commerce Web site that would provide 24-hour access to case status and payment information for parents, both in-state and interstate caseworkers, and employers. It was anticipated that by doing so, Colorado CSE would see significant improvement in its ability to provide better and more current information to its key stakeholders.

By the end of the grant period, this purpose was realized by providing real-time case and financial information to Web-registered parents with child support cases and registered workers from other States.

Project Findings (From the Final Report):

The project met the following three goals:

1. Rapidly meet common user needs: With the child support information and case-specific detail now available, many common user needs have been met. At the end of the grant period there were over 16,000 registered users on the Web site. Registered users are made up of about 78 percent custodial parents, 22 percent noncustodial parents and 2 percent out-of-State IV-D workers. It is also interesting to note that based on a recent analysis, 15.5 percent of the cases are for registrants residing outside of Colorado.

   Finding: Publications and forms are being accessed regularly. For example, the Colorado general CSE brochure was accessed an average of 950 times per month for the period of October-December, 2004. The application for Direct Deposit of child support payment was accessed an average of 1450 times per month for the same period. In December 2004 there were over 3,600 user requests for a printable version of case payment histories. Previously, all of these could have resulted in a request to a Customer Service agent to print and mail these materials.

   Moreover, registered parents in general have made extensive use of the payment detail and payment history pages on the Web site. In December, 2004 there were over 29,000 hits on the payment history page.

2. Answer frequently asked questions: On the Web site, for example, there is a section for Frequently Asked Questions, with answers to 19 different topic areas. The FAQs cover all of the primary areas of child support enforcement in Colorado.

   Findings: This has proven to be a great resource for information and it is a common occurrence for the CSE receptionist to direct callers to the Web site when they call asking questions about the program. This redirection has saved valuable staff time and has allowed for consistent answers to the questions asked.

3. Demonstrate proof-of-concept: The project was successful in meeting user needs.

   Findings: With visits to the site averaging 2,250 per weekday, clearly the Colorado CSE Program has gone well beyond a 'proof-of-concept' scenario to a well-used production CSE Web application. A check on the volume of hits asking solely for specific case financial history for the period of October-December, 2004 shows an average of over 29,000 hits per month. Analysis shortly after the project ended revealed this number growing by more than 10 percent per month.

Lessons Learned:

• The project did not see an increase/decrease in customer support calls to the SDU or in the volume of calls to the IVR for use in comparing "before implementation" versus "after implementation" of the customer service Web site. However, while the IVR inquiries were relatively flat, information requests continued to rise on the Web site.
• County workers did not believe that the Web site had a noticeable impact on their workloads. This could be due to the fact that county workers in general often feel overwhelmed in keeping up with their workloads, and the most difficult issues they deal with cannot be solved by the Web. However, one major accomplishment has involved putting procedures and desk aids on the Web site for in-state users to access. It is believed that in the long-term this will be a huge benefit to county caseworkers as it will allow for quick turnaround of changes in the procedures and/or just in time procedural information.
• Open source software is a viable alternative for Web software. It not only comes at a low cost, but is also reliable, stable, scalable, and performs well. Having begun this project using Oracle Web software, the Colorado project team has a good understanding of the benefits of open source software over proprietary software packages. This software has also been successfully implemented by Arizona.
• An executive steering committee is very important to the overall success of the project. Another group that made a huge difference in the outcome of the project was the application Web workgroup. This was a group of CSE staff made up of members from each section in the Colorado Division of CSE. This group was instrumental in helping to produce an end result that represented the child support program and worked well for the public.
• After the initial prototype was built, Colorado conducted a user usability test. There were four user testers participating. These users were taken through a scripted test and the sessions were videotaped. The usability test proved to be invaluable. Colorado would recommend this process to anyone building a new Web application.

• The approach used to pull together content about the Colorado child support program was the following: existing documentation on the child support program was reviewed and the primary topic areas were determined. Various topic areas were divided among small sub-groups of the Web workgroup and each group was asked to put together content for each topic. Sub-groups were encouraged to use existing material they could find, but were also charged with the task of limiting the size of each explanation for the Web. This content was then compiled together into one document where it was re-worked by the project manager in an effort to give it one voice.

