Sam's currently the Chief Strategy Officer for DataStax. He's been instrumental in creating two, billion dollar markets. The Enterprise Service Bus during his time at BA systems and API management at Apigee. He's also done time at Google and Autodesk and is a prolific adviser to everyone from early stage startups to massive enterprises. We're so glad to share Sam's story with you all!
Today, I'm honored and personally delighted to welcome Sam Ramji. Sam's currently the Chief Strategy Officer for DataStax. In his 25 year Silicon Valley career he's been instrumental in creating two, billion dollar markets. The Enterprise Service Bus during his time at BA Systems and API management when he was at Apigee. Sam and I first met when he was laying the foundation for open-source at Microsoft back in the aughts. When, to do this was next to unthinkable. Now, of course Microsoft owns GitHub and is a titanic thought leader in free software. Thanks in no small part to Sam's work. He's also done time at Google and Autodesk and is a prolific adviser to everyone from early stage startups to massive enterprises. He has a degree in Cognitive Science from UC San Diego and still plays around with machine learning and AI just for fun. Sam, welcome to the show.
Rachel, Thank you so much.
When you started leading the Linux and open-source efforts at Microsoft Steve Ballmer was still calling the GPL a cancer. What was it like to work at a company undergoing such massive transformations in environment and culture? How did you steer Microsoft towards survival?
It was super intense in a great way actually. Many times we would reference the feeling of The West Wing, the Aaron Sorkin show. We would just feel like the world depended on our excellent work and our ability to convince Microsoft coherently, qualitatively, quantitatively, pragmatically, that open-source was here to stay and that Microsoft would be a much better company if it embraced open-source in the fullest sense.
It must have been hard though to keep the focus on that. The pushback within the company was apparent even to me. I was an analyst at the time and it was really clear that many of the top-ranking executives were almost personally offended by the idea of free software and of programmers not being compensated for their work. They were coming from a genuine place, but they couldn't make the switch to understanding that open-source developers were trying to build a commons of infrastructure.
It was almost constitutionally offensive to them. Because when you go back to the original Bill Gates memo to software hobbyists back in the mid-70s -
Not the Halloween memo that The Homebrew Club. Yes.
Yeah. Way back in the day, you can't expect people to write good quality software without getting paid because we have to pay our own bills. So that's really what you were dealing with in terms of a current at Microsoft. Interestingly, the Halloween documents then led to the Department of Justice antitrust investigation. Which led to the consent decree and in no small way there's a link between the consent decree and the funding that created the opportunity for me to do what I would be able to do at Microsoft. So it kind of goes back that far with Jim Allchin and others being asked. Why did they make it so hard to integrate with Windows?
And this is where the language around cutting off netscape's air supply came from this very aggressive internal language about the browser wars.
And so what happened was that was misconstrued to also apply to an intentional nefariousness in every single thing Microsoft did. So the browser wars were over the top, right? That's HTTP, it's coming over the public internet. But what got deeper into the consent decree and then The European Commission statement of objections - the SO actually had everything to do with Samba. So it was the server side connection, server to server. Asking questions like, “How do you do windows identity? How do you look at file systems? How do you look at users? How do all those things connect?”
So that created a space for us to start working on open-source. I kind of came to Microsoft in an odd way. Very prodmatically joined their Venture Capital group in 2004 and their organization was asking, “how do we get a startup to use more Microsoft software or any?”
I went to a software-as-a-service conference in maybe March or April of 2005 and in eight hours they only mentioned Microsoft Technologies twice for a couple seconds each. They said, “Oh and we also use SQL Server. So it was kind of a shocking wake-up call for me to say, “Hey, if you don't change fundamentally, you're going to lose this entire space of SaaS startups.” That became the BizSpark program. I wrote a 10-page strategy paper. I worked with an open-source consulting organization led by Andrew Akin which was pretty excellent. That strategy document made it all the way up to corporations, went into corporate strategy, and it went to a particular person named Bill Helf. Now Bill who is now the CEO of Vulcan which is all Paul Allen's endowment and sort of global trust. Bill had been hired in 2004 from IBM as Microsoft's Linux person.
He actually understood Linux and he had the courage and the constitutional fortitude to as a basis of this job, “I'm going to demo Linux and open-source technology to Bill Gates himself. I'm going to show you this stuff that really works, and that it's really real and that you need to pay attention.” That began a string of unbroken hits of quarterly Bill demos that he and then my team did for, going on six years. A record that I think I can be confident in saying, will never be broken.
We showed Bill all this staff from SELinux all the way to the last demo I did. After Bill had retired Ray Ozzie was all about open-source cloud and how you could do live migration using OpenNebula. We showed Appscaler. We showed an arrangement of things that my team of Linux generalists put together in two weeks with no specialist knowledge. That was the kind of technical core that we used both to drive some fear into Microsoft product organizations, as well as some enlightenment. So we could show everyone who cared about it that there's real stuff here that you can't just dismiss and if you're going to compete with it, good luck. But if you want to work with it, then you might have a really good opportunity.
