Log in to watch

Log in or create a free account to watch this video.

Log in
London 2016
Share

Closing Remarks

Closing Remarks

Chapters

Full transcript

The complete talk, organized by section.

Gene Kim

All right. The other two things that we have learned since The Phoenix Project came out, the hidden subject is, I wish I had learned these things before The Phoenix Project had come out.

So the first one is that DevOps is a subset of dynamic learning organizations. And so when you listen to the way Ron van Kemenade from ING spoke in terms of knowledge acquisition models, the relentless need to create a learning culture, or even from Darren Haze's talk about SAP, about this relentless need to keep increasing performance and experimentation; Disney, the need for courageous leaders who are willing to change the way they work; Barclays, the need to create servant leadership; and John and Damon's talk certainly touched on this, is this notion of dynamic learning organization.

And for me, the best book on this is this book here. It's called The High-Velocity Edge by Dr. Steven Spear. And I took his course at MIT, his workshop, and I would recommend that to anybody, and certainly read the book.

And so his claim to fame is that he's credited for decoding the DNA of the Toyota Production System. He did this when he was a doctoral student at the Harvard Business School. And as part of that, he worked on the assembly line at a tier-one Toyota supplier for six months. And before Toyota would allow him to do that, he had to work on the assembly line at a General Motors plant for 30 days.

And so he went on to expand his work, going not just beyond Toyota, but the safety culture at Alcoa, the culture behind the U.S. Naval Reactor core, Pratt & Whitney engine design, many, many others. And it just shows that, I think it describes better than any other work, what enables dynamic learning organizations to always beat their peers.

And some of the key tenets are four capabilities. Capability one is you have to see problems as they occur. Capability two is we have to be able to swarm them in order to solve them, as well as create new knowledge. Capability three is there has to be some sort of mechanism that turns local knowledge into global knowledge so that we can elevate the state of the practice. And then capability four is that leaders must create other leaders, just because sometimes leaders die.

So if you're interested in that, certainly I would refer you to the book. And we had the privilege last year for Dr. Steven Spear, he gave the closing keynote at the DevOps Enterprise Summit, and his video is online. And we actually specifically talk about this in chapter four of The DevOps Handbook, the Third Way.

Oh, by the way, this is actually the first time where I've actually cited The DevOps Handbook, because now I even believe it's going to come out.

So Dr. Steven Spear is credited for the why. Why does the Toyota Production System work? But there's actually another work out there that was referenced by Jason Cox, which is: how does one create dynamic learning organizations? And the person who's responsible for this is Mike Rother, who wrote the book Toyota Kata. And I think he really describes the processes which leaders must put into place to encourage and create learning, to create these dynamic learning organizations.

So that's number one. DevOps is certainly a subset of dynamic learning organizations, just like Toyota, Alcoa, U.S. Naval Reactor Corps is. So that's lesson number one.

Lesson number two is, I think DevOps is also a subset of disruptive innovation. So I think we heard this in any of the organizations where they're trying to change the way they work. So whether it's Disney creating the culture of rebels, right, to be fearless in taking on the status quo. In terms of Hiscox and LV, taking on the largest part of their business portfolio, not starting at the periphery, but at the core. Zurich Insurance and HMRC, right? The HMRC I love just because they were doing work in a way radically different than the way they had done in previous decades and creating incredible benefits.

So the one resource I would point to for this is The Other Side of Innovation by Dr. Govindarajan and Dr. Trimble. They're in the School of Business at Dartmouth College, and they studied what they call disruptive innovation, and how difficult it is to do in large organizations that are successful, of which almost every organization represented at this conference is one.

So they studied the first BMW electric car project at BMW, the first online channel that was profitable at The Wall Street Journal, the radically different shoe at the Timberland company, the first specialty insurance at Allstate.

And so the concepts they talk about is that when we try to do work in a different way in large, complex organizations, we run into what they call the performance engine, or what we might call the mothership, right? Whenever you have large, successful organizations that have maintained success for decades, they are supremely great at daily operations. So I don't mean like IT operations. I mean all those business processes that must run, like product design, like order administration, supply chain, and so forth.

