Skip to content

August 30, 2018

Digital Transformation: Thriving Through the Transition

By IT Revolution
Jeffrey Snover DOES18

The following is an excerpt from a presentation by Jeffrey Snover, Technical Fellow and the Chief Architect at Microsoft, titled “Digital Transformation: Thriving Through the Transition.

You can watch the video of the presentation, which was originally delivered at the 2018 DevOps Enterprise Summit in London.

Today I’m going to share with you three things.

  1. The role transitions play in your career
  2. The role transitions have played in my career, particularly around taking the Windows out of Windows server or command line interfaces and PowerShell
  3. This big transition that’s happening in the industry — digital transformation— and how this provides you an opportunity to supercharge your career

My goal is to help you get this dynamic in focus so that you can make some changes.

But first, I’m Jeffrey Snover.

I am a Technical Fellow at Microsoft. I am the Chief Architect for Azure Storage and our Cloud Edgework, I’m also the Chief Architect for Azure Stack, but most people know me as the inventor of PowerShell.

I work for Microsoft. Microsoft has about 120,000 employees. We have $90 billion in revenue, and we are in a race to become the first company with a trillion dollar market capitalization. So pretty big company.

There are two types of jobs

First, there are stair jobs.

Stair jobs, during normal periods, you climb up the stairs of your career slowly, there’s somebody in front of you, and then every now and again, things go wrong. But slowly and steadily, you advance in your career.

Then there were these periods of disruption, transitions, where everything changes, and it’s an opportunity for your career to advance, or to go the other direction. These periods are what I call the elevator jobs. With an elevator job, you can have a period of time where things line up nicely and you’re able to go up one, two, three levels of your career in a relatively short period of time. And I know that there are a number of you who’ve experienced this with your DevOps journey.

But transitions are tricky. They can really be the wind in your sail, or not. The key point is that during a transition, there are new winners and there are new losers. So you want to get this in focus because there’s a difference between those two, and you want to be the guy ahead.

Now, Since I’m a technical fellow…

I get to participate in the review of the performance and the reward conversations for a lot of senior people. And the conversations typically go like this, who are the people moving this business forward?

We identify those people and then we ask the next round of questions:

  • How are they doing?
  • Are they secure in the socket, or are they loose?
  • Are they happy?
  • How do we reward them?
  • How do we keep them in the boat?

And guess what? We compensate them, at Microsoft, we compensate them very well. But like everyone else in this room, we have a budget. When we take a bunch of money and we hand it to the people who are moving our business forward, it comes from somewhere, and it comes from the people who aren’t moving the business forward.

So you want to be one of those people moving the business forward.

Let’s take a shift and talk about transitions in my career.

How they have helped me, and sometimes, how they haven’t.

I started off as a Unix developer, for many, many years, and I went to work at Digital. And at Digital, I had a stair job. Unix was not critical to the success of Digital, and I was doing fine.

Then at some point, Digital got involved with NT. Microsoft invented NT and when Digital got involved I was all in. I was very thrilled about this technology.

So I went all in on this transition to NT. I read everything. I mean, there was nothing you’ve written about NT that I didn’t read, I evangelized, I went around the company telling people why this was important, what they needed to do, and helped them get their project started.

And at some point, we had one of the biggest NT projects the company had. We were partnering with IBM to port NetView to NT, and they tapped me on the shoulder and they said, “Hey, we’d like you to be the architect for that project.” I grabbed that and we just ran with it. It was fun— I worked my butt off. This was not a period in my life that I would call ‘work-life balance’ at all. But I was just having a blast. We won a number of awards, etc. And this was my elevator job, because in fairly quick order we released that, and then things worked out very well for me. I became a digital consulting engineer. This is a big deal.

Now, whereas that transition was doing very well for me, the industry was also going through a big transition. And the industry was transitioning from a vertical integration industry into a horizontal integration industry.

Digital did not survive this transition. In fact, it was Bill Gates who said, “Most people get the decline of deck all wrong. The reality is that Digital declined because they were excellent at something that no longer mattered, and therefore they died.”

Part of dying was that they sold things off, one by one by one. They took my division, the management division, and they sold us to computer associates. I met with Charles Wong and decided he was insane, and that I did not want to work for him. So instead, I cold-called their competitors, a company called Tivoli and I said, “Hey, why don’t you hire me? I’ll take my product and my team, and we’ll go compete against these guys.” And that’s what happened.

