Modernization report comments

[craigschneider]

Respectfully submitted by Excella Consulting.

1. What are major attributes that are missing from the targeted vision? a. The report promotes the use of agile development practices as key strategy for modernizing government software applications. Agile development should absolutely be used by the Federal government to build and maintain government software solutions, but procurement rules and regulations make this a difficult sell for many agencies. Procurement officials routinely demand defined requirements and measurable deliverables when soliciting for bids for software development, but agile software delivery is defined by identifying the business problem and desired outcome and adjusting requirements through learning and discovery. To promote agile procurement, increased usage of the 8(a) Program Digital Service Initiative is a great start. We also encourage the administration to fully utilize the Joint Venture Partner program at the National Technical Information Service (NTIS). Does the administration plan to adapt procurement policies to simplify the solicitation for and execution of software delivery contracts that comports to the needs of and agile software delivery model?
b. The report references the term “legacy systems” but doesn’t define the term. Do you intend to define what is considered a “legacy system” in this report or is this simply a placeholder for older agency systems whose meaning can be defined by agency policy? c. Many agencies and their technology decision-makers like the idea of moving software to the commercial cloud, but the rules, policies, and ramifications surrounding these decisions are not always easy to negotiate as noted in the report. Difficult leadership decisions nearly always default to inertia and prevent the desired outcomes this strategy seeks. Technology leaders need clearer definitions about what types of applications can be moved to the commercial cloud, what types of public and private data can be stored in the commercial cloud, and what needs to be done to certify the security of these cloud solutions would make these decisions much easier for system owners. Is there anything more the administration can do to make these decisions simpler and more clear cut for agency decision makers?

4. Are there any missing or extraneous tasks in the plan for implementing shared services to enable future network architectures? The authors of this report are spot on in their desire to push the adoption of shared services, where possible, across the federal government to achieve positive outcomes included in the report. Shared service solutions, especially those built by and for government, are much more adaptable to changing legal and policy requirements and can more effectively adapt to the changing unique needs of the Federal workforce. As noted in the report, the use of shared services has long been a strategy of the Executive branch, but adoption has not come as fast or as easy as desired. Does the administration have any plans to address the issues noted below to improve execution and adoption of existing shared service solutions?

• Make it easier for agencies to buy into shared service solutions. Requirements to compete shared services in the market place can make their use a more difficult sell to decision makers. We should make it simple for agencies to buy these solutions if, in fact, this is what the administration wants them to buy.
• Make it easier and fairer for small agencies to buy into shared service solutions. Large agencies can negotiate larger discounts for shared service solutions, leaving small agencies to pay a larger amount of money for often lesser service. By finding a more equitable way to distribute the costs of shared services, smaller agencies won’t feel like they are paying more for less.
• Similarly, make it easier for government-provided shared services to collect payments from agencies. Many shared services operate in a cost recovery world, yet customer agencies can be slow in making payments, leaving the shared service provider without the funds needed to operate the service effectively. Anticipating costs for shared services and withholding costs at the OMB level for shared services could ensure timely cash flow for shared service operators and further simplify the procurement processes for agencies.
• Finally, OMB needs to provide a more concise and targeted policy communication to force agencies to use shared service solutions where possible and to incentivize agency decision makers to choose shared services where it’s not possible to mandate. Agencies are given significant control over these decisions, and they routinely select non-shared service offerings for questionable reasons. Creating a set of policies that make shared services the easy answer will get a lot more agencies taking shared solutions seriously, without the need for a larger stick that OMB can and should choose to wield when necessary.

[johnaweiler]

Craig makes some great points. However, we need to distinguish between "make biased" agile development and "buy/adopt biased" agile acquisition. The federal govt is the worlds largest BUYER of IT capabilities, but the worst performing sector. As it is not a major IT manufacturer (Google, IBM, Oracle, MS), it should follow commercial practices of other large consumers like Finance, Telecom, Healthcare, etc. Agile Acquisition is a relatively new approach that improves IT management and sourcing by framework three key elements of the IT SDLC; business/mission value, lifecycle cost and most importantly, RISK. There are many dimensions of risk; execution, quality, performance, and cyber vulnerabilities in code. Free Open Source S/W has the greatest risk if it is not commercial grade.