Watch and read this 90 minute Ask Me Anything with Gene Kim, author of The Unicorn Project, from January 8, 2020. Recorded for the IT Revolution Book Club.
1. Jerreck (jump to question in video) — Have you considered publishing an “Oops All Crunch-berries” book where it’s just Dr. Erik Reid telling the reader what to do and why?
2. Megan (jump to question in video) — Have you considered publishing a 3rd POV for this timeline illustrating the perspective of business stakeholders like marketing and sales and fin – maybe a bit lighter on the tech speak? I’d love to use a book like that to help business leadership understand both the IT experience and how IT and Business can truly become strategic partners to deliver business value.
3. Roman Pickl (jump to question in video) — Which educational opportunities could you recommend in the devops space(certificates, MBA, PhD)? I read about The High Velocity Edge Course, which looks promising. I already have a master in business administration and a master in software development. What could be the next step?
4. Roman Pickl (jump to question in video) — In “Beyond the Phoenix Project” there is a discussion (I think it is in chapter 3) on how bottlenecks tend to shift after improvements and what to do about it (E.g. from getting environments to deployment, etc). Is there more info about this somewhere?
5. Chris Combe (jump to question in video) — Loved the book, question is: how do you see enterprise architecture fitting into the context of the book. The main character / protagonist became the first distinguished engineer and the only time I observed the architects piece was on the governance side that everyone wanted stamped out. Enterprise and domain architects could add value by playing a role in the product teams and adding value upfront rather than at the end when it’s too late. I’d like to see architects adding more value and getting on-board to add value.
6. Chris Combe (jump to question in video) — One more: the book was very positive once the big hurdles were overcome, I worry the challenges of the transformation post initial success were glossed over (room for a third book perhaps. I’d love to see how the transformation was embedded across the wider org in more detail, in large orgs like mine we have so many IT functions which are business aligned compared to a Dev (incl. infra engineering) Vs Ops world which seems quite simple by comparison. What metrics were used for measuring value and weighted lead times etc.
7. Roman Pickl (jump to question in video) —Why does it need a crisis to make change happen? Do we have to prepare in the underground and let it crash and burn before we can change things with the support of management?
This is discussed in:
- Mark Schwartz: At a seat at the table: Transformational projects occur when the amount of debt has become too much to bear
- Jez humble in Q&A: GOTO 2015 • Why Scaling Agile Doesn’t Work • Jez Humble
- John Cutler: Too Busy To Improve (for Roman) by John Cutler – 3-Minute Rambles • A podcast on Anchor what if you sit in the phoenix project at year 2 years, and know it will take another year to crash and burn before things change? Of course there’s always the option to change jobs…
8. Michele (jump to question in video) — In the company where I work we are embracing both Agile and DevOps methodologies. For me both the TUP and TPP books were eye opener, helping me to realize that we IT professionals all have the same problems.
Some of the team members, working in the “old way” for the last 30 years, are having really hard time embracing the change. Unfortunately they are also the resources with better product knowledge, so their input is really valuable when discussing how to bring value to our customers.
In case a solution does not exist, and you have to let one or more resource go, how would you suggest to do it without spreading “fear” to the whole team?
9. Dejan Mengues (jump to question in video) — I’m curious myself how to – in a more people-friendly way (read: avoiding people getting fired) – inspire the change bottom-up? Architectural Committees and groups like that are mostly (if not always) something that’s enforced bottom-up, not to drive the change but to stop it before it even happens. I’m curious myself what other people experience in going through this wall.
10. Fernando Bitti (jump to question in video) — My question to you at the moment: the book says Maxine didn’t check her phone during 2 entire weeks of vacations. How realistic is this in the companies you’ve worked in the past? Weren’t employees pressured to keep an eye on the business during their PTO or didn’t they check it anyway, regardless?
11. Andy Tinkham (jump to question in video) — The book touches on how the QA role is changing, but then Kurt changes roles and Adam stays a minor character. I’ve had lots of conversations with people in QA about how the field is changing, but I’m curious to hear from across a broader range of roles. Could you say more about how you see QA evolving, please?
12. Roman Pickl (jump to question in video) — I have experienced that moving from svn to git and github enabled something I did not expect at first at least not to that extent. Discoverability and fast fixes using a user friendly interface. And not so technical people now contributing via the online editor and pull requests fixing minor issues like typos and translations in seconds rather than days. Did you experience something similar? Are there other tools which revolutionized your work? (edited)
13. Rainer Hansen (jump to question in video) — Hi @Gene Kim, I like the book so far. I have read only part 1 until now. However, I am a little bit surprised that Maxime seems not to try to find out more desperately what the whole project is about in the beginning. She has worked mainly with the manufacturing system and less with the retail system before so this seems to be a new area for her. If I remember correctly she spent already in the very first days a lot of time drinking coffee out of frustration. I would have expected that she would have tried more to speak with senior people in the project to find out what all is about. Any special reason for not including that in the first chapters?
14. Proctor (jump to question in video) — What would be the list of books Maxine would hand off to Maia and Page (the middle school girls working in Python) or Tom after she had piqued their curiosity about Functional Programming for them to learn more at the entry level (real or even books that are missing that need to be written with what kind of content)
15. Adam Hawkins (jump to question in video) — Question from my friend and I: What organization size fits the “Distinguished Engineer” role?
16. Steve Elgan (jump to question in video) — Have you considered writing another POV from John’s perspective? Would be cool to see a book about transformation around infosec / devsecops type stuff. What did John learn from the plant safety manager? What were his 1:1 conversations with Eric like?
17. Ian Ceicys (jump to question in video) — The Second Ideal of Focus, Flow, and Joy – Why the term “Joy”? It feels strange to use the term Joy. Joy can’t be measured can it?
18. Sreeni Kotpati (jump to question in video) — What other conferences do you suggest for Engineering IT leaders?
19. William Judd (jump to question in video) — Gene: Did Maxine (you) get one of the first Ember mugs? That part of her story resonated with me. Example of excited engagement.
20. Theo Studer (jump to question in video) — How can you find out if there is a “Rebellion” inside your company ? If there isn’t how to start one?
21. David Levine (jump to question in video) — Gene – what is your perspective on How the Product Drive Org model applies to Infrastructure Services. Basically convincing a CTO group this product stuff applies to them and they are critical.
22. Richard Tarantino (jump to question in video) — As an agile based enterprise recently acquired by a monolity/waterfall based company, how can we prevent their practices from changing (regressing) ours. Or better, changing theirs?
23. Ryan Wakefield (jump to question in video) — Speaking of dashboards, I completely agree that they are fantastic, but you can easily go overboard on the amount of dashboards. How do you balance the quantity of dashboards you are showing verses the quality to not have information overload?