That was kind of the general drive that we took over those years and I was blessed by Bill's insight to create a group that was worldwide. I had 80 people in 80 countries and 40 people in corporate. So a good hundred twenty person team and we owned a corporate scorecard item. Talk about Corporate Innovation. How do you do it? Well, you have to measure it. So Microsoft is a balanced scorecard company. We had 33 scorecard items and these directly drove Corporate Vice President compensation. Every Regional Vice President in the field was a Corporate Vice President. You were going to get paid differently based on whether you obtained the right results on the scorecard. I owned the Linux vs. Windows market share scorecard.
That created consequences for field organizations and product teams if they weren't doing the right things. So based on that we were able to have a place to stand and we were included in all the big corporate reviews, all of the annual planning cycles. We were able to run plays in the field with an 80 person field team and a 40-person corporate team. The Linux and open-source organization that I led had the brains, the hearts, the arms and legs, and the ability to think new thoughts. The goal was to be able to communicate these thoughts and then to be able to take action - and we did that on a consistent cadence. That was how we were able to support Microsoft Corporate Innovation to move off of its old models of being extremely open-source hostile to understanding how open-source could work really well for their business. Going from open-source apostasy to open-source apotheosis.
Right and one of the big pieces of pushback we heard over and over again was that open-source had not yielded a billion-dollar business other than Red Hat at that point. But of course right there in the mid-aughts was when the big open-source business started to appear and they didn't look like Red Hat. They were Facebook, Google, and Amazon Web Services. People were building on top of these huge open-source stacks and building SaaS services that work proprietary, but depended entirely on the work of open-source developers. That was a moment at which everyone's perception started to shift and you were instrumental in that.
Yeah, that's super well said. If you wanted to get the cloud workloads that were going to make you successful, you were going to have to lose your religion and realize all workloads are created equal and you want all of them. That was a huge driver in shifting the Azure strategy. But the most foundational element of Microsoft is that they are a platform so they make money only when somebody is running something useful on top of it. That was really what let us change the Microsoft orientation. If you want more web market share, more web workload market share, if you want IIS, Internet Information Services to be a better competitor to Apache - then you need to look at what people love about Apache. Which was most notably PHP. We tied all of that conceptualization together and said, “Let's rebuild PHP for Windows.” We found out that they were still 32-bit binary if the people had lost the compiler and last time I'd been built for Windows was 1999.
There were a lot of low hanging fruits to make PHP amazing on Windows. Then we followed that up with a gold sponsorship of the Apache software foundation. Then we followed that up with a worldwide PHP on IIS campaign. It's the kind of thing that Microsoft couldn't have done before and what we tried to teach the organization. You can have insight and you can understand the technology. You can build a technology integration to make Windows a better destination and then you can communicate about it in a way that you can be open-handed and you can be heard. That can drive market share. I think that was something that was a pretty extraordinary project that we ran. Bill Staples who was my product group partner is now the chief product officer at New Relic. So you kind of look at the ways that we've had to think about large-scale corporate Innovation meeting communities where they are learning about technology evolution and somehow blending that into the particular cocktail that works for one company or another. That's kind of the lifeblood of the kind of strategic work that makes me excited.
That's really interesting because now that I think about it, New Relic arrived on the scene as a tool for developers that afforded them the same kinds of convenience and immediate insight that people expect from Microsoft software. It brought that customer centricity into the developer realm and it was really a revelation to a lot of developers when it started to appear. That was a really interesting point as well because now we're in the early teens, Satya Nadella is about to become the CEO of Microsoft thanks to all of the groundwork that you and Andrew and your team laid. I was involved in the open-source communities at this point and we were still very skeptical whenever the enthusiastic bright-eyed Microsoft people turned up. But such an appointment was really The Crossing the Rubicon point where we were like, yeah, the the forces of open-source have decisively won here.
Yeah, It was interesting. We had to build the infrastructure, right? We had to build the spinal column to enable all of that activity to take place. So one of the most interesting meetings I've ever been in my life was June 20th 2008. It was a week before Bill retired. One of the things that he specified that we were absolutely going to get solved was fundamentally changing the rules of how Microsoft engineers could work with open-source. Because until then there was this concept that if it was open-source code, engineers would have some residual knowledge and that can create an IP infringement on Microsoft software. They actually use the term that is not technical and it's quite scary - they called it “taint”. You could become a “tainted engineer” and then it would be N years before you could work on Microsoft software - which is like career suicide for a Microsoft engineer.
We were able to completely change that by working collectively with Herasia Gutierrez, who is now the general council at Spotify. He was the head of intellectual property and licensing work with that team, worked with other open-source product needs and my organization. We sat down with Bill with Brad Smith who's a Microsoft president with Steve Ballmer with Craig Mundy and a few other members of the project team. I think David Kafir, Ray Ozzie were there as well. In the presentation, we had done nine months of work to prove comprehensively that working with MIT and Apache licenses should be green lighted and it was essential for Microsoft to be able to change its approach and put any engineers to work on those kinds of projects. That would allow us to put our shoulder to the wheel to become a more effective company. There was a direct link between that and our ability to participate in Apache Qpid and help AMQP.