  This was a difficult task, and the success of finding one voice for all of the different topics could be questioned. It would probably have been better to hire an outside content editor, if resources were available. This approach did allow Colorado to gather a large amount of content in a relatively short period of time. Colorado also used the same approach in compiling the 'Frequently Asked Questions' for the site.

• A very simple user site survey was made available for the first 12 months of production. This proved to be invaluable in getting user feedback to determine what worked well and what did not. It also gave insights into additional features that users wanted on the site. Even a simple survey asking general questions about navigation on the site and allowing for users' comments produced a wealth of information, particularly during the first 6 months.
• One of the biggest challenges faced was when registrations for the site began growing faster than expected. This increase often came from issues that ultimately had a good result. The registration process required that the user give their social security number (SSN) which then had to be matched to the SSN kept in the primary Child Support System (ACSES). When people could not register because Colorado did not have their SSN entered into ACSES they often called and their SSN was than entered into the system. This was a big plus, as the SSN is the key to many functions, such as locating individuals. Colorado also had numerous instances where they mailed the user's password by UPS mail (Colorado chose this route for security reasons) and it was returned because Colorado had the wrong address in ACSES. Once again, Colorado was able to contact these people and get a current address in the system. Both of these scenarios created work, but resulted in better information in the system.
• Colorado also learned that there is significant interest in real-time access to Colorado State case and financial information by other State CSE workers. By the end of the grant period Colorado had over 300 other State workers registered for this access.
• As time progressed, there was a need to appoint a technical resource person to be available for people with registration and log-in issues that are primarily technical in nature. This technical resource person came from existing staff. This responsibility has been rotated to various members of the Web team from month to month.

(NOTE: Includes Findings from the Final Report)

Project Plan

This grant responded to 2003 Priority Area 4: Expanding Customer Service through Agency-Initiated Contact. A large percentage of delays in children receiving their child support were due to missed appointments, forgotten hearings, and delinquent support payments. This project planned to use automated dialing as a non-confrontational means of reminding both custodial and non-custodial parents of upcoming appointments and payments due without the involvement of CSE staff. The project sought to reduce the number of missed appointments and delinquency payments, and increase payments due on support orders and the rate of collections among the approximately 44,000 clients served in the State.

The project used a "PhoneTree 3500" automated dialing product based out of Winston-Salem, North Carolina. The automated dialing system has three components: 1) Non-custodial parents were called and reminded of upcoming appointments with the child support agency and court hearings up to three days in advance; 2) Non-custodial parents were called with a reminder that the first payment will be due on new child support order, and on an initial delinquency on a new order; and 3) Custodial parents were called and reminded of upcoming appointments with the child support agency.

Project Findings

The project monitored call volume through the use of North Carolina's Data Warehouse System, and showed an average of 1,877 calls per day or 59,345 calls per month. Between August 2003 and November 2004, statewide appointment attendance showed an increase from 56% to 61%. Statewide hearing attendance rose slightly, from 85% to 87%, for the same period. Between August 2003 and December 2004, delinquencies on new orders that were resolved within 30 days rose from 31% to 38%. The project also received positive feedback from local offices regarding increased court appearances by non-custodial parents.

The project did not evaluate cost-effectiveness; however, the automated dialing product's cost was exceeded by the increased amount of collections brought in. The software cost only approximately $16,000, yet there were unforeseen increased costs to the State due to the need for additional staff at the customer service center to answer follow-up phone calls generated by the automated calls.

The project plans to continue the established automated dialing protocol, and is proposing the following enhancements: 1) Calls to all non-custodial parents with cases in delinquency/enforcement status for more than 60 days; 2) Calls to all non-custodial parents with cases in delinquency/locate status for more than 60 days; and 3) Calls to custodial parents with cases in locate status for more than 60 days to obtain updated information about the non-custodial parent.

Lessons Learned

Best practices and lessons learned by the project include: prepare the database with as many updated and missing phone numbers as possible; have procedures in place to correct incorrect phone numbers in the database; eliminate unnecessary data from the batch files that are downloaded to the automated dialer, i.e., social security number and date of birth are not needed for reminder calls; save data for six months on CDs.

Note: The May 2004 Child Support Report contains an article on the NC Automated System - http://www.acf.hhs.gov/programs/cse/pubs/2004/csr/csr0406.html