24. Anonymous (jump to question in video) — How do you go from showing interest/getting more involved in the platform while not dropping the ball in your day-to-day feature work?
25. Steve (jump to question in video) — Do you have a personal experience you can share where you decided it was better for you to move on rather than continue fighting the good fight?
26. Steve (jump to question in video) — Along the same lines, do you have any advice to those feeling burnout that look to your books for inspiration and fuel?
27. David Levine (jump to question in video) — Is there a link to the CapitalOne and CSG experience Reports from DOES?
Hi everyone. Welcome to the first AMA for the IT Revolution Book Club featuring “The Unicorn Project.” The author we have today needs no major introduction. I think we’re all very happy to have Gene here to join us.
Delighted to be here. It’s my first Ask Me Anything for the book club, it’s exciting.
Yeah, awesome. Just a couple things before we get started. Again, thank you all for joining the book club. We have about 800 of you joining, which is incredible. The discussions the past few days have been really insightful and thoughtful, and we’re really pleased with how it’s all taking shape. One sort of procedural piece, we added a new channel for unguided discussion. It’s #dockside. If you haven’t joined it, it’s in the pinned links in the first message that we put in the discussions channel. So you can just click that and join, or you can search for it and find it yourself.
The way the AMA is going to work today is people have been posting questions into the #ama_unicorn_project channel and so we’re going to work through those. If there are any followups that you want to add or any additional questions, put them right in there or use the Q&A feature here on Zoom. But, I think the best place would honestly be in the AMA channel, that way we have a log of all of those, and so if we get through the ones that we’ve already got, we will start working through the ones that are posted as follow ups.
So, without further ado, I’m going to pass it over to Gene. Is there anything else you wanted me to say before we start?
No, no. This is great. In fact, I hear it’s kind of just exposed my thinking, we have 90 minutes, and if we have a goal of doing 20 questions that would be about four and a half minutes each. So, I have a stop watch here.
And so I’ll time box it to four and a half minutes just to make sure that we get to some meaningful number of questions. And you’ve queued up 16 of them, which leaves some time for some more questions. Alex, does that sound good in terms of desired outcome?
That’s perfect, that’s perfect.
All right. Cool, cool. All right, so without further ado let’s jump in and start. And by the way, let me first thank everybody for all your enthusiasm and helping get the word out about the book. I couldn’t be more pleased about that reception. This is coming from a place of having spent three years on it. Basically the book launch consuming the last half year. So, my most profound and sincere gratitude to all of you. All right, so you’re here for an Ask Me Anything. Let’s start.
Question number one, Jerreck, “Have you considered publishing an “Oops All Crunch-berries” book where it’s just Dr. Erik Reid telling the reader what to do and why?”
If I were smart enough to, I would. I think the closest I’ve come to that is the “Beyond The Phoenix Project” work with John Willis where that was eight and a half hours of recorded audio. And that was 150 plus hours, probably more than that, and plus the recording, that’s like 200 plus hours of research that went into that work. I was using it all the time as I was doing the research for the book and then sort of since the “Beyond The Phoenix Project” came out there was a blog post, the Part One I put out about the resources for reading for “The Unicorn Project.” And that’s based on the citations and references in the book. My goal was to write a couple paragraphs on each one. That’s the closest I’ve come to and I would love to get to a point maybe in the future where I could further elaborate on it.
I guess the other thing I’ll mention is so much of the learning came from the DevOps Enterprise Summit through Dr. Steven Spear, Dr. Andre Martin, Dr. David Almina, and Dr. Cristina Maslach. And so I was often watching those videos again, taking kind of more pointed notes. And so if you want to follow in the foot trails of kind of where to find the authoritative sources for this, I would actually do that and almost in all cases I was putting those in the references blog post. So, anyway, I hope that helps and post a followup question if you want more, ask a follow up.
Awesome, two minutes. Alex, one down. If I could get in the Ask Me Anything, could you start a reactions thread and just maybe post an emoji, how are we doing here? Thumbs up, thumbs down, happy face, whatever. Feedback is always good, right?
Question number two, Megan, “Have you considered publishing a 3rd POV for this timeline illustrating the perspective of business stakeholders like marketing and sales and fin – maybe a bit lighter on the tech speak? I’d love to use a book like that to help business leadership understand both the IT experience and how IT and Business can truly become strategic partners to deliver business value.”
Yes! In fact, I think a bunch of you noted that I kind of left the door open for that in the epilogue of the book, where Maxine meets with Sarah for lunch. And it’s not what she expected, and that there will be future meetings in the future. One of the things that I plan on doing is recording a video or podcast with some friends of mine, who would love to explore the Sarah character in more detail. So yeah, there is… in some way, Sarah was a cartoon villain, right? With even more powerful friends in the Phoenix Project, I love that.
But Elizabeth Sullivan… Elizabeth I’m sorry Henderson. She’s one of my heroes, she helped pioneer automated testing and quality kind of in the 2000’s. She’s a VP at Pivotal for many years. And she asked me about Sarah’s background. We were hanging out for lunch and when she asked me that question, I was kind of thunder struck because I’d written personas and kind of character backgrounds on every character except for Sarah. So I sort of burst out laughing because kind of revealed to me kind of how shaky, kind of the Sarah character was compared to everybody else. And so I mentioned that maybe she came from a mergers and acquisitions background. I think she was a fantastic merchandising person, right? But I think she’s been burned by technology before and that’s why she behaves the way she does. We started brainstorm about kind of what her bookshelf was like, which was great fun. I think she loves strategy books, but she’s not a very people person. And I think her favorite book that she likes to give out is Who Moved My Cheese, which is a kind of a terrible book to give to someone.
Anyway, so yes, I actually anticipate in the next couple of months, being able to do a couple of hours of recorded material in terms of like really kind of speculating on why does Sarah behave the way she does? I’m just dazzled by the idea of like, what does Sarah’s perspective look like? Like what is her self talk? How does she perceive things that caused her to act in the way she does? Because I think that the feedback I’ve gotten really does validate that I think we’ve all had people like Sarah in our lives, people who maybe have good intentions, or maybe they’re genuinely bad actors. And so let’s neutralize a bad actor aspect and really see if we can have some compassion, and empathy, and understanding for Sarah like characters. Anyway, so the answer is, yes. If you’d like to follow up and brainstorm about that, just reach out to me. Why don’t you reply in a thread and DM me after you’ve mentioned Alex, and he’ll make sure that he routes me to it. All right, cool. Three minutes, 12 seconds. So on track.
Number three: “Which educational opportunities could you recommend in the DevOps space(certificates, MBA, PhD)? I read about The High Velocity Edge Course, which looks promising. I already have a master in business administration and a master in software development. What could be the next step?”