That was a seminal moment. Balmer and Mundy were so hostile in that meeting that Bill actually had to stand up in the middle of the meeting and say some pretty strong language to them. Then he went up and wrote on a whiteboard about how everything would work and why it was a good thing for Microsoft and did not increase the company's risk. That constitutional right offense that Steve carried deeply in his body made it extremely difficult for him to understand how open-source would work and why it would work. In fact many years later after he had left and he was the head of the Clippers, my dear friend Bill Helf was sitting with him in the crowd chatting. Steve was talking about how he had discovered this amazing source of data about how US Government funding gets acquired and how it's spent and the gap between that and the social contract we're in. He was just going to share all that data out. So other people can look at it as well. And Bill goes “Huh, Steve you're kind of taking an open-source approach to this.” And Steve's like “No, I'm not!”
How dare you! If you told me 30 years ago that I would be anxiously awaiting the next post to Bill Gates’s blog so that I would know what had been reading over the summer - I would have had you put away. But one thing that I really admire about Bill Gates is that he did completely change his mind. He did a 180 on this and he's shown that ability to assess the evidence and decide that he was wrong and he needs to change his approach many times over the course of his career.
Yeah, Bill doesn't get enough credit as an outstanding listener. He really sits and listens closely and he is able to put his total focus on whatever is happening and listen to everything on multiple levels. Not ask a single question until you're done - that could be five minutes, could be 45 minutes. Then he will open up with an encyclopedic array of really on-point questions to fill out the rest of the space of knowledge. Every time I got to prepare for a Bill Gates demo, which my team would typically put about 400 hours into prepping for a half hour Bill Gates demo, I put in about 50 hours of my time to make sure that we never wasted Bill’s. He understood open-source extremely well and was not really in the public eye for that at all in 2005, 2006, 2007, 2008, and 2009. I think the PR organization again speaking of corporate Innovation, couldn't figure out how that would foot with their stories. He didn't get the credit in the public for how well he really understood both the intellectual property structure and why it worked - why the technologies were important. He used my team extremely strategically.
For example, we show him demos on how fast Linux could boot and then he would make sure that the head of Windows was at the demo. So they would have to deal with the fact that open-source was moving really quickly. Then Bill Gates would say, “Well, why can't you do that?” He would use us as a very thoughtful club or prod to get the Innovation he wanted to see out of Microsoft Corporation. Effectively we were kind of the black sheep squadron. We were flying for the opposition to show his own teams where the gaps were in their thinking in their velocity.
I think that was a really good move on his part to create and empower a team like ours and then to use it the way that he did. because it's so easy in corporations. You've been there for a long time, when you're super senior people want to wrap you in cotton wool and tell you what is safe for them for you to know. He managed to continue to find a way to undo that and get new sources of information and caused the Innovation that he needed to see in the company.
When you look back on your work, and I guess in particular this turnaround at Microsoft, can you pinpoint one or two things that you're proudest of?
Sure, I think there were three moments that really stand out. One was just deeply personal. I'm an introvert. I have a lot of social anxiety and early on I found out that the Samba Community was very angry with us. But it became pretty clear to me intellectually that what I needed to do was to go out, understand them, make peace, be direct and honest, and figure out what can be done. I flew out to Szczecin in Germany for the Samba XP conference. I was kind of a stranger in a strange land. I learned a lot of people were curious about me, some were quite kind to me. I wrote a 20-page paper on the flight on the way home from Germany. That was a breakthrough for me and kind of pushing through personal barriers to go and just purely listen.
The orientation I had for writing on the way home, having come through the experience - I pulled on my college experiences in cognitive science studying anthropology and I wrote an ethnography. The whole thing I wrote was as true as I could to the world view that I had picked up from the Samba community. And then I used that as the basis for understanding them and trying to figure out how we could bridge the two.
That put me in a really good position for the second thing that I'm really proud of which is that we created lemonade from lemons. Everybody was freaking out about the Novell-Microsoft deal about Linux. I couldn’t change that the deal was going to happen and I couldn't change people's reaction. But what we could do is we could take the enormous amount of money that had been allocated to the deal and build out a full-fledged interoperability lab for Linux and Windows. So I got to work with Nat Friedman a little bit and with Miguel de Icaza
a lot and we built out the machines, the software, and we put the teams together to make sure that they wouldn't ship Windows unless it had proved compatibility with Samba. We had a reference implementation for the protocols that allowed that connectivity from operating system, to operating system. That was a moment in time that we were able to take everything that we learned and make the world safer for Linux.
Then the third thing is another Linux contribution. We contributed under the GPL to Linux 2.6.32, a range of drivers that we had built and that was all spearheaded by an amazing principal engineer who worked for me by the name of Hank Jansen. Realizing that the work we were doing to virtualize Linux on top of Hyper-V, which was some of the work that we were doing as a result of the Novell lab. Really would be much better suited to the market if it were part of the Linux kernel drivers project. We figured out, if you can imagine, this is kind of all in this ongoing momentum we had from the Bill Gates and Brad Smith review where we changed the rules for how we contributed. Even after that, despite Apache and MIT being green-lighted GPL was redlighted. I think it was a combination of like, alright, let's knock that one out, too.