And so I worked at Tivoli for a number of years where I ran their network management development effort. That was very successful. It was going great until one day the CTO of Tivoli asked me out to lunch. We went to a barbecue place, we were having a nice conversation, and he said the following,

“Jeffrey, you’re one of my hardest working, smartest guys. You’re doing great, I love it. But here’s the issue. You’re working on a product that every time I sell it, I earn $15,000.” I said, “Yeah.”

He says, “You’re not working on the product that every time I sell it, I earned two to 20 million dollars. So I just have one question for you, Jeffrey. Do you want to be relevant?


I had what I think alcoholics call a moment of clarity because I was passionate about my technology. It was NT. I went and did this incredible thing, and won all these awards. I moved to Texas to work on my project, and here he’s telling me I’m not relevant because I’m not moving the business forward. In the end, I decided that, actually, impact mattered more to me than technology.

Eventually, I got on Bill Gate’s radar screen, and I got the call to go work for Microsoft as their lead architect for all their management technology and products.

At the time, Microsoft was heavily focused in on the GUI, GUI, GUI, GUI. If I can make it about the GUI, I win. So let’s make it about the GUI.

I was great, except I could see that there was going to be a transition to very large data centers, and a GUI centric world was not the right thing to do there. So, I wrote down the Monad Manifesto, which was the basis for Windows PowerShell. This was a fundamentally new approach to automation, and I was totally in. I was going to do it again. Now you’ve seen this org chart of Microsoft.

I was, as they say, an industry hire. Bill Gates sponsored hiring me, I came in at a high level. I decided I was going to be all in on this automation technology to take advantage of this transition and was demoted. Yes, I was. And that lasted five years.

So fine, I was demoted. I didn’t care so much because, again, I get very passionate about technology, and I just knew this was the best technology I’d ever come up with. It was another moment of clarity. This is the right way to do automation. The other ways were wrong. This is right. I’m going to pursue this.

While that was going on, there was something called Windows Vista.

Does anybody remember Windows Vista? Not our finest release.

But here’s the thing, everybody knows it was not our finest release, but they forget that Vista was the save for something called Longhorn.

Bill was on this campaign to get everybody to transition to .NET. I met a bunch of people who were like, “Well, Bill says .NET,” and did incredibly stupid engineering decisions, like trying to get that .NET into the kernel, and replacing the open dialog box with .NET. So Notepad will go from 15k working set size to when you say “save as”, it would then balloon to 15 megabytes as .NET got paged in. Crazily, bad decisions. So it was a disaster.

People got pretty angry, not at themselves for making stupid decisions, but they decided, “I’m so smart. It must be this .NET technology.” And so within the Windows organization, there became this theme to never, ever, ever, ever use .NET. Those applications guys can use it, but us OS people, never use .NET. And, everyone abandoned it and they produced what was called the seven rules for using .NET in Windows. And these were incredibly hard, incredibly draconian rules that you had to overcome in order to use .NET in windows.

But not everyone abandoned .NET

I did not abandon .NET. I knew it was great technology, I knew it was the right technology for the problem I was solving. And so I withdrew from the Windows product, developed this on the side, and that began a difficult period.

While I was still working on this, they found out that I was working on automation and on command line interfaces. It was the most miserable two or three years of my life. You cannot imagine how miserable they made my life.

I had executives say things like this, “Admins don’t want command line interfaces.” I said, “Yeah. What did you do before you came to Microsoft?” “I came here from college.” I said, “Yeah, I knew that.”

I had one executive literally pull me aside and say, “Jeffrey, exactly what part of effing Windows is confusing you?” It was a very, very difficult period. But I was absolutely convinced that I was on the right path, so I stuck with it. And basically, if they weren’t going to fire me, I was going to get this technology out the door.

So what did I do? I found the coalition of the willing. The exchange team had the problem, they understood the problem, they saw the light of what I was doing, and they became our partner. And every time they tried to kill me, I’d say, “Let’s talk to the exchange guys.” And the exchange guys would say, “Hey, I’m betting my multi-billion dollar a year business on this technology. What are you trying to do?” And people would back off. So that provided me some protection.

But then, then it was time to get back in Windows.

Remember, there were the seven rules of getting .NET in. And I’m telling you, these were herculean tasks. But I said, the heck with that, I’m in. Each one of these things, I nailed it. Every single one of these rules, we solved, we solved it well, we solved it by a wide margin. So we knocked on the door and then we said, “We’re bringing PowerShell and .NET into Windows.”