You remind me of a friend of mine in Seattle, who asked a similar question, he too had an MBA and a Master’s. He was asking what degree do I need? And I laughed because I don’t think he needs another degree, and listen I basis this on just your question. I don’t know anything about your background or kind of what your aspirations are. So take this with a grain of salt, but I feel like probably another degree is not what you’re looking for. And I think the problem with certificates is … I don’t think it will give you what you’re looking for. Although, I’ll tell you before I leave the certificates there’s one that I found very, very useful. And I’m not sure if I recommend to everybody, but I’ll just tell you what it is.
It’s a National Association for Corporate Directorship masterclass, I think it was like a $3,000, and is three days. And it was amazing. I mean, that was 10 years ago, back when I thought I wanted to be on corporate boards. And now learning more about it, that’s the last thing I want to do. The last thing I want to deal with is other people’s problems with no ability to actually do anything about it. But I mean, it was a fascinating glimpse in terms of just the breadth of the training and the materials covered were awesome. And just to paint the composition of the people in the class, they were all 50 plus years old. I was the youngest person there by at least 15 years. They were typically lawyers, accountants, auditors, all looking for what’s next in their career, looking for board seats. I thought it was great National Association for Corporate Directors masterclass. And I even have a link, I actually wrote a whole bunch of blog posts reviewing the course. Alex, let’s work together and I’ll get those links posted.
The Stephen Spear Certificate course was great, and there I think you’re going to fit right in. And in fact, I met a whole bunch of people in the DevOps community there or rather I met them there, and then we reconnected in the DevOps community. But I think these days, we’re in a frontier where it’s just what we need to know, it’s just not well known. I think if you look at the DevOps Enterprise Community of which “The Unicorn Project” is modeled after, inspired by, and dedicated to I think we’re all trying to piece together our bodies of knowledge. Because it can’t just take a course and learn what we need to learn. And I base this on as evidence, right? Just looking at the books that are on the bookshelves of these kind of incredible pioneers is often they have a ton of overlap, right? We’re all reading the same books trying to piece together what we need to learn. So I would actually say the best place to go, that is helping to find that frontier is the DevOps Enterprise Summit. That’s where I learned the most.
I measure the success of the conference by how much I learn and for two and a half years in a row it’s been our goal to ‘out program’ the last conference and I would say each year that goes by, for seven years, I’ve learned more and more, and the categories of what we need to learn, keep getting elevated. So anyway, that’s where I would go. And if you’re not at the conference all the plenary videos I think are worth watching. So those on the YouTube channel, Alex, let’s get those posted because all of those are carefully curated, and very carefully selected by me to teach me what I want to learn. And a community is more than conference, the community is a gathering of people with shared goals, and with the ability to help each other. And I think that’s one of the main reasons why I think people get value out of the DevOps Enterprise Summit, and why the program committee are willing to invest so much time in helping make that conference successful.
We just did a video with Jeff Gallimore, who co-hosts the conference with me, Jon Smart from Deloitte, and Dominica DeGrandis and it was amazing to hear them talk about their own aspirations that they’re trying to achieve, and how that get served through the conference.
I went a little over on that four minutes 45 seconds, but I thought that was worthy of a longer response. We’ll make it up later how are we doing so far? Emoji is positive. So that’s good. Okay.
Number four, Roman, “In ‘Beyond the Phoenix Project’ there is a discussion (I think it is in chapter 3) on how bottlenecks tend to shift after improvements and what to do about it (E.g. from getting environments to deployment, etc). Is there more info about this somewhere?”
The answer is yes. In fact, someone remarkably posted what exactly the constraints tend to be, and I think it might have shown up in “The DevOps Handbook.” The punch line is this, right, in general, here’s typically how the constraints move and this is really was demonstrated… I use this as a technique in “The Unicorn Project.” The initial constrain was deployment. Well, sort of… Anyway, you can’t get 10 deployments a day, if every deployment takes eight months, and you’re always waiting for things.
So the countermeasure is to automate those, and ideally enabled developers to do their own deployments. After that the constraint becomes testing. Wait no, it’s around environment creation, I’ll say it’s like environment creation first, right? Maxine definitely lived through that where she was unable to do anything despite her incredible unmatched skills without environments. She’s really unable to do work that she feels is productive, and rightly so. Then deployment, then testing. You can’t get to 10 deployments a day if every test cycle takes three weeks. So the countermeasure there is, automate those probably paralyze those so that the test rate can keep up with the ideal deployment rate. And then after that becomes architecture, right, we can’t be doing 10 deployments a day, if every change requires 12, 20 teams all to work, have to communicate, coordinate, synchronize, schedule, marshal, deconflict and so forth.
And so that’s because we don’t have locality, and so the goal is to create the kind of loosely coupled architectures where teams can work independently. And then after that, it really becomes how many good ideas do we have that are worth working on? And so that’s actually in a list I think in “The DevOps Handbook,” I saw that in the slack thread. And that absolutely matches my belief on the classic path, and that’s very much of what “The Unicorn Project” is modeled on maybe not explicitly, but kind of through the narrative, what the constraints are. Okay, two minutes 30 seconds.
Number five, Chris, “Loved the book, question is: how do you see enterprise architecture fitting into the context of the book. The main character / protagonist became the first distinguished engineer and the only time I observed the architects piece was on the governance side that everyone wanted stamped out. Enterprise and domain architects could add value by playing a role in the product teams and adding value upfront rather than at the end when it’s too late. I’d like to see architects adding more value and getting on-board to add value.”
Yeah, the answer is yes. So the answer is, I don’t know. There was a great talk by Levy Geithner at Target last year. No, I’m sorry, 2018. He’s now at US Bank and he gave a talk with his boss, Werner Loots, which was amazing. But I’m thinking about the 2017 presentation about next generation governance at Target. And there the role of high level architecture is really to categorize values into three areas and it’s all done in public, where their goal is to categorize technologies as ‘Avoid,’ as being actively deprecated. These are vendors that we’re actively avoiding and trying to stamp out, so that we can stop giving them money. It’s I think, maybe ‘Neutral’ and then ‘Recommended’, right?
Recommended means it has an active community of practice within the organization, there’s a lot of expertise, there’s a lot of battle tested libraries and components and vendors have been using anger. And if you need help, you can get help.
So that’s one thing and then I would recommend you would see any other talks by Dr. Mik Kersten at DevOps Enterprise Summit, CEO of Tasktop, he wrote the “Project to Product” book, which created the Flow Framework, and just to see how he talks about architecture, he says ‘ an architect’s only job is to increase flow,’ right? Any other kind of measure of what the architect’s job is in the way, and when he goes through it, it’s kind of laughable, right? How many things can you reduce things to or is it how many boxes on Visio diagram? It really is the only purpose where it’s being served is to increase flow, maybe to increase reliability and decrease risk, right? So the risk category and the risk flow framework.