While we were building the code, in parallel we were running through legal, the ability to release it under the GPL V2. We got it done. And I think this was August of 2009. It was one of the very last things that the team did while I was doing Microsoft and I felt like that was pretty much a crowning achievement of both technologies. It worked really really, well. It basically hooked libvirt and made Linux think it was running on Linux when it was really running on Hyper-V. It was a pretty cool community achievement because we worked with Linux kernel contributors to make sure that it was done correctly. It was a really good market achievement because what it meant was we grabbed a lot of attention from people like “Oh my gosh, you know pigs are flying Microsoft has contributed to the Linux kernel. This is absolutely wild.” But fundamentally it meant that we could run Linux workloads on our Cloud. All of the different things that I've mentioned before came through as elements in a complete strategy, we're going to create a completely coherent contained end-to-end strategy to serve Microsoft's business interests by doing the right thing and by doing the right thing in the right way. So that was pretty awesome.
And it was really received as an epochal moment in the open-source community. I was close to XenServer at the time and Andrew Morton who supervises the Linux patches because he's an Australian. He's an old friend and came to our wedding of course, but I remember Andrew and my Xenserver friends saying to me Microsoft released these patches and they are really good. The astonishment from the open-source folks was palpable.
Turns out we had really good engineers, but particularly I want to call attention to Hank Johansen. He worked on Unix and he did the multi-threading code for system 5 release five. It was 1994 I think when he put the multi-threaded code in there. He worked with Brian W. Kernighan and Dennis M. Ritchie when he was at Bell Labs where he came out of the AT&T Unix mother load. So, incredibly competent and it was an amazing day in my career when I was able to recruit him to Microsoft because he was very much like I was previously, an anything-but-Microsoft, kind of person.
The deep care and competency that Hank brought to all of his work really showed up in that contribution. I think it changed the game for Microsoft. I think a fifth of all of the Azure workloads are running Linux. You can draw a direct line between that success and Hank’s vision, inspiration, and frankly perspiration, because he wrote the code.
Even three years ago AWS was still the default choice for early-stage startups building on the cloud. Now that calculus is changing and Azure has become a real force to be reckoned with.
The other thing to make sure that we understand in corporate Innovation; the Chinese expression, I believe is, “The nail that stands up will be pounded down.” That's totally true. I had these funny conversations with people like, “Oh, you must be the loneliest person in Redmond (Microsoft Redmond Campus).” Far from it. As the Linux and open-source person, there were so many people in Microsoft who just wanted to do the right thing. Why did it create an open-source foundation? Why open-source parts of Windows and want to embrace more of it?
So these folks were all across the company. There was a collective community that wasn't super carefully organized like communities generally aren’t. But there was a movement and there were a few of us who ended up being lightning rods for moments in time. But all of that is built to the community that's there today. So that community obviously has strengthened and with Satya’s ascendancy. They've come strongly out of the woodwork and they've changed from the old normal of “anything but open-source” to the new normal, “of course, we're going to start with open-source.”
Clearly Microsoft's transformation was decades in the making. Do you think this is part of what makes you such an optimist? Because you really are. Do you think having done that work and seen it bearing fruit so many years later is what gets you up in the morning now?
I'm certainly an optimist and I'm a pragmatist. That was what let me leave Microsoft feeling really good. I left Microsoft to go to join a close friend of mine who was building a new company called SONOA Systems and it eventually became Apigee. Long story short we took it public and then it was acquired off the public markets by Google. As early as 2009, as I was learning about Red Dog and demonstrating open-source to Ray Ozzie and the Red Dog team. That is what Azure was called before it was Azure - and showing it to Bob Muglia. I could see that the changes were in place that Microsoft would figure out how to do open-source, if only because of the cloud and it's need for the workload.
I tend to see over long horizons. And once something is a solved problem in my head, even if it's not really solved in the world. I'm kind of satisfied. I'm kind of done. I know it's going to turn out a certain way and I thought it would take about five years and when I left in 2009, I said 2014 it's going to become really public that Microsoft loves open-source. Well behold. Built in San Francisco, 2014 they launched all of the Azure pre-built open-source applications. These are third-party applications that are actually popular rather than new fake ones.
Some things just become really obvious to me once I stick my head into them for long enough. I try to understand the physics of space and if you call that optimism, I guess maybe it's excessive confidence in the power of intellect. I tend to think that little bit like being patient, whether your patient or not doesn't change the result. It changes how much suffering you're going to experience. The choice between optimism and pessimism is the same. I think we optimised suffering a lot less.
If you had a do-over what might you have been done differently?
Anywhere in your career. You get a mulligan you get to have another go.
