PurePerformance - 019 DevOps Stories, Practices and Outlooks with Gene Kim: Part 2

Episode Date: November 7, 2016

Gene Kim has been promoting a lot of the great DevOps Transformation stories from Unicorns (Innovators) but more so from "The Horses" (Early Adopters). The next DOES (DevOps Enterprise Summit) is just on its way helping him with his mission to increase DevOps adoption across the IT world.In our 3 podcast sessions we discussed the success factors of DevOps adoption, the reasons that lead to resistance as well as how to best measure success and enforce feedback loops.Thanks Gene for allowing us to be part of transforming our IT world.Related Link:Get a free digital 160 page DevOps Handbook Excerpthttp://itrevolution.com/handbook-excerpt?utm_source=PurePerformance&utm_medium=organic&utm_campaign=handbookexcerpt&utm_content=podcast

It's time for Pure Performance. Get your stopwatches ready. It's time for Pure Performance with Andy Grabner and Brian Wilson. All right, hello and welcome back to our Pure Performance podcast. We're continuing our conversation with Gene Kim. Andy, hello. Hi. Hi, Brian.

Hi. Hi, Gene. Is he still there? Yes. Yes, I am still here. Thank you so much, Andy, and thank you, Brian. Has anybody ever do Mean Gene the Dancing Machine? Have you heard of that? Reminds me of elementary school, yes. From the Gong Show, yes. Excellent. That's what I'm here for. I'm here to keep it juvenile.

You know, I wanted to bring up a topic that, you know, looking at DevOps, right. I'm looking past DevOps now. Right. So DevOps is awesome. Everybody loves DevOps and it's, it's super, super great. Um, but in general, everything kind of has its time, right? There's, there's, things are always changing in the world. Just like how, uh, DevOps replaced waterfall because waterfall was inefficient. DevOps is probably going to be around for a long time. And I think it's, you know, not saying anything bad about it. But what do you have you ever thought about what it would take to make DevOps obsolete? Like what would have to happen within the technology world to say either we no longer need this or this is no longer efficient

kind of more of a future forecasting idea if you you know have a crystal ball kind of a thing oh gosh you know i think one of the problems that's been verbalized by the devops enterprise community uh is that it's not enough for dev, test, ops, and infosec to change their way of working. I think one of the bottlenecks that quickly become evident is product management and product owners. We need, in order to achieve these amazing outcomes of fast flow of work from dev through test, through operations, so customers are getting value, is we need user stories to be articulated in a way so that, you know, that can be amenable to being worked on in small batches. You know, people brought up, you know, how do we do, you know, financial budgeting in a way that, you know, doesn't, you know, preordain waterfall like outcomes.

But I think, you know, those are I think I would call a part of the DevOps journey. I think those will become the natural bottlenecks that as technology leaders we will need to overcome. I think names are very powerful things. I mean I think the genius of Patrick Dubois to call this problem that we've felt for decades, to give it a name, the DevOps problem, which DevOps is a solution, I think is just, I think that coming up with that easy handle for both the problem and the solution was a flash of genius. So, you know, I have, I don't think it serves as well to change the name of DevOps to, you know, business dev test ops,osec, whatever new handle we want to come up with. I mean, I think what makes DevOps so breathtaking is that it is this massive pendulum shift, whether it's from waterfall-like releases to being able to have multiple deployments per day, whether it's kind of the siloed functional orientation where we put all the developers in one silo, testers in another silo, ops in another silo,

right, the DBAs in yet another silo, right, to move to more platform-oriented teams or actually embedding perhaps those people directly into the feature teams themselves. I mean, I think in so many ways, DevOps represents this massive pendulum shift from one extreme to the other. And I think that's what makes DevOps so exciting. But do you think it'll ever become obsolete in any way, shape, or form? I mean, I know this is kind of a crazy concept of a question, but just like how people might have thought at one time waterfall was good, is there some sort of technology disruption that could come out? I don't know

if it would be maybe self-learning machines or something else. Um, and that might be going too far, but is there ever a time? I mean, I don't know. I guess it's, it's more of a question of, is this sort of the end or, and it's just going to be modifying and adapting to those ideas that you outline? Or do you think there's some kind of – I guess – There very well could be, but I guess here's the math that goes through my head. IDC says there's 8 million developers on the planet, 8 million ops people on the planet. And, you know, I think by the most optimistic of estimates, you know, let's say one to 2% of, you know, that population of 16 million engineers are actually using DevOps principles and practices. So in my mind, right, that means we have 98%

to go, right? And we deserve it, right? We know that, you know, the benchmarking work that we've done with Jez Humble and Dr. Nicole Forsgren in Puppet Labs, we know that the high performers that are using DevOps principles and patterns are massively outperforming their non-high performing peers, right? We know they're doing 200 times more frequent deployments, 2,500 times faster lead time from a change committed to version control through testing, through deployment. So it's actually running production. They have better production outcomes. But we know that they have less burnout. They are 2.2 times more likely to recommend their organizations as a great place to work to their friends and colleagues.