And that’s when I found out that there were eight rules to using .NET in Windows, and you only find out about the eighth rule when you pass the first seven. The eighth rule, you guessed it. No .NET in Windows.

But the reality is that there was this transition to large data centers that people were being able to see it. And so the owners of the server business said, “Hey, wait a second. It’s time to go revisit this conversation.” And ultimately that was successful and I shipped Windows PowerShell in Windows!

I eventually got my promotion back, it only took me five years.

Now, remember how we talked about performance evaluations earlier, what do we do? We take all the money and we give it to the people moving the business forward. Well, guess what? For five years, I was not viewed as the guy moving the business forward, so this effort cost me a lot of money.  But when I eventually got my position back, and then people are like, “Yeah, this is great stuff. This is really important for our business.” I started getting those promotions to Distinguished Engineer, and then I was the Chief Architect of System Center, and then the Chief Architect for Windows Server. Eventually, I became a Technical Fellow. Now that’s the top of the rang. My joke is, I can no longer be promoted.

But again, the point is, when sometimes you go up, sometimes you go down, and if you get the timing wrong, sometimes it goes down before it goes up.

Just because it’s a transition, doesn’t mean that it’s going to help you, and it doesn’t mean it’s going to happen on the timeframe you think it’s going to happen.

Now, there is another big transition.

In fact, I think this is the biggest transition we will see in our careers. I could be wrong, but I’m pretty sure of this. And this is the transition that Marc Andreessen identified in his paper, Why Software Is Eating The World.

If you have not read this paper, you should read it, but here’s the gist of it.

  • He says software is eating the world by eating traditional businesses. Traditional businesses are being replaced by software businesses, bookstores by Amazon, ads by Google, telecom by Skype, recruiting, LinkedIn, et cetera.
  • But even for the industries that software is not replacing, software is responsible for a large and growing share of the value of the product. So cars are a great example. It’s not a piece of code, but there’s a lot of code in it. Cars today have somewhere between 10 and 100 million of lines of code in the car. They do things like safety and engine control, they entertain passengers, they do navigation, they do conductivity. Software comprises a lot of that value of the physical object, the car. These are the two key dynamics.
  • He said, companies in every industry need to assume that a software revolution is coming. Now, here is a list of the top companies in the world ranked by market capitalization in 2017.  And over half of them are businesses whose value is primarily driven by software.

When Marc Andreessen wrote that document, this was absolutely not the case. In fact, none of the businesses ranked at that time had value primarily driven by software. 

He pointed out the phenomena before this was the world, and now, a few years later, over half the top companies are primarily driven by software. This is the largest transition we will experience in our careers. Whether it’s banking, retail or automotive, each one of the major players are all figuring this out. They’re not looking at their traditional competitors, they’re looking at the disruptors, they’re looking to Silicon Valley for the potential threats to their business.

Microsoft bought LinkedIn. Last week at our staff meeting with Satya, he said at LinkedIn we do a lot of recruiting, and we found that more jobs for developers are going to the industries outside of technology than inside the technology. More programmers being hired outside tech than inside tech. That’s crazy.

you’ve heard of Moore’s Law

Doubling of transistors every year, one of the foundations of our industry. I’m here to tell you, not so important, not so important.

There’s another Moore’s law, this one is really important. This one is Geoffrey Moore, and the Geoffrey Moore’s law is more important to the next 20 years than Moore’s Law was for the last 20 years.

Geoffrey Moore said “Every business participates in two activities— core activities, and context activities. Core activities are those things you want to invest in because they deliver differentiated value to your customers and you can charge a premium for them. Context activities are everything else.”

Let me give you an example. At Microsoft, our core is software that helps people do more. So we invest in trying to find the world’s best talent to help us do that, We scour the world looking for people that help us do that and we make sure that they are happy, et cetera.

Context— in order to do that, we have to have receptionists, we have to have cafeterias, we have to have shuttles. We have no technical fellow of the cafeterias, no distinguished engineers of the shuttles, no vice president of the receptionist. We want these things, we need these things, we pay a lot of money for them, but Satya Nadella has never held a senior leadership team to discuss the menu for next week.

Now, there’s an interesting complication, and it’s a complication that the people in this room are going to succeed or fail on, and that’s society of mission critical. The definition of mission-critical is screw up, you’re in big trouble.