Career mulligan, what an interesting thought. I'm not a person who lives with a lot of regrets. Not sure that I have any particular do-overs - I was fired from my job as Director of Engineering at Ofoto about two months after we were acquired by Kodak. I went from the person who was doing behind-the-scenes work to get the Kodak acquisition to happen in 2001. With the executive saying, “Oh my gosh, this is amazing work. You should get a percentage of the acquisition over and above whatever stock you have because this is so amazing. Our bacon has been saved.” To like, “Oh my gosh, you're a terrible communicator. You're putting us at risk. You can't stay in your job.” Which obviously neither of the two things were true. Yes. I did a lot of work, but the company had been around for a while to be able to hold the promises and attract Kodak's attention. It wasn’t true that I was a terrible Communicator or that I shouldn't be in my job. But that's just how things go.
I spent six months after I left Ofoto coming through September 11th, the job market dried up. It was a super hard time. But that hard time led me to ask really fundamental questions like, “Why am I still in engineering when the people who are in charge might not even know what they're doing?” I need to go figure out what's outside of engineering. It's kind of like if I hadn't had that experience - “A” I'd probably have less empathy with people who get fired but “B”, I would probably have stuck with engineering all along and I wouldn't have had the opportunity to work at Microsoft or do any of these things. I realize that's again sort of an optimistic philosophy. You get knocked down nine times and get up 10. You're so much stronger the tenth time you got up and can’t regret any of the nine times you got knocked down. I really can't.
What do you think makes corporate innovation in particular so difficult?
This is a great question. I've thought about this a lot especially since getting an opportunity to work with you so closely at Autodesk, which was one of the great parts of the last decade for me. Coming from highly innovative cultures and then landing in a very old and slow moving company led me to realize something. A lot of people are not just status quo folk because status quo is easier to maintain, but a lot of corporate organizations have all these complex cultural rules that nobody quite understands. Nobody ever actually developed, but they're encoded in everyone's behavior. A little bit like that nail that sticks up and gets hammered down. If you want to stay tenured in an old company - you don't rock the boat. And that is the antithesis of innovation. You don't bring up scary ideas. You don't know how they're going to be taken. One of the things I started to understand was Network math. I was trying to figure out the math behind my social anxiety. Why do I do so well in one-on-one situations and why are much bigger situations difficult for me?
It turns out in one-on-one, two people have one relationship. But the relationship map scales up really fast. With three people there are four relationships. With four people there are 10 relationships. It's an additive factorial, all of the different ways that those groups can be combined. So now we're just up to 4 people and we get 10 relationships. Now take a 10,000 person company - not to say that every single person is in a relationship. But what is the network math of character, the fabric, the beliefs, the implicit behaviors, what's rewarded in all of those? So as that number gets embarrassingly large, you can start to imagine the actual inertia behind change if you don't have a change oriented culture to start with. I do tend to think that there are Innovative pockets in large corporations, and there are entire Innovative corporations. But if you haven't built Innovation into your culture fundamentally, it's very difficult to bolt it on or add it in. Because despite their very best intentions and people's ability to say “Yes, absolutely I'm doing the new Innovative thing”, 99% of the time they are using new words to talk about the old thing. They're not actually going through a process of innovating, changing their orientation and figuring out a way to do things fundamentally different or not at all. Because sometimes what's required for Innovation is a vast brush fire that wipes out all of the stuff that's kind of started to accumulate that you have to do every day. It gets in the way of being able to do anything new.
We have this tendency, particularly now in California, to think of wildfires as intrinsically destructive. But what we're glossing over is a ten thousand year history of fire farming on the part of the indigenous people and an ecosystem that has evolved for fire. I've been finding that idea of regenerative fire a good touchstone for the work that I'm doing in Corporate Innovation now.
Absolutely. What are the things that you care about? All of those scrubby huckleberry oaks and blackberries or do you care about the pine trees and the spotted owls. If you care about spotted owls and pine trees, you are pro brush fires. You can’t hold those two ideas the same because when you knock that out, the spotted owls have an easier time hunting, you know for voles and mice on this open forest floor. The pine trees are less competed with so they've got more ability to stand taller, sequester more carbon- there is a lot more carbon sequestration that happens in these larger older growth trees. So, there's all sorts of good reasons to burn out bushes that just kind of get in your way. But you can start to believe that it's everything that matters to you because it's taken up 95% of your working hours.
Just to take this analogy a little bit further - we fetishize this idea of wilderness as something untouched by humans, but in fact humans are part of the environment. There's a positive interaction that we can have with it that is probably as decent a definition of Innovation as we're going to have.
How would you distill all of this experience into say two or three lessons for our listeners?
The Archimedes principle is the most important one if you're going to create large-scale Innovation in your company. What is your lever and where are you going to stand? If where you're standing isn't related to revenue, things that the management chain cares about, things that might impact something that the board of directors or the public market cares about, then you are probably wasting your time.
If your lever doesn't include technology and unity and an ability to deliver that Innovation to your customers again, you're wasting your time. To turn that upside down, focus on something that is close to the core of what the business is. Because by definition, businesses have to care about them and make sure that you have the social and political connection to the people who are fundamentally responsible for delivering those results. Ask yourself, what is being measured? You will end up finding one metric. That simplicity of strength that you believe you can move, that's your place to stand. Now figure out how the technology in the community and the customer value will be your lever so that you can move it. That's a major thing that I would offer is to use the Archimedes principle.