We know that they have better organizational performance. They're twice as likely to exceed profitability, market share, and productivity goals. So in my mind, this is moral imperative that we get as many people to adopt DevOps as we elevate the state of the practice so that every engineer is as productive as if they were working at a Google, Amazon, or Netflix, or Facebook? I mean, I think that's, in my mind, that's sort of the challenge at hand. And, you know, until we're there, you know, I guess, I'm not so worried about kind of what is the next paradigmatic change that could change. I love the passion you have for it. I think we all kind of feel that when we read these books.

There's definite passion for it. And hopefully the other people will get on board and start making the change. You know, it's quality of life comes down to, I think, so much. You know, actually, just to add one more element to that. So a friend of mine, he was telling me at the bank that he's at, he said we had a reduction of force of 500 people. A lot of it was in operations. Some of it was in release management. Some of it was project managers.

And he said, on the other hand, we had 600 open job recs for anything related to automation. So that's performance monitoring, that's continuous integration, continuous testing, platform engineering. And I asked him, holy cow, that's amazing. Did you offer retraining opportunities? And he said, absolutely, yes, but only 20% of the 500 people took them up on the offer, which blew me away. And my next question was, how is that possible? Why wouldn't they take the training opportunity? And he said the most common response was, I only need one more job. Somewhere in this big

metropolitan area, there's another release management job for me, which I thought was startling. So I told that to my friend Tom Limacelli, who wrote the marvelous book, many books, including the cloud administration book. He said, isn't that interesting? In the ops community, we're at this weird, strange fork in the road right down one path. Our salary goes down by half because we're working at the Apple bar, at the Genius bar at the Apple store. And down the other branch, you know, our salary doubles because we have the

hottest skills in the marketplace. And, you know, so I think that kind of incredible reskilling of the workforce, I mean, I think that is what's happening inside of most organizations. And so, you know, one of the things I just have a tremendous amount of passion for is how do we make sure that we don't leave anyone behind? How do we elevate the state of the practice so that everybody gets to feel, know what it feels like to be working in this type of environment where we can actually get work done instead of waiting or associate deployments with fear, right? I think we hear a lot of, a lot of our, not colleagues, but a lot of our, a lot of people

we talk to have that similar situation where the company offers, you know, there's, there's an opportunity to level up and change to the new style. I think it's very important for companies to do that because those people have all that knowledge. There's so much institutional knowledge. And I'm sure, yes, the transformation is difficult for any company to go through. And sometimes people might have baggage that's hard to break through. But also the knowledge they have, it just seems to me it makes more sense to try to convert these people into the new way and retain all the inherent knowledge they have than to just toss them out with the, you know, I guess it's going to say morning paper, but I don't know if those exist anymore. In fact, one of the talks that I'm super excited about next week at the DevOps Enterprise Summit is Adrian Cockcroft.

So he's famous for many things. He was one of the 50, 60 distinguished engineers at Sun Microsystems back in the day. He was part of the eBay transformation and reconstruction in the early 2000s. But I think what most of us know him for is the architect at Netflix who was responsible for transforming the video delivery service from a monolithic J2EE app that ran in a private data center to one that runs entirely in the Amazon cloud. And he told me that he was inspired by a talk that was given by Jeffrey Snover last year at the DevOps Enterprise Summit. He's one of the 11 technical fellows at Microsoft. And he said so many organizations these days, as part of this engineering transformation, they realize they need a career track for individual contributors. And you had Adam Auerbach from Capital One on the podcast. I mean he's – one of the people that he works with is Topo Brada Pal, who helped create the DevOps movement at Capital One.

So he's a technical fellow there. And so he's – Adrian Cockroft will be talking about how do you create career tracks like a ladder that includes a distinguished engineer or a technical fellow to allow people like Topo to do what they do best without having to manage people. Right. And so I think it's, boy, what an exciting time to be in the game where, you know, so much of how we work and how the organization is supporting those people is changing. Hey, yeah, I've well, I was like listening in and and just astounded again by – with all the numbers. I mean going back to a little earlier where you said all these numbers that you have from do it. Otherwise, they will eventually become disrupted by new startup companies that just provide something much better and faster. Why? Because they have the talent, they have the people that actually want to learn something new and not these people that feel comfortable in their second to last job

of their life. So that's just sad to me. And I hope a lot of people will, I hope a lot of people will wake up. And I hope that the work that you do, and we try to do our little part of it by talking about DevOps on the podcast and blogging about it and speaking. So this is, I think we just,

we need to get the needle of the 2% adoption up to definitely next year. It needs to be the one metric we should measure ourselves against. Next year, we will be at 3% and then 5% and then 10%. Here's a – maybe to make ourselves feel better. I think one of the people who addresses this phenomenon very well is Geoffrey Moore. So he wrote that amazing book Crossing the Chasm.

