“The DoD was capable of picking winners and losers, and that’s what stopped in the 70s. A bunch of us in the startup world think that one of the big problems with these incubators and accelerators is that giving out $1 million [SBIR] grants is detrimental to getting effective weapon systems. No, you don’t want to hurt anybody’s feelings so everybody gets a prize for participation. The goal isn’t to show up on the soccer field, the goal is to win the game.”
That was Steve Blank on my recent podcast episode with him. Erik Jessen comments on LinkedIn:
I’ll point out that if the DoD can’t pick winners and losers, then it also won’t be very likely to pick the right mega-vendor on mega-contracts. So there’s a much bigger problem than SBIRs. Which is a concern. My personal observation: what has *also* happened since the 1970s is the huge growth in commercial engineering jobs, where there’s more respect for employees, more opportunities for innovation and personal growth, and far less bureaucracy. So MilAero companies, I’m told by the old-timers, had to switch to “fighting for the bottom third of engineers”. If your engineering staff starts going from top third to bottom third, you’re going to see failure and really slow progress all over.
Here’s my response:
Erik, thanks for another insight[ful] comment! A lot to break down there, I’ll give it a quick shot. (1) No doubt you’re right on mega-contracts. But those are conditioned by 5-10 years of requirements, detailed specifications, multi-layered approvals, etc. — in other words, rely on process > insight.
(2) That rigid/prescriptive process means [contractors] parrot back the technical/business volumes, and only differentiate/win on the thoroughness of the cost volume. Some have said their business is not production, but bidding.
(3) This process forces a “fishbowl” environment not just on govt PMs, but on the whole [contractor] workforce, which cannot contribute genuine creativity. By contrast, Dick Spivey at Bell in the 1960s right out of college was able to “flower” by developing his own tiltrotor designs and bugging pilots to test it out — inventing the “Whisper Tip” blade as a youth. Of course, the DoD eventually screwed up the reqts on the V-22 in the 1980s, nearly causing a revolt.
(4) The DoD is simply too big to “compete” with private world for talent. It has to be a source of building talent, giving them ways to use their knowledge, and boomerang to industry. Rickover’s a shining example. I think the DoD software factories today are trying that.
Erik replies:
Imagine if the DoD, instead of spending 5 years on studies, simply started paying for variants? And also paid to get rid of the Architectural Debt in existing products, so that more variants could be created both far faster and far cheaper? Coming from the commercial world, I’m all about the OODA loop.
The problem is (1) just because there’s a process, doesn’t mean it generates quality results, let alone optimal. In fact, the saying in process is that first you want predictable results (even if predictably crappy) so that you can then start to fix the process. What you’re describing is a process that both generates predictably bad results, AND does it consuming large amounts of time and money. But the good news, there IS a process to fix. 🙂
(2) creative workers don’t go to large organizations. It’s an axiom. Which is why large organizations have to create small R&D groups, “Skunk works”, etc. Highly-productive/creative people won’t put up with manual outdated processes. Which is another reason why DevOps is so critical – to attract and keep the employees you desperately need.
(3) Japanese cell phone companies changed to this approach to deal with an extremely rapidly changing market: they simply got rid of marketing. By the time a marketing plan was written, approved, executed and results collected, that market didn’t exist any more. So instead they took their highest-selling model, created 5 variants (bigger screen, more battery, better video, more storage, etc.) and just put them into production.
And I reply back:
Thanks for the reply. Definitely agree on #1. I didn’t mean to imply that the mega-contract process was effective, just that it is an inferior substitute for technical in-house knowledge.
I’m a bit agnostic on you’re #2. I tend to defer to Ben Horowitz who says that it’s not small vs. big firms, but young vs. old firms. I’d presume the FANGS can still attract talent despite their size bc they are still run by hungry founders. Another way I would frame it is this: creative/talented people have a lot of intangibles that don’t show up well on our current culture of credentials and resumes (at least in DC). They’re better off as entrepreneurs and at startups where they can have the freedom and the potential to earn rewards commensurate with their abilities.
#3 is interesting and I think relates to your agile/devops view in #2. Defense is different because there’s not really the same market structure. Rather, you have a monopsony buyer with rigid processes and its rather hard to actually get to a large number of diverse users and find the niches. The DoD, wrongly in my opinion, demands singular “economy of scale” solutions.
Erik responds:
On #2: every organization goes through the aging process. It’s been observed that the grandchildren of every group of Communist Party leaders pretty much act, talk and think like trust-fundistas – doesn’t matter if their grandpa was Mao or Fidel Castro. But it’s important that in the long term, if you don’t figure out how to cope with the long term, you’re dead. For commercial world, they use Skunk Works, R&D labs, etc. For DoD, SOCOM has been very impressive. The Marines have much more of an Agile culture than the Army, etc. So it can be done – just a question of leadership deciding that the pain is worth the gain.
#3: I think of international affairs as ‘business by other means’. Companies compete in the marketplace. Governments compete as well. Within each corporation, there are multiple divisions, each with their purchasing departments, etc. DoD, CIA, State all look like divisions within a large corporation. CIA is Business Intelligence, State is Sales&Marketing, etc.
And I write back:
Hallelujah for multi-buyer nature of govt. That’s how it used to be in the bureau/arsenal days, but the whole purpose of the PPBS was to stamp that out and create a monopsony. That’s why I push for budget reform, so that organizations can take differing perspectives as to technical or military feasibility, rather than presuming that can be settled before the fact in singular winner take all programs.
Leave a Reply