It's some powerful customer development on the grand scale. It's understanding how an entire customer base needs a company to evolve such that it continues to be relevant in the future.
Yes. And so it was the corollary there. There's a wonderful book I read maybe eight years ago by John Hagel and a few others who were writing a book called The Power of Pull. What they were looking at was to reimagine the corporate hierarchy from a top down pyramid to concentric circles. They place the customer outside of the last concentric circle and say that that is the edge. A really strong company these days is not focused on push - how do you push your intent out to the world - but it's focused on pull. Then the most important and valuable people for Innovation in the company are not the ones who are closest to the center who know exactly how to run the old business; they're the ones closest to the edge listening to weak signals from the customers about what the customers want from you that aren't yet fulfilled. They called those marginal behaviors and marginal requests. The point that they made was that the great transformations of companies and societies have generally come from marginal communities - literally meaning on the margin between the edge of the organization and the world outside. Often not real well understood, often not real well championed or explained, but that is where you can find your next business model.
When I link that back to what you and I did together at Autodesk. We have an enormous number of the world's buildings. Over 50% of buildings are designed in Autodesk software. There's a lot of different building information models and records are great. So we know what the buildings are. But if you're at the edge and you're listening to the constructors of the buildings, you're saying, “I need you to answer different questions for me. Is the project on time? Where are the risks in the building? How can you help me understand all that through data? How can you point that out to me?” Those marginal conversations then end up creating transformation if they're properly listened to and properly curated and you've got a serious leadership team that is seeking truth outside the company. If you couple the Archimedes principle with the power of pull you end up in a pretty strong place not just to figure out where to be and how to do it. But what to do listening to those marginal voices.
That's a really fascinating framework for thinking about it. I mean, there's a lot to criticize about Google and Facebook, but when you think about the ways in which they and Amazon have been extremely effective, it's in empowering small groups of engineers to listen and act. That has given them a way to in-source outsourced Innovation. So where Sun would throw off these misfit bands of engineers who would build a company that would then be reacquired by Sun - Facebook, and Google managed to keep that activity happening internally. I think it is because they have these circles and they pay attention to these weak signals from the customer base.
There's no reason to say that the Sun’s model is worse than Google or Google's model is better than Sun. The question is, is the Innovation happening? Cisco has a similar model. One of the interesting things about throwing off engineers, whether you spin it out intentionally or whether they quit out of frustration and inspiration. While they're outside the company, they're Gathering new signals and they are derisking the innovations that they're doing. Let's be real 98-99 percent of Innovations fail just because you invented something doesn't mean anybody cares about it.
Derisking it and then being able to say “Oh, you know now that's something worth 250 million dollars, let's hire those 10 people back.” On the one hand you'll get the push back inside the company going that's outrageous. But on the other hand look at all the other things that failed, didn't work, that you didn't have to pay for. So, effectively you're reaping the rewards of having good relationships with your ex-employees and figuring out how to pull those folks back in. Because they understand both the Innovation culture and the market. They understand your source culture. So they are more likely to be able to have something that sticks. Facebook and Google can do that internally because they have such an extraordinary and profitable business model then they could just ignore an enormous amount of waste. The vast majority of their employees do not produce revenue for those companies. All of that insourced Innovation is actually driven by these platform business models that are throwing off astonishing margins.
Not to toot Alchemist’s horn, but you did set it up for me. One of the things that I'm proudest of this little accelerator for achieving is taking strategic investment from Cisco and building three companies that Cisco then turned around and reacquired. I think that's a real testament to the model that Ravi has built.
It’s huge. Ultimately Facebook and Google will come crashing to Earth. Their margins will start to shrink and then they will have to adopt more of this pragmatic spin-out, spin in. Whether that's implicit - people just leaving and swarming companies and being required or whether it's explicit and there's actually an intellectual property clearing and spinning out competency.
How do you think the pandemic might affect corporations long term?
It's informative to see how it's affected them in the short-term. Most global 2000 enterprises that have been surveyed in the last few months have said they have accelerated their digital transformation. Which is a wide word. It incorporates everything from remote work to changing how you provide value to the customer. They’ve accelerated the digital transformation by anywhere from 2 years to seven years just in the last six months. There's two things that I find very interesting.
One is when you digitize, you reduce the cost of Innovation. The more data that's available to you, the more simulations you can run, the more what-if scenarios you can have. You can bring people together around a new idea of the future. “Hey, what if we did this when I crunch this enormous amount of data that we now have because we've digitized? I think our results might look like this in three years.” So, your ability to tell a story has everything to do with your ability to create that new reality.