And he essentially says it's a Gaussian distribution. So you have the innovators and the early adopters and the early majority, late majority, and then laggards. And I think in my mind, this is part of the journey. I think the Googles and Amazons and Facebooks, right, the unicorns, they're the innovators. I think the early adopters, you know, those are these heroic transformations that are being pioneered by the people that we see speaking at DevOps Enterprise. And then, you know, I think the goal is, you know, how do we get that early majority, you know, who know that there's some change required.

And then we get to late majority. But I think one of the really neat phrases I've heard from Geoffrey Moore is that that early majority, they're doing it because they see an advantage both for the organizations and themselves. They say, hey, here's a chance for me to get promoted. The late majority, however, they're the ones who are saying, if I don't do this, I get fired. So I think we're just going through this kind of natural progression where we have these very courageous pioneers who are figuring out how do I lead change, right? How do I get started? How do I start and finish these type of transformations without getting fired? How do I convince stodgy upper management? How do I get them on board? And how do I avoid the holdouts that can jeopardize my initiative? I think it's

been so fun to collect these stories and to be able to, I think, help the next wave go through. I think in my mind, it's all about how do we increase the likelihood, how do we accelerate DevOps adoption, and how do we increase the likelihood of them succeeding? And I think we do that by having each one of these leaders share experience reports. Can I mention one more thing just in terms of the method behind the madness? I think as leaders, as adult learners um i think we tend to learn less effectively you know from listening to people saying here's what i think you should do right or um yeah when we just listen to theory um i think i think the more effective way that

we learn is through experiential learning, through experience reports by actually people telling their exact story of here's our organization. Here's the business problem we're trying to solve. Here's who I am and where I fit in in the org chart. Here's what we did. Here's what the outcomes were. Here's what we learned. Here's what we still don't know how to do and you know for that reason every one of the talks at the devops enterprise uh summit is given in a 30 minute experience report so there's really no room for kind of theory it's all about you know

saying here's what we did here's what we thought was going to happen here's what actually happened sort of like the scientific method and uh know, I'll send out a link where you can find all of the DevOps Enterprise Summit videos and talks. And if you have any interest in any of this, I would recommend you check out some of the talks because I think somewhere in those talks, you're going to find a story that will very much resonate to your situation. Yeah, I think that kind of goes back to, you know,

in the previous episode, we were goes back to, you know, in the previous episode, we were talking about the, you know, the horror story of the Phoenix Project of, you know, how it instilled nightmares and all that. But that kind of, to me, that brings in the same thing, you read the story that terrifies you, because you've lived through it, right? But it has a happy ending, right? And you get to see that transformation of that happy ending. And it gives you that inspiration. And I think inspiration is a lot of how we can learn and move forward as well. Seeing that other people have been in a similar situation and found ways through it is very inspiring and can push us to where we need to go.

So I think all those stories together really help drive that excitement because that's what it's going to take, right? It's going to take people wanting to do this change. Or as you were saying, the people are going to say, I'm going to be out of a job as the late adopters, was that? Yeah, yeah, right. But they're still going to need to be inspired, right? So that's why I think this stuff is also wonderful.

Yeah, this is also why I love our own story so much because our founder, he said, guys, our market is changing. And if we still want to be the leader in our space, then we need to do something. And here's the goal that I set out. Gene, I'm not sure if you remember the presentation, the part of the story that I told at our webinar. But Bernd basically said, we need to go from two releases per year on our on-premise product to a SaaS-based model where we deploy every other week with the option to deploy within an hour in case we need to because of a catastrophe or something like that, or like a hot fix or something like that.

And he set out this goal but he also said you guys i trust you and you guys figure it out what what you need you tell me what you need and i'll get it done and i'll shield you from the rest of the organization that works right now in a different mode and i think he was inspirational and he set up the goal but not only did he do that he helped along the way he he was i i believe he's he was a great he he was great in a way because he didn't tell us what to do but he made sure he could he removed all the roadblocks and the bottlenecks that he could remove in his position and i think that was great. Boy, you know, I got chills listening to that.

And I absolutely love your description of that in the webinar. It reminds me so much of this quote from the CIO of Target, where on the first week on the job, when he came from Tesco, the UK retailer, he said, my job is not to direct and to control. Instead, it is to guide and enable and remove obstacles. And I just – I love that just because I think it just underscores how – I think even the job of leaders is changing, right? It's not so much to set goals and make all the right decisions and allocate resources and check for compliance. It really is about empowering teams to do work independently and productively. And how do you mobilize those teams so they're achieving a greater goal?