Flow framework. Feature the four categories of work, mutually exclusive, completely exhaustive, features, defects, risk, and debt. And so, architecture is the opposite of debt. Architecture is there to advance the ability to deliver features quickly, decrease the number of defects and reduce the risk to the organization. I think, just hearing how he talks about how he integrates the role of architecture into his organization is awesome. One example that he gave was across all their product teams, they never had a canonical work item ID, and so that means every team had to deal with that. And that really should have been a cross cutting concern extracted from all the feature and product teams. By doing that, that help feature flow immeasurably. So I thought that was just a great example. I hope that helps. And if you’re interested in following up on that, we can arrange an introduction to Mik Kersten, right, Alex?
Yes. All right, three minutes, 45 seconds, a little bit behind pace.
Number six, Chris, “One more: the book was very positive once the big hurdles were overcome, I worry the challenges of the transformation post initial success were glossed over (room for a third book perhaps?) I’d love to see how the transformation was embedded across the wider org in more detail, in large orgs like mine we have so many IT functions which are business aligned compared to a Dev (incl. infra engineering) Vs Ops world which seems quite simple by comparison. What metrics were used for measuring value and weighted lead times etc.?”
A great question and definitely something that I just couldn’t go into as much detail. And actually, I guess I don’t really know. I mean, I would say if I were to learn, I would actually look at the case studies that are coming through DevOps Enterprise, through the repeat experience reports like Capital One. I think they have 14,000 engineers right now, that’s up from 8,000 two years ago.
I said just the scale of these organizations, right, the fact that they broke up the centralized ops and infrastructure function and move those into the four business units. What are the other… CSG, I think is probably the most well documented in terms of that. But there’s maybe not as large as some of the other organizations, but yeah, I would look at the repeat experience reports and if you’re interested in that, follow up with me and maybe we can actually work on getting that story told and specifically ask a repeat DevOps Enterprise speaker to talk more about that. I think that would actually be fascinating. Let’s learn together, reach out to me, Chris. All right, two minutes.
Number seven, Roman, “Why does it need a crisis to make change happen? Do we have to prepare in the underground and let it crash and burn before we can change things with the support of management? What if you sit in the Phoenix Project at year 2 years, and know it will take another year to crash and burn before things change? Of course there’s always the option to change jobs…”
Yeah, I think this is actually really well discussed in… is it “Leading Change” by Dr. John Kotter one of the prerequisites is have a burning platform, and in the absence of a burning platform, it’s really difficult to motivate large scale change. I mean, as you probably all know, if you looked at the resources guide for “The Unicorn Project” they had a burning platform. I mean, they were probably hundreds, if not thousands of people working on a rebellion that was just not effective and led to the ultimate demise of Nokia, where it was sold for $7.2 billion, a shadow of its former self formerly valued at 100 billion dollars. It was soundly eliminated in the marketplace by the Apple iPhone.
And so they had a burning platform, there was a famous Stephen Elop memo from the CEO talking about the burning platform, they still didn’t make it. So I think it just says without a burning platform… even with a burning platform, these changes are hard, and so imagine how difficult it is without one. I think that’s the reason why these rebellions are so effective because they have the information and intelligence networks to be able to find burning platforms, so that’s why I think Kurt is always actively looking for people who need help. They don’t need convincing to work in a different way, right? They’re the people who are immediately poised to say thank you when you help them. And so it means that you have smaller wins, but at least they’re incremental, and you’re solving problems that people do appreciate. And so eventually that earned them the right to work on a real burning platform when they were able to work with Maggie.
So that reminds me of like one of the best piece of advice I got from one of my mentors. His name was Ari Balogh, he was a CTO at Yahoo and then worked on the storage infrastructure at Google. He wrote the forward to The Visible Ops Handbook. He was formerly the CTO at Verisign that ran the dot com DNS registry. This is a one of the engineering miracles of the 2000s. Anyway, I mean, they built cloud scale infrastructure before there was cloud. Anyway, when I asked a similar question, his advice to me was you don’t choose a gene. When I asked him, how do I make people care? He said, You can’t make a moral issue out of it. That advice stuck to me for 15 years, right? And I think what he meant by that is, you can’t get people to care about something they don’t care about, right? You can’t make a moral issue. It’s like trying to convince someone to change their morals, right? You have to wait for them to actually have a sense that they have a problem before you can offer a solution. So great advice from Ari Balogh. Three minutes, 30 seconds.
Next, eight… Actually let me check in the Ask me anything channel for the emojis. How are we doing Alex?
Good. Okay, I’ll take your word… oh wait, I’m in the wrong one. Ask me Anything Unicorn Project. More emojis, please. Okay. All right.
I’ll pin the message too, so you can add to their reaction message.
Thank you. All right.
Number 8, Michelle, “In the company where I work we are embracing both Agile and DevOps methodologies. For me both the TUP and TPP books were an eye opener, helping me to realize that we IT professionals all have the same problems. Some of the team members, working in the “old way” for the last 30 years, are having really hard time embracing the change. Unfortunately they are also the resources with better product knowledge, so their input is really valuable when discussing how to bring value to our customers. In case a solution does not exist, and you have to let one or more resource go, how would you suggest to do it without spreading “fear” to the whole team?”
Yeah. Great question. So this is something I had to go through a couple of times during my days at Tripper, couple of times, way too many times, right, because we raise too much money, we could never hit our revenue targets.
And the problem was is that we went through a period where we just never were honest with ourselves about what revenue we could actually do. And so what led to was kind of… I think we did two reductions in force per year, four years in a row and that was terrible, right? Because I think this is what led this dynamic that you’re talking about Michelle where there was a lot of fear, and was valid, right? Because there weren’t decisive steps taken to actually fix the underlying problem. And so we actually had a new CEO come in Jim Johnson, he was amazing. He was a former site manager at Intel. Sounds like it’s a pretty humble title, but essentially, that was the general manager for all the non P&L responsibilities for Intel for the entire state of Oregon. So he was responsible for 3,000 to 6,000 employees. So he was amazing. I mean, so we did one last reduction of force. And he set very explicit targets about what we need to do in order to make a buck.
In fact, that was… somewhere around here I have a Lucite plaque that he gave to the entire company when we made our first dollar, it was our first profit, Tripwire. And it was just amazing, right? I mean, that was sort of the brave leadership we needed. So that we cut once and we did it with a plan of how to turn our first profit. So that was what I was trying to allude to or suggest in the scene in The Unicorn Project when they did a reduction in force because of Sarah and the hostile board member. And really making the structural changes necessary for the long term viability of the company, and they would only have to do at once. Yeah, and by the way, the other question that’s in the… Michelle, that you were asking is like, how do you get these kind of conservative leaders on board and that was really my aspiration for “The Unicorn Project” is just say, hey, look, there’s something about the way they’re working that is just dramatically not working. And here’s a way to evidence it.