The other thing is why didn't they accelerate themselves two years ago? It's the brush fire problem. Right? COVID has been this brush fire that has cleared things out. And for people who are stuck in the old normal, there is the new abnormal. They might look at this time and say “Oh my gosh, we really can move fast.” What was it that led us to move slowly in the past? Perhaps some of these companies will end up inoculating themselves against some of the historical anti-Innovation forces. Now that they have round truth, they have lived experience that you can go fast and you don't have to have a meeting with 16 people - 14 of whom were there to shoot the idea down. You can instead say, “Hey on the digital foundation we built, we can fail fast and hey if we fail five times you can turn off the project. And by the way, we're shipping every two weeks so it's only going to be a 10 week cycle.” You can fit this all neatly within a quarter and now you can be within the cadence clockwork that corporations run on.
I think there may actually be an uptick in the Innovation cadence for the companies that have survived through this. Clearly like any brush fire, there's going to be a lot of death and a lot of companies will disappear. It’s not about the old normal. It's not about the new abnormal. It's about what's the next normal going to be and we have much younger people coming to bear. This is their world war experience, right?
COVID-19 is the defining experience for people in their 20s. You think about people who are younger, more plastic, their orientation for our ethics, how we support and serve others - the basic expectations about the rate of Innovation to get from apps on mobile phones. What their web experiences are. All of these things are going to come together into the cocktail of the next normal. I think they'll have a lot less patience for, frankly, crusty hierarchical folks sitting around using their experience to stop new things from happening. Instead, adopting a model where you move fast and change things is going to be the birth of a lot of new value. The way that we think about business will be fundamentally different. Because the only ones who survive will have adopted this new style of playing and that will be what we call business five years from now.
It's another realm where your comment about where change is coming from. The margins are so pertinent because anyone who has been listening to disability activists, let alone public health folks 3 years ago, are way more prepared for the changes that are coming than people who disregard those voices.
How do you avoid burnout?
I take vacations. I have really good hobbies. I live with my best friend, which is super helpful. We've been married for almost 24 years. I meditate every day for 20 minutes to 60 Minutes. It just depends on what's happening. I have a gratitude practice. I think being grateful every day to one or more people or things in your life is really important. It's important for me to start my day off there. It's almost like rebooting a computer - start the day, clean in a positive mood with your thoughts about you. I also create a lot of space for creativity outside of my job and I give myself permission to stop working. Especially in the COVID era, I think it's really important to stop working at about 5:00 pm. Or certainly no later than 6:00. And then I do some creative writing. I still play Dungeons & Dragons. I play guitar, I write original music with a close friend of mine.
I think to be human is to be creative and don't let anybody tell you “You're an accountant. You're an engineer. You can't be creative.” It is fundamentally a creative existence that all of us are meant to live. So, if we create space for creativity outside of the nine to five and the imaginary walls of our work, I think it's a lot easier to fight burnout.
I just pay attention to when I really need to take vacation and I give myself the permission to do that. When I'm on vacation I am on vacation. The thing that we have to acknowledge in corporate life is that we are athletes. There's a great essay called The Corporate Athlete. It's by Jim layer and Tony Schwartz who created the discipline around the power of full engagement. One of the things that they identified was that, unlike an Olympic athlete who's going to have a 4 to 10 year career and is practicing 95-98% of the time, is performing two to five percent of the time. Somebody in a corporate environment is going to work for 40 years and are expected to perform 95-98% of the time and we're only supposed to train two to five percent of the time. It's an inversion of athleticism that is a recipe for exhaustion. You can't sprint a marathon so I would highly recommend their work. They really focused on not what you're doing, but what you're not doing - so the fundamental question they ask is what are your practices for renewal? Everybody gets renewal a different way. You need to figure out what yours are and give yourself the permission to do that. I think that's a real key for avoiding burnout because if you don't have those, you are burned out. You just don't know it and your normal is this terrible feeling that you have. So if you don't have a place to escape, you won't know how you're supposed to be.
If you were God Emperor and could dictate the next 10 years of the software industry and the outcome in ten years was exactly the software industry that you would like to see, what would it look like?
I think 10 years from now is a place for me as an infrastructure nerd that I really like to think of fully automated and augmentation of intelligence. We should be in a place where we are generally using open-source and open standards. We're paying people for their expertise and we're paying companies for their ease of use - taking all of the different abundances of different types of technologies. Sometimes multiple things for the same problem. Sometimes very closely related things that have to be integrated together. We would be paying for operational simplicity because that's what lets companies move really fast forward. I think data would be standardized. We would have seen the completion of the Kubernetes re-platforming wave which is standardizing how we do computation and would have created an environment for standardizing how we do data.
The next step beyond that once you can take computation and data for granted. Then you can create a lot more augmented intelligence. Things that are doing a lot of automated work. Maybe letting us see complex multi-dimensional problems in a way that we can fit into our human brains. So that we can make decisions, we can be creative, and we'll be seeing a lot more small ephemeral programs that spin up and exist. Which exists really only to help a person make a better decision and make a better connection to reach out to someone else. I think the more that we give things away the more that we can figure out how they work together and the more that will end up coming back to us in a sort of creative surplus.
That sounds really great. What about the future for you personally?
