I remember when I was at AN AGENCY in 2003 and I worked to develop a good navigation to get to the content. Then I worked to make an alternative to that navigation that was called a HyperFAQ. I got to the end of that project and I realized that I had made two separate methods to get people to the right content, but each was developed with the logic and biases that I had. I woke up one night, like from a nightmare (because, yes, I am that geeky) and I thought, “What about the people out there who don’t think like me?” My point is that we technologists are in pursuit of connecting people to the data and processes as efficiently as possible. My goal is shared by most people, to reduce the friction to get what you want.

That is part of what Yegge is referring to in his rant. If you are in one Google product, like Google+ and you want to leverage something from another, like YouTube or Gmail, the connections didn’t exist to bring them together seamlessly. That is what we mean when we say that the whole is greater than the sum of its parts. And listen to me, brothers and sisters, we have a lot of friggin parts. Go to the IT Dashboard and look at the major investments. We have about 850 major investments listed there and there is a whole bunch of shadow IT that never shows up on the Dashboard.

Some would argue that Google has actually learned the lesson that Steve Yegge was ranting about. My evidence of this is that there is now one place that a person can go to see (a portion of) the data Google has on him or her. Go to https://myaccount.google.com/privacy. The fact that you can see and, to a certain extent, control this data proves that Google is working to be much more integrated than they were before. This is the evidence from Google, which seems to have been late to the integration party; imagine the extent of integration from Amazon, Microsoft, and Facebook, the companies that Yegge claims were way ahead of them in 2011.

When I was at ANOTHER AGENCY, we jumped into this pool. I am a lazy person. I can’t even believe that I am taking the time to write this book. Apparently being pissed off by the repetition of collective mistakes is overcoming my innate laziness. We’ll see if I actually get to Chapter 15. But when I was at the AGENCY I had a bunch of development projects. I didn’t want to waste my time developing stupid shit like authentication and backup and recovery capabilities for my applications. My thinking was, why the frick do I have to develop this stuff when someone else has already done it? At the time the department had just rolled out a new capability called eAuth. Instead of building my own authentication service, I used that one. It cost $50,000, but we ordered it in such a way that all of my applications could use that service to authenticate without an additional charge.

Next, we were in the position of buying metal every time we introduced a new application or capability. Buying metal means that I had to buy a server for every new application I wanted to develop. I call this the server-hugger mentality. I remember when I was working to develop a program monitoring capability for a grant program the infrastructure guy came to me with a bill to buy a new rack of servers. I was like WTF are you talking about? I’m building a capability that, at its peak will have 20 users per day, why do I need a rack of servers? Our office was dominated by one really big program that every American has heard of. We had a big system that supports that program. We worked within our little bureau CIO shop to abstract the application from the infrastructure. This allowed me to park my applications on that infrastructure and leverage all of the services that were already existing. Thus backup, reporting, and a bunch of other services were available to me without having to lift a finger. Also that authentication service that I got from the department was made available on the larger system, so the big program could use it too. I’m going to get into this in a lot more detail in the next chapter.

The important thing to keep in mind, and I’ll probably say it 17 times during this book, locally oriented decision-making is short-sighted thinking. My bureau is only thinking about our processes and functions. My little office in my previous department was only thinking about what we needed. It is the departmental CIO’s job to think local, but act global. When you have an opportunity to meet local requirements by thinking globally, you should always do it. I could have bought my rack and done what everyone else before me did and owned my whole environment and not have to worry about someone else jacking up my stuff. But I would probably still be doing it, nine years later. The problem is that I have a short attention span; oh, and I am fricking lazy. So push everything to enterprise thinking.

What we did, organically, or just out of laziness, was create a platform. Now the services that were available to the application that managed the big program were available to my grant monitoring application. Collectively our maintenance is less and our infrastructure is better leveraged. There may never be an opportunity to mashup the data from these two applications, but by creating a center of gravity, other applications came, and that presented an opportunity to integrate the data to deliver more value.

Topic-EA_slide1_lg four operating models
10

If you take what I was doing at my bureau and apply that more holistically across the government, you can see that where we are going is away from the Coordination, Replication, and Diversification models and toward the Unification Model.

We can do it like Jeff Bezos did at Amazon, and mandate it with an iron fist and a bunch of memos from OMB. There are instances in which some people have been bold enough to do it. Chuck Christopherson was both the CFO and CIO during his time at USDA. He made some very bold decisions, like consolidating their email down to one and collapsing all their data centers down to one. He had the power both from a budgeting perspective and an IT perspective and he used it effectively. Of course, he didn’t make any friends, but we are here to move the government forward, not make friends. It would be more effective if agencies understand the opportunity and came to these conclusions for themselves.

The litmus test to see if the government is making headway in platforms versus applications would be to see if there is a service like that privacy service from Google that lets you see all the data they have on you. Well, shouldn’t the government have a service that allows you to see all the data that the government has on you, regardless of agency? Anyone who has ever gotten a Social Security card, paid their taxes, or gotten a loan to go to college should be able to go to one place to see all of the data that the Social Security Administration, the IRS, and the Department of Education have on them. We will know that we are integrated when each of these and the thousands of other services that the government provides are available for query, just like Google. And I don’t want to just see the information that the government has on me, I want to take an active role in improving it. If it is wrong, I want to be able to tell someone that it is wrong and I want to hold them accountable to fixing it.

So Enterprise Architecture is the first place to start on this tour of stuff to fix. It is probably the most difficult part to fix because it requires people to think globally and that is hard because the tendency is to think that you are taking on additional risk.

 

In This Series:

The Federal IT Papers–Part 1

The Federal IT Papers–Part 2

The Federal IT Papers–Part 3

The Federal IT Papers–Part 4

The Federal IT Papers–Part 5

The Federal IT Papers–Part 6

The Federal IT Papers–Part 7

The Federal IT Papers–Part 8

The Federal IT Papers–Part 9

The Federal IT Papers–Part 10

The Federal IT Papers–Part 11

The Federal IT Papers–Part 12

1 Lewis Carroll from Alice in Wonderland

2 Page 4 http://www.gpo.gov/fdsys/pkg/PLAW-107publ347/pdf/PLAW-107publ347.pdf

3 VA EA, 2002, https://commons.wikimedia.org/wiki/File:Simplification_Zachman_Enterprise_Framework.jpg – you have to go to the Wayback Machine to find this PowerPoint

4 Ross, Enterprise Architecture as Strategy, page 29

5 From https://spaces.internet2.edu/download/attachments/35980201/Delta.png?version=1&modificationDate=1358346473993&api=v2

6 From http://cisr.mit.edu/files/2009/12/Topic-EA_slide3_lg.png

7 Section 3504 of Title 44 of the US Code

8 NARA IRM Strategic Plan page 52 https://www.archives.gov/about/plans-reports/info-resources/nara-irm-strategic-plan-2015.pdf

9 Education Strategic Plan, page 30 – https://www2.ed.gov/about/reports/annual/ocio/irmstratplan2015.pdf

10 Jeanne Ross – http://cisr.mit.edu/files/2009/12/Topic-EA_slide1_lg.png

 

Read More About
About
Demosthenes
Demosthenes
Demosthenes is a pseudonym for a senior Federal IT official.
Tags