What is valuable about your organization?
I recently wrote an op-ed about the use of technology in the government, specifically the Intelligence Community (here). My thesis was that government organizations are not particularly good at designing or acquiring modern software systems. There are dozens of reasons why and I get into a few such as design by committee, contracting process bureaucracy, and the loss of integral government staff on large projects that take years to build or certify for use. Had I been a better writer, my essay would have looked a lot like this. Hana Schank and Sara Hudson published a piece in the Washington Post on Sunday, 21 January that does a better job explaining why the governments struggles so much in this area. Most importantly, they were able to tie the issue to the recent nightmare that Hawaii went through, believing a nuclear missile was headed for the islands. In this day and age, technology REALLY matters.
If you are the US Intelligence Community, your sweet spot is the top 3 boxes: Unique Expertise, Proprietary or Unique Data and Leadership. ... It's what you do best.
I won't rehash all of the arguments here but I will introduce an idea for thinking through how an organization should decide what work it is or is not going to do. What work should be given to an outside expert and how do you evaluate whether they are the right provider? It is conventional wisdom that a company or enterprise should do what it does best and not dilute itself with non-value-added work. If you are awesome at building rocket engines you should do that. Even though you could probably build a car engine as well, and you need cars to get out to the test stands, you probably should not take time out to design and build a car engine. You should stop, ask yourself who is the best at building cars and engines and if they can give you one for a reasonable price (i.e. substantially less than what it would take you to build it.) you should just go buy it from them. You also should probably not ask another company that you work with that is good at bending aluminum let's say, to build you a car, just because you have a comfortable relationship with that company. Specialization is the driving force behind innovation and progress.
Clearly this is a ridiculous example but it happens every day in the government and lots of private sector organizations as well. Many of you will recognize the colloquialism "not invented here" (NIH). The toxic and often wasteful corporate characteristic that implies that if you want something done, you should do it yourself. I used to work at Lucent with lots of Bell Labs PhD's running around. I would go as far to say that AT&T and later Lucent believe they invented the NIH syndrome. We always thought we could be more special if we invented and built everything in-house and in some cases, this was probably correct. When you have the patent machine that was Bell Labs, its an easy trap to fall into. But in fact, I would argue that this led to enormous wastes of time and money and distracted us from doing what did best (making optical fiber in my case).
When the government or any organization is thinking about how to acquire technology, specifically software, they should focus on what makes their enterprise special. I offer one way to think about it in the pyramid below. Various boxes can slide up or down depending on the type of organization you're in but as you move up the pyramid, you have an increasingly stronger argument for doing things in-house. It's your special sauce. As you move down the pyramid you should be looking to others who specialize in this area to satisfy your requirements. If you are the US Intelligence Community, your sweet spot is the top 3 boxes: Unique Expertise, Proprietary or Unique Data and Leadership. You should focus on these 3 areas because only you can do these. They can't be outsourced to another best-of-breed organization. It's what you do best. For the other boxes, you should find providers that have that function at the top of their pyramid. If you go look at the company that bends aluminum, you won't find making cars at the top of their pyramid and that means they shouldn't be doing it for you. As Schank and Hudson pointed out, maybe you don't get a company that has fighter planes and missiles at the top of their pyramid to build your web application.
