Project managers with a forty-person team and a two-hundred-line schedule: the fastest project teams in history were small enough to fit in one room. That wasn’t a budget constraint they worked around. It was the strategy.
This is the second post in our Elite Teams series. The first covered the trait every elite team shares. This one covers how they’re shaped: small, fast, and built to learn from each cycle. It happens to be the definition of agile project management, and the best examples of it were running long before anyone wrote the word “agile” on a whiteboard.
Unleash Your A-Game: Building an All-Star Project Team (and Becoming an A Player Yourself)
143 days
In the summer of 1943, the U.S. Army Air Forces asked Lockheed to build a jet fighter. Germany already had one flying, and the schedule was measured in months, not years. Lockheed handed the job to an engineer named Kelly Johnson, who set up a walled-off group in a rented circus tent next to a foul-smelling plastics factory. The team named itself after a comic-strip moonshine still. The name stuck: Skunk Works.
Johnson’s team delivered the XP-80 Shooting Star, America’s first operational jet fighter, in 143 days. The same group later produced the U-2 spy plane and the SR-71 Blackbird, which still holds speed records more than sixty years after it first flew. Johnson eventually wrote down how he did it in fourteen rules. Read them today and they look like an agile manifesto written fifty-eight years early:
- The project manager has complete control of the program, with authority to make decisions quickly.
- The number of people connected to the project is kept viciously small, ten to twenty-five percent of what a “normal” program would use, but they must be good people.
- Reporting is minimal, but the work that matters is recorded thoroughly.
- The team and the customer trust each other and talk directly, without layers of liaison in between.
- Funding arrives on time, so the team never stalls waiting for money.
Keep the number of people connected to the project small, and use only good people. Kelly Johnson wrote that rule in the 1940s. The Agile Manifesto arrived in 2001 and said the same thing in different words.
Every one of those rules maps to a line in the 2001 manifesto. Individuals and interactions over processes and tools. Working product over comprehensive documentation. Customer collaboration over contract negotiation. Responding to change over following a plan. Johnson wasn’t doing software. He was doing project management under a deadline that mattered, and he arrived at the same answers.
About two months
On January 11, 2020, Chinese researchers published the genetic sequence of a new coronavirus. Within days, a small team at Moderna had designed a vaccine candidate. On March 16, 2020, the first human volunteer received a dose in a Phase 1 trial. Depending on which milestone you count from, that’s somewhere between six and ten weeks from sequence to first injection. The historical average for a new vaccine was closer to ten years.
Pfizer and BioNTech ran a program they called Project Lightspeed, and reached U.S. emergency authorization on December 11, 2020, about eleven months after starting. Both companies did it the same way, and the way has a name in the PMP® vocabulary: fast-tracking. Phases that normally run one after another, Phase 1 then Phase 2 then Phase 3 then manufacturing scale-up, were run in parallel. Factories were built and doses were made before anyone knew whether the vaccine worked. If it had failed, the manufacturing investment would have been written off.
That’s the part people miss when they call these programs “fast.” They weren’t reckless. Every overlap was a risk the team named, sized, and chose to accept, with the money to cover the downside. Fast-tracking is a schedule-compression technique on the exam because it works, and it works only when the team is honest about what it’s betting.
Iterate on hardware
Software teams take short build-test-learn loops for granted. It’s harder when the product is a rocket. Between 2006 and 2008, SpaceX launched its first vehicle, Falcon 1, three times, and all three failed. The company had funding for roughly one more attempt. On September 28, 2008, the fourth flight reached orbit. Each failure had been treated the way Etsy treats an outage: find the cause, fix it, fly again as soon as the fix is in, and don’t spend six months on a design review when a launch will give you better data.
The same approach shaped what came next. In 2014, NASA awarded fixed-price contracts to two companies to fly astronauts to the space station, one for $4.2 billion and SpaceX for $2.6 billion. SpaceX flew its first crew in May 2020. The other provider’s crewed test flight in 2024 ended with the capsule returning empty and its crew coming home on a SpaceX vehicle the following spring. Same customer, same requirement, and the program with the smaller budget closed the gap first. Short loops on hardware beat long loops on paper when the unknowns are physical.
The 2010 Chilean mine rescue is the version of this story with a whiteboard-simple lesson. Thirty-three miners were trapped 700 meters underground. The rescue team didn’t pick the best drilling plan and commit to it. They ran three plans in parallel, each with different equipment and a different risk profile, and let the ground decide. Plan B broke through first, on day 66. All thirty-three came out on day 69. Three plans running at once is what a risk response looks like when the stakes are human.
The agile project management rules hiding inside these stories
Strip the aircraft, the vaccines, and the rockets away and the same six rules are left. They’re the ones the PMP exam expects you to recognize in a situational question, and they’re the ones your next project can use on Monday:
- Cap the team. Amazon made the “two-pizza rule” famous: if two pizzas can’t feed the team, it’s too big. Johnson said the same thing in the 1940s. Every extra person adds communication paths faster than they add output.
- One decision-maker with real authority. Not a committee, not a steering group that meets monthly. Someone who can say yes today. If your project has to wait two weeks for a decision, the schedule is a fiction.
- Overlap phases on purpose, and name the bet. Fast-tracking is fine. Fast-tracking without writing down the risk you’re accepting is how projects surprise their sponsors.
- Report less, record more. Johnson’s rule in one line. Cut the status reports nobody reads. Keep a rigorous record of decisions, assumptions, and test results, because that’s what the next cycle learns from.
- Ship something testable every cycle. A prototype, a demo, a launch, a dose. If a cycle ends with only documents, it wasn’t a cycle.
- Colocate, or fake it well. Skunk Works put everyone under one roof. A distributed team can get most of the effect with one shared channel and a hard norm on response time. What kills speed is the wait, not the distance.
Master of Project Academy Blog – FAQ: Go‑to‑Market Strategy for 2026
Where hybrid fits
None of these teams were purely agile in the software sense. Skunk Works had a fixed deliverable and a hard date. The vaccine programs answered to regulators who required sequential evidence. The mine rescue had a plan A, B, and C with predictive schedules for each. What made them elite was choosing the approach that fit the work, phase by phase, instead of picking a methodology and forcing the project into it. PMI calls that hybrid, and it’s where most real projects live.
The PMP exam tests predictive, agile, and hybrid approaches, and situational questions routinely ask which one fits a described scenario. PMI’s July 2026 update to the exam expanded business-context content without dropping approach selection.
Next in the series: elite teams adopt the tool first. What Formula 1 pit walls, a 360,000-hour legal review, and a public reversal at Klarna say about AI in project management.
Learn all three languages
The PM who can only run a Gantt chart gets handed the predictive projects. The one who can only run sprints gets handed the software. The elite PM speaks predictive, agile, and hybrid, picks the right one for the phase in front of them, and can build the plan in the tool the sponsor actually uses.
The 9-Course Project Management Training BundleOne bundle covers the full toolkit: PMP® exam preparation, Agile Scrum, and Microsoft Project, so you can prepare for the certification and run the project the way the work demands.
|