And really the mental model for the book for the first third of the book was like Robinson Crusoe or Tom Hanks in Castaway, right? You have Maxine who’s exile into the Phoenix Project who’s at the height of her skills at the height of her profession, and yet cannot get anything done so very much like Tom Hanks, right? She can do nothing by herself. And I think the goal was to really kind of convey to development managers, QA managers, engineering directors, and so forth that this is what it feels like to be on these teams when you just can’t get anything done, right? Just like Tom Hanks couldn’t do anything that he wanted to do. So hopefully, that book was effective and is effective in conveying that story. Right, four minutes.
Next question, nine, Dejan, “I’m curious myself how to – in a more people-friendly way (read: avoiding people getting fired) – inspire the change bottom-up? Architectural Committees and groups like that are mostly (if not always) something that’s enforced bottom-up, not to drive the change but to stop it before it even happens. I’m curious myself what other people experience in going through this wall.”
Hang on let me just study this question a little more. So, how do you avoid people getting fired and still kind of allow bottom up work? Yeah, I was really inspired by the talk from Ron Van Kemenade DevOps Enterprise. So he’s the CIO of ING. So back then there were 8,000 engineers and I think like Capital One they’re probably much larger now. And so there’s ING bank, and one of the questions I asked him to talk about was like, what do you do if you can’t get permission? Yeah, he said something just so remarkable, he said, I expect every leader to have the bravery to do experiments, right?
So if that means hiding capacity from your product owners, you do it so that you can actually generate a win and have a business case to do more of it. And so the birth of the DevOps at ING bank, in his words, were we used to be like Parts Unlimited. Our mobile app was the worst rated app in the App Store. The only reason why we weren’t worse is that you couldn’t actually give an app zero stars that line came straight from Ron Van Kemenade. And the birth of a new way of working came from a development director who said what we’re doing is insane, right? Here’s what I want to do. And he wanted to do an experiment, where working in this new way that doesn’t involve… it’s not RP, what do you call them? Like requirements documentations I used to write them all the time at Tripwire. And he asked for one team and Ron Van Kemenade he said, no, I’m going to give you three teams, right? Because he was such a believer.
Anyway, that was one of the best talks. And I’m really hoping that I can get him to speak this year at DevOps Enterprise in London. It’s just one the most inspiring talks I’ve seen and I think in a just world, every leader would have a leader like him. And I think the results he’s been able to generate speak for themselves. Anyway, and to answer your question. Yeah, I think by doing experiments from the bottom up you create wins, and I think the reward of doing that is you get to tasked to do more of it at that bigger scale to make even bigger impacts. And people won’t get fired, they get assigned to you. So I think that’s great, so have no fear, just do the right thing as modeled by the rebellion in “The Unicorn Project.”
Number 10, Fernando, “My question to you at the moment: the book says Maxine didn’t check her phone during 2 entire weeks of vacations. How realistic is this in the companies you’ve worked in the past? Weren’t employees pressured to keep an eye on the business during their PTO or didn’t they check it anyway, regardless?”
Yeah, I think that’s a hallmark of older ways of working, these unjust cultures that are trying to grind and extract every bit of energy that we have. And I’m laughing, there’s nothing really funny about it, it’s terrible, right? We don’t want that. But I mean, I think that’s kind of the way that all too many organizations manage technology people, we just feed them more pizza and extract every piece of energy out of them like in… what is it? The Matrix, we’ve turned people into batteries, right? And I think in kind of these organizations that we admire, it’s the opposite, right? We want to protect those people so that we can always get the best out of them, right.
At Etsy, one of the missions they have was monitoring people sleep, because they wanted people rested. They want to make sure that on call rotations were not only justly divided, but people were actually able to unplug, that people were able to go on vacation, and go on holidays that there was enough on call support, and rotational support so people could go on vacation. I think you’ll find the same thing at Amazon, and Facebook, and Netflix, and Google, and Microsoft. So I’m not denying that the opposite exists. I just think that… it’s just obvious to me that in 50 years, we’re going to look back and say that was savage unacceptable behavior and just not very good business. If we want to win the marketplace, we can’t do that, if we treat people like that. So that’s actually one of my near term aspirations is really, how do we surface that to business leadership.
And let me just… Alex, if you could take a note maybe that should go part of the into the Sarah point of view. Maybe that will be one of her aha moments. So yeah, excellent question. So I would look at kind of on call rotation at Etsy, there’s a kind of a thread there of like what people do to make sure that people don’t get burned out, have adequate coverage on call coverage and so forth. PagerDuty has done some great work there, and hey, write about it. And I think it will be widely referenced, including by me. All right, two minutes 45 seconds. So Alex, I’m going to leave it… here’s a request I have for you is, could you just tell me when people are sort of posting positive emojis, and more importantly, negative emojis?
Question eleven, Andy, “The book touches on how the QA role is changing, but then Kurt changes roles and Adam stays a minor character. I’ve had lots of conversations with people in QA about how the field is changing, but I’m curious to hear from across a broader range of roles. Could you say more about how you see QA evolving, please?”
Yeah. Great question. So there’s a guy named Adam Auerbach at Capital One. I had mentioned that operations and infrastructure got decentralized into the business units. And Adam Auerbach owned the last centralized kind of function, which was QA. And he knew that it too would be decentralized and moved into the business units. And he was preparing for it. He left the organization to take an even bigger role at another bank, but I mean, he had a sense of equanimity. And his job was to make sure that QA objectives were defined in a way that made sure that QA engineers had a ladder.
So I have like three things in my mind. One is Microsoft… I met a person with Sam Guckenheimer… what was his name? He was in the Bing search group. And he was a part of the group that basically reduced the number of job functions at Microsoft from what they call the chair four pillars to three. What was it? It was Program Manager, developer, QA person, and infrastructure. And they eliminated the… I think they called it E-STAT role software development engineering test. They eliminate that role and as a job category in the HR systems, and basically made everyone a developer. And it was just this amazing kind of… I think, it pre-stages what’s coming just like it’s another I would say very much in the same category of what Adam Auerbach experience at Capital One. Alex, if you could take a note, should ask them to talk at DevOps Enterprise talking about the future of QA because based in real experience.
But let me tell you about the big aha moment that I had like that sort of made all of this kind of feel right to me. So I mentioned Elizabeth Henderson, her Twitter handle is @testobsessed in the context of the Sarah discussion. She helped pioneer kind of what next generation QA look like, and she told me about a story that came out of the Los Altos workshop for software testing. So they ran like 50 plus workshops in the 2000s, that they invited the best QA people in the game, right? The best consultants, the best people in… No, maybe there were no consultants allowed, they had to be building products, maybe… Whatever, right? It was the best QA people in the game. And so they did a whole bunch of workshops. And one of the exercises they did in one workshop was what was the dev to QA ratio of your best project and your worst project.
