<aside> 💡

The program design stage is about taking your high-level vision and the results of your field exploration and turning these into an actionable roadmap. Remember that the goal here is not defining the solution but defining the problem(s), framing the solution and then structuring a program to achieve your goals. Program designs should be a guide for the program, but should also leave plenty of room for adaptation.

</aside>

⚡Practical outcomes

• A full Heilmeier document (or similar) with ROM budget, timeline, evaluation plan, etc. that has been iterated multiple times with colleagues and direct managers
• A full pitch deck ready for your program approval meeting (that has been road-tested multiple times with colleagues and direct managers)
• Specific examples of key potential performers and methods in each technical area
• A draft of your transition strategy/options (at a high-level)
• A road-tested argument for why this won’t happen without your program

Many aspects of effective program design require good strategic skills as well as scientific intuition. You need to assess whether the technical tasks are doable, who is up to the challenge, and whether all elements of the project can come together cohesively. Understanding incentive structures, clarifying roles, and ensuring alignment between performers are all tasks that fall squarely on the PM's shoulders. While scientific challenges can sometimes be difficult to predict, management problems are often within your control. Careful and thorough program design helps you stay ahead of these challenges.

In the design phase you will continue to iterate on your Big Four HQs and answer “the other four” as well:

• What are the risks?
  • This should include the technical blockers you’ve already identified, transition/adoption challenges, program structure risks, etc.
• How much will it cost?
  • We’ll cover this later, for now just calibrate to something like more than $10M but less than ~$100M.
• How long will it take?
  • What is time? Use this prompt to think about sequencing, phase structure and intermediate milestones.
• What are the midterm and final "exams" to check for success?
  • Metrics and evaluation are essential to a good program. The best way to approach this is not “how do I prove that my program is working” but “how do I design evaluations and metrics to help drive towards the outcomes I care about.” This is harder than it sounds in most cases…

🔮Planning for Problems

In DARPA-style programs, the key performers are spread across separate organizations. As the PM, you are often the sole unifying force tying these performers together into a functional team. In many cases, you won’t just be an orchestrator; you might need to take a strong leadership role.

In a perfect world, the performers’ skills and incentives would fit seamlessly with your program’s objectives. You could focus on ensuring they have the resources to succeed and that they collaborate well with end users. However, this is rarely the case. As the program progresses, you may discover that a performer’s incentives don’t align with the program’s success, or that they are simply underperforming. While you’d ideally uncover these issues during the selection process, some challenges won’t become apparent until later.

This is where effective program design comes into play. Planning for problems from the start can help prevent these issues from derailing your entire program. For instance, you might establish contingency plans, set up intermediate checkpoints to evaluate progress, or build in mechanisms for changing course if needed. Adequate program design can prevent management and contracting issues from sinking your program.