I'm going to spend the bulk of my time on data on Kubernetes. I had an amazing opportunity in 2016 to join Google from my position as CEO of Cloud Foundry. They were working to create an open cloud and they wanted me to come in and run the Kubernetes business that all started with Craig Mcluckie talking with Brian Stevens to say Sam would be a good person to come in and lead Kubernetes for us. That was the beginning of an extraordinary journey where I got to learn a lot and meet a lot of extraordinary people including folks I've talked about before and who you've met like Eric Brewer and now Damon Fessel a number of other amazing people.
But thoughts stuck in my head are all about the need to make data easier. With Kubernetes solving for how you build out applications , how you handle stateless environments, how you scale out, and even how you manage the abundance. Because you have an abundance of new services and new service nodes, it's because you have a scarcity of insight. “Oh my gosh, how do these all relate together?” So you create a service mesh. With the service mesh, now you understand your services.
What's unsolved in that data is really hard and every pot, every micro service team, has kind of done it a little bit differently and that creates sprawl and that's a mess. I think we have a really special opportunity to move beyond microservices 1.0 into microservices 2.0 by using Kubernetes as the lever to drive change in how we use data. The way that we can fix data is to create a coherent plane. Where just like the Kubernetes API is declarative and tells infrastructure what to do. We can have data for the Kubernetes API that lets you look at the security, the storage, the replication, the restoration, the backup, the quality of service is a range of different affordances that lets many different data technologies and storage technologies work together to be orchestrated automatically by Kubernetes. If we can create an application data environment, we've created the foundation for I think the big change we're going through. Which, “Is the world going from app driven data, to data driven apps?” We have an opportunity to make that a lot easier. I think because of the experiences I've had, the thing that speaks to me most is creating data on the Kubernetes community and a technical reality and seeing what that revolution has in store for us over the next few years.
And so for example, that might make the problem of tracking and tracing every case of COVID-19 in the UK a lot more straightforward than it is in an Excel spreadsheet.
Yes. So many things will get radically easier once the data is fluid.
What is the best way for our listeners to connect with you or to follow your work?
I'm semi-active on Twitter. I blog here and there. You can find me at DataStax and around I would love to hear from you. Probably terrible at email. My service level objective is only about a week and sometimes I don't even hit my SLO, but on Twitter and I'm fairly quick.
Is there anything else I should have asked you that I didn't?
A question I would love to answer is, why does AI matter? Because as we look at all these things. If we look at all of the Innovation reports that we have now, the number one transformative technology on most people's minds in large organizations and small is that in the next three to five years, AI is going to be the most transformative force. It's easy to write about people getting excited. I have a degree in cognitive science, which means AI and Neuroscience. I think about that a lot, but here's the thing - I don't have a technical answer to this. I have a humanistic answer. The technology is all very well covered. You can do smarter, better, faster things. You can do lots of pattern matching, but I think what is going to be extraordinary about AI is that the more that we can automate the things that we used to think we're special about being people. We will slowly uncover what is actually special about being people. I think that's our ability to love, it's our ability to create, our ability to connect with each other. As we keep looking into these almost cognitive prosthetics that we are able to get with Kubernetes and data and cloud computing - a lot of the stuff that we've been confused about whether that makes us especially human or not will fall away. That's my optimistic-humanistic perspective on why we should all be excited about the growth of AI over the next 10 years.
It's really timely for me because I've been interacting with a bunch of the new bots that have come out this year for various reasons. At first I was struggling because my mental model of a bot is a kind of repetitive and slightly dull human. I had a revelation a couple of weeks ago, which is that in fact, what they are is chatty texts. Text in the sense of they are books that can answer questions and interact with you. That really changed the way I feel about engaging with a bot.
That is super cool. I always look back to science fiction to tell us what human beings are when you change a few particular constraints. When I think about great near future science fiction writers, I always come back to Neal Stephenson. What you said invoked for me the diamond age or a young lady's Illustrated primer, right? What if your books could have conversations with you? My goodness, that would be pretty amazing.
That would be pretty amazing. Looking forward to building the diamond age with you. Thank you so much for coming on the show, Sam. It's always such a great time talking to you.
It's a privilege. Thank you so much, Rachel.
"You want to keep them motivated, you want to keep them moving forward, but at the same time you want to make sure that they understand that there are lessons to be learned and they should listen. People get too close to what they're passionate about and sometimes ignore all the facts." - Alan BoehmePlay Episode
"When you have intellectual work, you're not always going to have good intellectual days. Sitting there and trying to make it happen is just burning out something that we don't have an endless capacity for. If you're not doing great work today, forget it, go for a walk." - Cindy AlvarezPlay Episode
"There is purpose in all sorts of things that we do. If you can make someone's job easier, if you can help someone sleep better at night, or whatever it is. Reach out and have those conversations. Understand the context of the people that interact with those products and services." - Christi ZuberPlay Episode
"You can't believe everything you read, see, or hear at first glance. Just because we're pattern-matching animals doesn't mean that your conclusion is accurate. It just means you matched some pattern. That pattern may or may not be useful at all." - Mike DolbecPlay Episode