And so the way that you maintain great daily operations in general is process and bureaucracy and standards, right? And so you can take out half the bureaucrats and the bureaucracy will still win. It will still function, right? Bureaucracies are incredibly resilient. But that's exactly what we can't have when we want to do something as disruptive as DevOps.

And so their diagnosis, or their recommendation after having studied these efforts, is that you must have a separate, dedicated transformation team that is freed from many of the rules and regulations that would govern the performance engine, right? They're still held accountable to a result, but they can't be shackled to the existing processes. Really, their goal is to figure out the next generation of processes in order to help the organization survive and thrive.

And so some of the other tools that they talk about is shared resources. How does the performance engine and the dedicated transformation team work together? And one of the things I found so exciting was, what happens to the transformation team after they succeed? And in general, they are asked by the organization to help elevate the state of the entire practice.

And studying DevOps transformations, there's no doubt in my mind that this is happening in our space as well. We found out of the DevOps Enterprise speakers over the last two years, at least one out of four have been promoted at least once, some twice. And I think that's really reflecting this last point. They are now being asked to help improve the health of the enterprise as a whole.

So I think it's a phenomenal book, The Other Side of Innovation, and it's also chapter six in The DevOps Handbook, selecting our initial value streams.

So, is there anything on that? Oh, and now it's almost time to wrap up.

It's a shame.

Paul Muller

It's a shame.

Gene Kim

All videos and slides that you have seen throughout the last few days will be available online eventually. I think give us a week or two.

Paul Muller

A few people had that question, so yeah, they'll be sent through to you.

Gene Kim

Yes. Yep. Just like in the previous DevOps Enterprise Summits, all videos will be made online, unless if we were not allowed to, and all the slides will be online. You can go to that URL and expect an email from us in the next week or two, when we have gotten those posted.

Two is save the date. We are in the early stages of planning DevOps Enterprise Summit London 2017. So just bookmark all of June.

Paul Muller

Yes, exactly. It's that early.

Gene Kim

Right. It's not going to be 30 days. It'll be two days, probably. Anyway, so...

Paul Muller

Just set your expectation that we do plan on doing this again.

Gene Kim

And the third one is, also expect a survey. Like Erik says in The Phoenix Project, the opposite of love is not hate, it's apathy. Right? Feedback is love. So please, take some time and please give us some feedback, because feedback is love.

And then Paul Muller.

Paul Muller

Ah.

Gene Kim

Paul.

Paul Muller

What?

Gene Kim

So, what's been your primary impressions and surprises over the last two days?

Paul Muller

It's kind of hard to summarize two days, and in fact, it's hard to summarize, what has it been, like six-plus years, in many ways. But I think a couple of things jump out to me.

I think the first is, any of you who've spent any time in Japan know that when people give you a gift in Japan, they often actually apologize. They'll say, "Sumimasen," as they give it to you. And the implicit assumption is that the gift's not good enough for you. My gift isn't worthy.

And to hear so many people here who've shared incredible experiences, where they've slashed deployment times, they've transformed the way their business runs, to actually stand up in front of everyone and say, "Look, I've done this work. I'm sorry it's not good enough to share with all of you, but here's my humble offering." It's amazing, the humility in this group.

And frankly, I think combined with unrelenting standards, I mentioned this yesterday, everybody clearly has a desire, and the culture of this group is ultimately to keep pushing the envelope. I don't think anyone here seems like they're ever satisfied, and I think that as a cultural trait, I think all of the DevOps community tends to share that.

In terms of, I guess, other feedback or thoughts: multidisciplinary. The number of people who cited not only their collaboration between dev and ops, which is a portmanteau that's really handy, it kind of explains what we're trying to do, but the silo-busting culture between sales, marketing, security, HR.

I think John spoke about it great. Unfortunately, we can never use your acronym for it. What was it? BizDevHRSalesFinancialServicesDevOps. But I think ultimately multidisciplinary is another summary. I already created a Wikipedia entry for it, by the way, so.