And so in people’s worst projects, she said something troubling kind of came up, it was the highest dev to QA ratio, some were closest to one-to-one, right? In other words, the closer you got to one-to-one, or the more QA people you have as a percentage relative to developers, the worse the outcomes, right? That was the hallmark of their worst projects. She said even more troubling, was like what people’s best projects were, and they almost all had no testers. So this is like very prone to misinterpretation. But it’s astonishing, right? She said the learning was that if everybody knows that there’s no one out there, no department who are going to catch your errors for you, suddenly you care a lot about testing, which I think is this amazing observation, right. And this was actually validated in the state of DevOps report where the more that the automated test library is owned by an external group outside of developers the worst the outcomes, right?
So really QA is a job of development. And so the job of QA becomes less about testing other people’s code, it really becomes about coaching and consulting and helping developers test their own code. And then there’s another great talk I’ll point you towards, which is from Rohan Singh at DockerCon 2015 about how their code quality at the unit level coverage got so good that the majority of arrows were coming through kind of integration problems where it was spanning different components. What they did was created the entire Spotify… He was a dev lead at Spotify, so they created an entire Spotify environment inside of Docker containers. And basically, the dev teams required to write and run normal unit tests, but write and run integration Test as well. So moving integration testing left, I thought that was just an amazing talk.
So, Alex, yeah, that entire topic should be a focus for DevOps Enterprise. Let’s do that. Five minutes and 30 seconds. Definitely went over on that. But I thought it was such a meaty topic. It was worthy.
We’re getting reactions somebody says she’s wearing the socks, we’ve got praise on the pacing.
Great. Good, good. Okay. Thank you for the feedback. As Bill said in the Phoenix Project, right, the opposite… wait, feedback is love. Absence of feedback is apathy, something like that. Okay, what’s next?
Question number 12, Roman, “I have experienced that moving from svn to Git and Github enabled something I did not expect at first at least not to that extent. Discoverability and fast fixes using a user friendly interface….” Yeah, man, GitHub is phenomenal at that meaning amazing interfaces, “…And not so technical people now contributing via the online editor and pull requests fixing minor issues like typos and translations in seconds rather than days.” Yes. In fact, Capital One talked about lawyers and legal counsel, putting pull requests for other people’s repos and fixing license agreements. Amazing. “Did you experience something similar?”
Yes. “Are there other tools which revolutionize your work?” Oh, by the way, yeah. In fact, I was talking to some friends at GitHub, and they said, yeah, the finance people like doing revenue recognition. This was about five years ago, I should follow up on that. Alex, let’s follow up on that where people in the finance group were actually making kind of delicately code changes to I would assume the revenue recognition areas or something, right. Yeah. So I think this is really kind of showing kind of code moving further into the business areas, which I think is great. I think you see this a lot, which I think is interesting with designers, right? Where you get a better integration between developers and designers, they’re working on kind of like CSS elements, and graphics, and so forth where they’re using the same tool chain. And that’s a very different style of work than what I was used to 10 years ago where every handoff between design, and graphics people, and development would take forever.
And so business functions. So the question is, what are other sort of like revolutionary things? Oh, yeah dashboards. In fact, I was just watching a presentation on dashboards being kind of one of the best cultural change tools and in fact, I even have a browser window open. I’m going to post it while I… In fact, where did I get this post? I think it was from the Slack channel. Stand by. Alex, I’m going to DM it to you so that you can repost it as a right reply in the right place. Yeah, and it’s basically saying that kind of before you start telling other people how to do their job differently, post a dashboard up, so that people can sort of see problems, right? And it actually forces the discussion around how to actually generate accurate data as opposed to denying that their data is actually inaccurate. It’s called Dashboards and Culture: How openness changes your behavior by Steve Pool at IBM. I think that’s another one, right?
And so with things like hosted Graphite, where it’s a tool where you can get a dashboard up and running in a minute, I mean, it’s just really great. I use that all the time. It’s a little spendy, if you’re tracking a lot of metrics. But in my mind there’s no… I have found it’s been the easiest way to get the dashboard up and running. We actually generated one right before one of the state of DevOps report surveys because I just didn’t want to. And we actually were able to get one up and running like the night before launch, and it was great. So yeah, it saved my butt on a whole bunch of times. Anyway, so hopefully that helps.
Question number 13, Rainer, ” I like the book so far. I have read only Part 1 until now. However, I am a little bit surprised that Maxine seems not to try to find out more desperately what the whole project is about in the beginning. She has worked mainly with the manufacturing system and less with the retail system before so this seems to be a new area for her. If I remember correctly she spent already in the very first days a lot of time drinking coffee out of frustration. I would have expected that she would have tried more to speak with senior people in the project to find out what all is about. Any special reason for not including that in the first chapters?”
Yeah, I think show how explicit I was. But here’s my mental image of her first couple of days is she’s just trying to get a build running and no one has time. And she goes shopping for meetings, trying to find interesting meetings to go to, but they’re all kind of all mired in the features and she doesn’t care about that, right? She doesn’t want to learn about features she wants to understand… I love this model. This came from the Kersten, this kind of three levels of categories of things that developers work on. One are the features, the second are the kind of back end… the features, right? That’s the features on the website or the mobile app or whatever, right?
Anyone can get budget for those because you can see them, then you have the back end API’s and those are tougher to get. Those are systems record and so forth, and those are tougher to get budget for because no one sees them and no one values them, they’re maybe running on 20 year old mainframes. And they like to modernize, but they can’t get the budget for it. Third, are the systems that developers use in their daily work, right? The CICD pipelines, and the dev productivity tools, and the dashboards and all that stuff. And so Maxine has no interest in the first, right. Wait, so what’s the point? So anyone can just get budget for features, right? And typically who gets assigned to the dev productivity tools are the interns, right? So you have the best people on features, the people not good enough for features on systems record, and the summer interns on dev productivity. But in the tech giants in Facebook, and Amazon, and Google, it’s the exact opposite. They put their very best people, people with PhDs on dev productivity they have…
Google has 1,500 engineers working on dev productivity, some people invented programming languages, the basil build tool, right? Of which PhDs were created from or at least papers of people getting PhDs cite. Then you have the next best engineers working on platforms. And then the other developers on features, right? So they invert the priority. So Maxine understands that, right? So that’s why Maxine only cares about things that aren’t features. And that’s why she’s trying to shop around for and find people who care about platforms. And that’s why she’s not really hanging out with feature teams and she can’t find anyone who cares about it. So that kind of explains maybe what she’s seeking, why she’s not learning more about feature teams, and why she’s so adamant that she does not want to get sucked into feature work. Even though I think she would have found it… it would have felt like the work was more productive, but she knows that will take her not in the direction that she wants to go in, and the organization not in the direction they need to go. Two minutes 30 seconds.