I mean, I think that's actually very different than how most managers and leaders have been trained. In fact, it's very different than the way I was trained. And so I think there's this kind of incredible frontier of how do we create a generation of leaders who can create that system of work. Well, I also think there's a different problem with that, at least how i saw it in my career in my career i think when you work for a small company and that company grows sometimes you get pushed into a leadership position because the natural growth of the company and then you just get put there even though you have no clue what this means you don't know how to lead and then sometimes the companies don't give you enough time and actually education for their job and so you just do what you think makes sense. But if you are in your early 20s or mid-20s,

you really have no clue that leadership is not about, you know, filling out Excel forms and then pick micromanagement people. But like this new way of leading, as we just said, is something – I think this is where we need to go. But the question is do companies realize that you need to invest in your people to actually get them to this type of leadership? Like what you said earlier with these companies that have an education plan or a career plan where the career plan should be figuring out how to put people into the positions where they are most effective and most happy. And those that want to become leaders, how can we tell them that they don't become leaders where people will follow, but leaders where they can inspire more people to become self-sufficient and basically become more, you know, more better value add to the company.

Yeah, I think, you know, I was reading, I read a while ago, Michael Feiner, the Feiner points of leadership, I think it was. Yeah, it's right behind me. But it outlines that whole process, that whole idea that people get put into management, right? And I think that's the big thing, right? There's this concept of management, right? And a leader is not a manager.

But everyone, but the title is always management, management, you get management and just that con just that conceptual title of management means you're managing resources. You're shuffling things around. You get, you know, whereas a true, what, what companies are really looking for, but they don't know it quite is, is, is the leader, the one who's going to enable people to, to strive and do great things. And that I think is one of the biggest challenges in, in, in the biggest challenges in the corporate world is not having those leaders and because everyone still thinks of them as managers. Yeah. You know, one of the quotes I found so inspirational was from Steve Farley.

He's the VP of IT at Nationwide Insurance. And he said, I had the privilege of being able to go to their internal technology conference. And from the highest levels of leadership, they were talking about how the importance of continual learning. And Steve Farley said, since 2011, we have been committed to create a culture of learning. Part of that is something we call Teaching Thursdays, where each week we create time for associates to learn. For two hours, each associate is expected to teach or learn. If they can't teach something, learn something.

If you can't learn anything, go teach something. And they have this internal technology conference where there was 1,800 people there where they had over 100 talks being given by internal staff. And it was just – to hear at the highest levels of leadership talk about the most important thing that you can do for the organization is to learn. I mean I think it's just – I think it is – in 10 years from now, I think that will be a part of every leader's vocabulary. But in the meantime, I think there's just such incredible competitive advantage in doing that. I think you see that a lot in the unicorns. In fact, I was at the Velocity Conference in 2011. It was in London.

And I was talking with a friend and we were saying, gosh, why are there so many people from Etsy and Netflix presenting? And then right behind me was Michael Rambetsi, who at the time was the director of operations. And so I turned around to ask him, Michael, why are there so many people at Etsy presenting? And he said, as a part of every engineer's performance review, they're required to – everybody's required to give at least three conference talks or write three blog posts or contribute to open source. And so that kind of notion of active engagement in the community and learning is actually baked into the performance evaluations, which I just thought was so amazingly neat on so many levels. I think one of them is that requirement really forced people to go to conferences and engage with the community. And I think the reason for that is a quote that I just love, this quote of, you're only as smart as the average of the top five people you hang out with.

And so when you're in a conference and you're hanging out with the best in the game, right, there's no – one of the inevitable outcomes is that you just get better, right? So that notion of continually – continual learning. That's interesting. Cool. I think, I mean, obviously these topics that we talk about now is very, I mean, actually moving away from a lot of technology towards, you know, people skills, which obviously we know DevOps in the is the biggest thing and then motivating people and kind of building a team culture that we all want to do better than we do it right now, meaning creating value for the company. I have another topic that I would like to touch upon, which is feedback loops. But the question is, Brian, I know we're probably again at the 30-minute mark, which is phenomenal.

Shall we do another quick break, if that's okay with you, Gene? Absolutely. All right. Yeah, that sounds like a good idea because, yeah, I think there's several more things we can – we can probably talk for hours. Yeah, yeah, yeah. So let's wrap up this episode. I'd like to thank everyone for listening again.

And again, if you want to follow any of us on Twitter, at RealJeanKim and at GrabnerAndy at EmperorWilson, you could always send any feedback through there. Or you can send an email to pureperformance at Dynatrace.com. We'd love to hear any thoughts, insights, or show ideas. Or if you want to be a guest, please get in touch with us. And we will be back momentarily

with Gene Kim.