But I'd actually challenge us as a group. We keep talking about business and IT like they're two separate things still. And I think one of the other observations I'd make is we're still kind of emerging out of our shell, and I think next year I'd love to see more sales, marketing, HR, operations people up here delivering talks, because if anything else, DevOps is a silo-busting... What did you call them?

Gene Kim

Boundary spanners.

Paul Muller

Mm. Spanners, not spanners. I thought you meant a tool at one stage.

And look, I think the last thing is just, frankly, it's the culture of sharing. Everybody here seems more than willing to help each other in a way that I think is incredibly profound. It's just people's desire to chip in and help each other is incredible, and it's great to be part of the culture.

What about you?

Gene Kim

For me, the biggest surprise is, even though this is the third DevOps Enterprise Summit for us, it just doesn't get old. Still now, there's no two or three days where I learn more than at this conference.

Paul Muller

Mm.

Gene Kim

And it's not just this. It's all the speaker interactions that led up to this conference. So it is my most sincere hope that you have also learned something that you think is valuable, whether it comes from Ron van Kemenade or the experience reports or the subject matter experts that we've invited, and that helps you on your own DevOps journeys.

As Paul mentioned in the opening remarks, Erik in The Phoenix Project said, "My goal is to change the lives of one million technology workers in the next five years." So who's actually going to make that possible, right? It's not Paul or me. It's certainly not Erik. Erik is a fictional character, by the way.

It's people like you here who have invested days to be at this conference, and people who've invested 15-plus hours preparing their experience reports.

And so after these sessions today, there's just no doubt in my mind that the economic value that will be created by transformations like the ones that we've heard throughout the last few days is going to create incredible economic value. It is implausible to me that it will be anything less than trillions of pounds, dollars, euros, et cetera.

And it's all because I think the goal is to make every one of those eight million developers and eight million operations people as productive as if they were working at a Google, Amazon, or a Netflix.

Paul Muller

Great point. So you mentioned it never gets old, and I think one of the reasons... Where's Patrick? Is he around? Where's your show of hands?

I think someone's buying him shots at the bar.

Gene Kim

I suspect there is.

Paul Muller

But when I first met Patrick back in Sydney a long time ago, one of the things we talked about was he was really worried about over-codifying DevOps, right? And getting too prescriptive with it, because if you've read The Goal, we all know you remove a constraint, and it simply moves somewhere else.

And I think this whole space is dynamic because as we're removing the code deploy constraint out of IT, we've suddenly got a business constraint or an education constraint. It keeps moving around, which means the problem always moves.

And I've got the same question for you, you forced everybody to answer, which is: what help would you like people to bring to you?

Gene Kim

By the way, when Paul mentioned that to me backstage, I was actually taken aback, and I actually needed about five minutes to come up with an answer. That's not fair.

Here's what I came up with.

Paul Muller

Now you know what it's like to be them.

Gene Kim

Don't like it.

No, no. I think my answer is: tell everyone about what you've learned here. Everyone needs to know about these heroic journeys that have been shared today and about what this community is doing.

And I guess, for me, that's why it's been such an honor to help chronicle this journey and help amplify these incredible stories that are being told, because it creates economic value, it creates organizational outcomes, but it's also good for technology workers, and it creates a more humane place for people to work and get joy out of the work.

Paul Muller

I think it's a great answer.

Gene Kim

So I think I'm now at the thank-yous.

So thank you to my fellow emcee, Paul Muller.

Paul Muller

Thank you to my... You did great.

Gene Kim

Jane.

Thank you to my fellow DevOps... What is that? Long day. DevOps Handbook co-authors.

Thanks to the programming committee who helped find so many of these amazing stories.

Paul Muller

Yeah. Thank you for the people who actually did some work, the speakers.

Gene Kim

Thank you to all the sponsors who made this event possible. Thank you to all the people on the IT Revolution staff and the event staff who actually made this event run smoothly.

And one more thank you. Many of you got a chance to meet my wife, Marguerite Kim. Thank you to her for essentially putting up with me for the last five years, when I've been on book deadline mode for five years. So thank you and sorry.

So, see you next year. Go home.

Woo!