Question fourteen, Proctor…Holy cow. Proctor We did a podcast together he runs a fantastic Functional Geekery podcast. He invited me on, and I’ve been a fan of his podcast. Anyway, it’s delightful that he’s in the channel and Alex let’s post a link to Functional Geekery podcast.
“What would be the list of books that Maxine would hand off to May and Page, the middle school girls, working in Python or Tom?”
Oh, I misread the question in this Slack channel… or Tom, right the data hub engineer after they pique their curiosity on functional program for them to learn more at the entry level? Yeah, so I don’t know about May and Page, right, the middle school girls. I think functional program is something that’s… I actually DM Proctor a link to a great quote about someone trying to teach functional programming to kids using kind of like acting out code. But it was great, and I’ll post that link, Alex.
But I mean, his observation was that maybe imperative programming is better for teaching the kids just because there’s more stuff to do. I mean, there’s only so much you can do with map filter and reduce functions, which I thought was kind of a funny thing. So yeah, I actually do have books I would recommend to Tom, right, the Data Hub Engineer, and that is in the Love Letter to Closure blog post. And so I think I recommend function… these are all closure books, which might be a tough haul for people. But one of the books that… I’m kind of looking for on my shelf, it’s specifically targeted at Python, Java and Ruby programmers that I thought was amazing. That was actually the first book I would recommend to people like Tom. And in the C Sharp community, I think the F Sharp is a phenomenal… by Scott… He has a blog site F Sharp for fun and profit. Alex if you could post that, he has a phenomenal series of books and videos. And that’s just a great on ramp for C Sharp programmers.
And there’s probably an equivalent book for Java people who don’t want to go to closure because too alien, maybe they can look into Scala. I don’t know what the right book for that is. Proctor I would love a recommendation from you. And I think in my mind, it has to be a book that sort of closes the imperative escape hatches, right? We don’t want to have people make the same mistakes in Scala that they were making in Java, right? That’s a total waste of time, and I think that happens all too often. So that is worthy of a reply in that slack thread Alex, so why don’t we take the actual time to do that later. Okay, three minutes. I’m going to pause for water. And scanning the slack channel for a second while I… Okay, from Adam number 15. How are we doing on time? I can’t see a clock 09:30, we have 35 more minutes left. Is that right or no?
Wait an hour left. Wait, help me. I have no idea.
Okay. How we doing Alex in terms of progress?
Great. We’ve got about six, and we’ve got some extra ones coming in through the Q&A, and I’m adding them to the list as we go.
So 30 minutes times 12 more questions is six more minutes, six minutes for question, right? We’re good, right?
Yeah, let’s keep the pace we’re going just for anything else that comes in. I think we’re good.
All right. Okay. Cool. It’s nice to have a forebrain on the call to everyone… Alex, is serving the same role as Jeff Gallimore plays at DevOps Enterprise where its so intense, sort of the Ops role, so that someone who has a very loose grasp of time and space. It’s always nice to have someone maintain situational awareness, even though that word is not allowed in the safety communities, but I can attest to the value of that. Thank you, Alex.
Where am I?
Question fifteen, Adam, “What organization size fits the “Distinguished Engineer” role?”
I honestly have no idea. I’ll say here, so I think the distinguished engineer role comes from companies where technology is viewed as a core competency. So that comes from like software vendors is probably where you’re going to find them or hardware vendors. They have distinguished engineers like Verizon, they have distinguished engineers and technical fellows, but typically those people are like the radio specialists, right? But now those companies, Verizon also has their first technical fellows. Ross Clanton joined as a technical fellow, I think that’s freaking amazing. And they created distinguished engineers because software is a core competency. So I would imagine Topo Pal, Dr. Topo Pal at Capital One was the first technical fellow. So at what point and so those organizations have at capital Ones, they had 8,000 engineers at the time. How many does Verizon have? I don’t know, we can look in the slide set.
Alex just pull that from the latest Verizon presentation from DevOps Enterprise Vegas, post how many engineers they have. I guess if you have 5,000 plus, maybe it’s time. Jeffrey Snover, technical fellow at Microsoft, they have 11 and he said something interesting. It’s like one technical fellow… statistically revenue of Microsoft, I think was like 20 billion, and he said, we have 21 technical fellows over 11 are individual contributors. He did the math, and he said, it was basically a billion dollars per technical fellow. So you want every technical fellows sort of carrying that kind of weight. So anyway, that’s Microsoft, I suspect they’re kind of on the high end. But I think that would be kind of interesting, right? I like the idea of like, one technical fellow per x thousand engineers, and maybe it’s hundreds, I don’t know. Or it’s across some sort of multiple, I don’t know. But maybe that’s where they out study.
Alex, let’s take a note on that, and we can ask the technical fellows, we know, that’d be really super interesting. Thinking about Suzette Johnson from Lockheed Martin, Robin Yemen from Northrop Grumman whole bunch of technical fellows. Cool. I’m sighing not out of frustration or boredom, just taking a deep breath.
Question sixteen, Steve, “have you considered writing a point of view from John’s perspective? It be cool to see about transformation around InfoSec, DevSecOps stuff. What did John learn from this plant safety manager? What were his conversations with Eric like?”
So yeah, there was a whole bunch of feedback I got that the John should have played a better role at information security, should have played a bigger role. I did the best I could, there’s a problem that happens in movies that you can only have so many number of characters before they become nameless faces. You see this problem happen a lot in like World War II movies like war movies, where like a Platoon, a squad is 10 people and how do you make the characters different enough so you can actually remember them and care about them?
So that when they die, you actually feel sad. So I was having that problem with “The Unicorn Project,” and so I shrinking the number of characters, especially in the rebellion. So that’s why Shannon had to do double duty, both as a security person as well as the data person, which I actually loved. Yeah, to your question. Yeah, I think that’ll be interesting. In fact, one of the things I really want to do is find some really great security officers and I actually put a couple feelers out. This was like, what do you do when you have data available everywhere? Right, yeah. Developers they said the goal was to get that into the hands of developers, and teams, and business people so they can use in daily work, but then how do you be responsible about it so that you prevent sort of unauthorized uses of it that are… and unauthorized meaning kind of like it’s not allowed by contractual agreement, Terms of Service, licensing agreements and so forth.