So, you have a mission-critical core, of course, that’s great. That’s where you want to focus. But there’s also this mission-critical context, what’s that about? And the answer is, that’s where the risk is. And Geoffrey Moore called this the “killing field of once great companies.” Because once you have that core mission critical and you’re making a lot of money, over time, the market will respond. It will have a substitution, other competitors will come in, there’ll be a different play, and you are no longer able to charge a premium for it. It’s no longer core, and instead it becomes context, but it’s still mission critical. If you screw it up, you’re in big trouble because you’ll still get a lot of your revenue from it.

Because often to get in this first square, you incur a lot of technical debt, people continue to invest in mission-critical context activities, and that’s where they die. The challenge is in managing the mission-critical context, so you can invest in a mission-critical core.

I’m going to tell you how to do it

You got to do two things.

  1. You’ve got to create bandwidth
  2. You’ve got to invest in innovation

How do you apply the other Moore’s Law? You transition from the way the world is to the way the world’s going. The world’s going to the cloud, both on-premises and off premises. You want to create bandwidth by using software as a service by doing lift and shift. Just by running your VMs in a cloud, you gain bandwidth because those environments deal with all the hardware management, they have management tools. Then you do some lift and modernize, put some of these things in containers. Small amount of work, nice benefit.

But then you invest in innovation. Don’t be confused here. This is a separate distinct activity, you invest in innovation. And when you invest in innovation, you really want to go cloud native. The difference between cloud native and some of these other things, is that cloud native has a better portion of your energy focused in on what matters, the conversation with the customer, to gain insight, to translate into a feature, to be delivered as quickly as possible, and again, and again, and again.

When you’re messing around with a virtual machine that needs to be patched, and a registry key, and a comp file here, blah! Things like Serverless, functions, make a ton of value. They do more of the work for you. When you’re investing in innovation, you really want to use these new cloud models.

What does that do to IT Spend?

Shifts it from context to core by buying infrastructure, software as a service, and automating, and investing, you invest in innovation and you automate. You automate on one side to create bandwidth, you automate on the other side to have repeatable processes that can get better through every iteration.

The heart of digital transformation is this:

Build the things that differentiate you, buy the things that don’t. I’ll share it again. Build the things that differentiate you, buy the things you don’t. This transition is going to need new heroes. You are those heroes.

So here’s a set of starts and stops.

  • Stop clicking next. If you’re just clicking next on a dialog box, you are not adding any value to your company. Stop it. Learn how to automate. If it’s tough, tough, do it.
  • Stop crafting no-value add solutions. Look, if you’re running an exchange, stop it, honestly. If you’re in the airline business, can you get more people in your planes because you run mail better than the other guys? No. No. Now it’s mission-critical, so we can’t screw it up, but we’ll run exchange for you. There are a lot of softwares as a services that you’re not adding any value to, you’re just getting up and running. If you can buy them from someone else, it costs money, but it frees up your people, frees up your talent.
  • Stop building snowflake servers, or snowflake clouds. Embrace DevOps.
  • Stop using low leverage architectures, and leverage the cloud. Learn these new architectures that allow you to focus more of your team on the customer conversation and less on the technology.
  • Lastly, stop dialing it in. Invest in your career.

This is all hard stuff

Nobody’s confused about that. The pace of things is such that you’re always going to be uncomfortable. Live with it, learn to embrace it, learn to push through it.

You want to be the guy where transitions are putting the wind in your sail, and you don’t want to be the guy floundering. Because remember, when it comes to review time we pay attention to the people who are moving the business forward and we reward them, and there’s a fixed budget. So the awards for those people come from the people who are not moving the business forward. Every transition has new winners and new losers. So I hope you can use this knowledge to navigate changes in your career.

- About The Authors
Avatar photo

IT Revolution

Trusted by technology leaders worldwide. Since publishing The Phoenix Project in 2013, and launching DevOps Enterprise Summit in 2014, we’ve been assembling guidance from industry experts and top practitioners.

Follow IT Revolution on Social Media

No comments found

Leave a Comment

Your email address will not be published.

Jump to Section

    More Like This

    Industrial DevOps: Bring Agile/DevOps to Cyber-Physical Systems
    By Suzette Johnson , Robin Yeman

    The following is an excerpt from the book Industrial DevOps: Building Better Systems Faster…

    My Fit Experience: Leaders
    By Lucy Softich

    In previous articles, I've been going through some of the excursions from André Martin's…

    2023 Fall Releases
    By IT Revolution

    We have had quite a year so far, and it only keeps going! As…

    Learning Sprints at DevOps Enterprise Summit Las Vegas
    By IT Revolution

    We are just three weeks away from (hopefully) seeing you all in Las Vegas…