De-risking custom technology projects
A playbook for state budgeting and oversight
August 5, 2019
18F, Technology Transformation Service, General Services Administration
Note: This is a lightly updated version of the original document, which was created by 18F.
Only 13% of large government software projects are successful.1 State IT projects, in particular, are often challenged because states lack basic knowledge about modern software development, relying on outdated procurement processes.
State governments are increasingly reliant on modern software and hardware to implement federal programs and deliver essential services to the public, and the success of any major policy initiative depends on the success of the underlying software infrastructure. Government agencies all confront similar challenges, facing budget and staffing constraints while struggling to modernize legacy technology systems that are out-of-date, inflexible, expensive, and ineffective.2 Government officials and agencies often rely on the same legacy processes that led to problems in the first place.
The public deserves a government that provides the same world-class technology they get from the commercial marketplace. Trust in government depends on it.
This playbook is designed for executives, budget specialists, legislators, and other "non-technical" decision-makers who fund or oversee state government technology projects. It can help you set these projects up for success by asking the right questions, identifying the right outcomes, and equally important, empowering you with a basic knowledge of the fundamental principles of modern software design.
This playbook also gives you the tools you need to start tackling related problems like:
- The need to use, maintain, and modernize legacy systems simultaneously
- Lock-in from legacy commercial arrangements
- Siloed organizations and risk-averse cultures
- Long budget cycles that don’t always match modern software design practices
- Security threats
- Hiring, staffing, and other resource constraints
This is written specifically for procurement of custom software, but it’s important to recognize that commercial off-the-shelf software (COTS) is often custom and Software as a Service (SaaS) often requires custom code. Once any customization is made, the bulk of this advice in this playbook applies to these commercial offerings. (See "Beware the customized commercial software trap" for details.)
As government leaders, we must be good stewards of public money by demanding easy-to-use, cost-effective, sustainable digital tools for use by the public and civil servants. This playbook will help you do just that.
About the authors
We work for 18F, part of the Technology Transformation Services team at the General Services Administration (GSA). Collectively, the three of us have many years of experience in government procurement, software development, and state-level elective office.
In work funded by GSA’s 10x, we spent a year meeting with state legislators, legislative fiscal staff, state budget officers, contracting officers, and gubernatorial policy advisors. This playbook came of learning from and teaching hundreds of people from dozens of states. We’re grateful to the many people who contributed their time and knowledge throughout that process.
1. Projects valued at $6M or greater, in Europe and the United States, that were completed satisfactorily, on time, and within budget. From The Standish Group’s "Haze," based on their CHAOS database. ↩︎
2. Of the $90 billion in federal IT spending in FY2019, 80% is allocated for maintenance of legacy software, according to the GAO’s June 2019 report, "Agencies Need to Develop Modernization Plans for Critical Legacy Systems." They write that inadequately-maintained legacy software leads to security risks, unmet mission needs, staffing issues, and increased costs. ↩︎
3. The National Association of State Budget Officers’ “2018 State Expenditure Report” finds that federal grants provide an average of 31.2% of each state’s budget, climbing by 5.7% in FY2018. ↩︎