So how do you simultaneously empower the edges with data and yet, be responsible for it, enable discoverability, and so forth. I think that’s just an amazing right topic that I don’t think it was well explored, so that’s something I would love to do. Maybe not top of the list, but if you have interest in that would love to… let me know, and maybe we can create a team to explore that.
Question seventeen, Ian, “The Second Ideal of Focus, Flow, and Joy – Why the term “Joy”? It feels strange to use the term Joy. Joy can’t be measured can it?”
So I’ve done a couple of presentations on this, and Alex, I did post the Portland DevOps Days presentation, but instead of posting that one, there are better ones out there now. I think I did one that’s an hour long at Heyo that will be posted soon. So let’s replace that eventually, when that’s available, Alex on the blog post. And if it’s available, post it. That was the one I did in Australia six weeks ago.
And so I talk about the work of Dr. Csikszentmihalyi. So he wrote the book, Flow: The State of Optimal Psychology. And it’s really about the work he loved, and he called it flow is a state where you’ll have so much fun and fulfillment out of the work you’re doing, you lose track of time and maybe even sense of self. That’s a very transcendental experience. And maybe he used the word joy, but I use the word joy because that’s the joy I get out of coding now after learning closure, that’s what I was putting in the love letter closure is this is what got me into technology. I love coding. And so joy seemed to be the best word for doing the work I want to do and not having to struggle with build problems, and secrets management, and infrastructure problems. Those are all work I used to value, used to love doing. But now I just hate doing because it takes me away from the joyful work that I love.
So yeah, joy just seemed like a very… I could think of no better word than joy. And, yeah, by the way, I would recommend that TED talk to anyone. I think it’s the best TED talk of all time. And I just think it beautifully frames kind of what platform and engineering… what platform to do, right? It’s enabled development productivity so that we can solve business problems, the problems that we want to solve, and not get mired down into all the stuff that we don’t want to solve or go to meetings or get approvals and so forth.
Question eighteen, Sreeni, “What other conferences do you suggest for engineering IT leaders?”
Let’s see here. So I had mentioned DevOps Enterprise Summit and there’s a blog post coming up maybe later today. And the title of the post is how to get your submission accepted. But it’s a 4,000 words blog posts about what the objectives of the conference are and how much care we go to make sure that the conference objectives are served.
And I just say that just because we are really trying to help technology lead to succeed in a way that… I had no other conference that does to the extent that DevOps Enterprise program community and community does. But the other conferences I’ve learned a ton from are the velocity conference, and now they call it I think, the O’Reilly operational and infrastructure conference or something like that. They renamed it this year, and I always love the stories of like how we saved Pinterest or how we save Twitter from collapsing in on its own weight. Just those transformation stories are leadership stories, as well as feats of incredible engineering achievement. Those are the talks I think we all learn from, and I think that’s why the DevOps Enterprise conference really 75% of talks are experience reports, because I think that’s how we learn. So the other type of conference, we see those but nowhere near the proportion that you find a DevOps Enterprise are like, velocity. Another one is QCon, and Go-To Con, those cater more for developers.
I haven’t been to one of these, but I want to go to one, hopefully, this year is the architecture O’Reilly conference, and they actually have a data conference as well. So those are all ones that… In fact, I met those people at the DevOps… For year one of the DevOps Enterprise Summit, so many of the speakers actually came from people I met at those conferences, so engineering leaders I would definitely go to them as well.
Question nineteen, William, “Did Maxine (you) get one of the first Ember mugs? That part of her story resonated with me. Example of excited engagement.”
Yeah, no, sort of. Dr. Mik Kersten “Project to Product,” right? He actually gave me and my wife Marguerite an Ember mug as a gift. We thought they were amazing, and so yeah, it definitely channeled kind of imagine what the Ember mug experience might have been if we had been through Kickstarter. I don’t actually know what the origin story of ember.
But yes, that was absolutely modeled after the ember mug, and I think I do mention that as such in the resources guide to “The Unicorn Project”. If not, Alex, we should probably fix that or something, that should be referred to.
Question number 20, Theo, “How can you find out if there is a ‘Rebellion’ inside your company ? If there isn’t how to start one?”
So there’s a tweet out there that blew me away. Alex, we should post a link to it. It said “I love “The Unicorn Project” and I’m starting to notice a whole bunch of people with rebellion stickers on their laptop.” So I think that’s amazing, right? So John Willis says something really interesting in “The Unicorn Project”. He said, the presence of the Phoenix Project is an incredible signal of… the executive’s office an incredible signal of kind of the value system. It’s a likely indicator that they recognize problems in the Phoenix Project, and probably care enough about it to do something about it if it’s sitting on their desk, or on their bookshelf.
I would say “The Unicorn Project” should be another marker, that they understand these problems and hell, if they have a rebellion sticker on their laptop, I think that’s a very strong signal, right? So there’s actually, in that Twitter thread that we’ll post there’s actually a link to an Amazon product page, where you can actually buy stickers, I actually got mine, but I think I left it at home. I’m going to put one on my laptop, just because I love that. So I think that’d be great. But I think that’s a great way. Another way is to start a book club at your organization, and I think books are really great for qualifying. So that’s a sales term, so When you ask someone to read a book and they come back and they say I loved it, here’s what you can get from that is that person probably spent six to eight hours reading the book. And they finished it, and they came back, and cared enough about it, they came back to talk to you. That’s a heck of an investment for that person to have made, right?
So I would say, give up your copy of the book to other people, give out other copies. And the ones that come back, and they have something to talk about is I think a very great signal of who the other rebellion members are. So, the book is really intended to generate a visceral response of saying, holy cow, this is us, right? Nothing makes me happy when I see people on Twitter or in Slack say it’s like they’ve been hiding like in Phoenix Projects or Unicorn projects it’s like he’s been hiding in our office for years, right? It says that, and they know it’s not good. And books create this kind of wonderful ability to talk about problems in an organization that aren’t you. So it becomes less personal, and I think it’s just a great way to start a rebellion. And that’s very much talked about in the Beyond the Phoenix Project book about like, how effective the Phoenix Project was at helping create rebellions in a slightly different context.
Time check for 20 minutes. Good, I think we’re on track, yes. Alex?
Hey, another thing. I’m not sure if this made it into the book or not. I was kind of kicking myself… I’d wake up in the middle of the night, worried about things I didn’t make into the book. Kurt, one of the reason why he spent so much time in the rebellion, dockside meetings, intelligence gathering, is he’s trying to answer that question is, who’s having problems? Who’s likely members of rebellion? Recruit them, and if it didn’t make it in the book, it should have. He actually meets with a lot of sales people from vendors, because he trades information, right? He’s using their… salespeople, their job is to understand organizations, so they can find prospects that are qualified that have a problem and have budget and can afford it to buy their product. So they have incredible… not always accurate, but they spend a lot of time intelligence gathering, and so does Kurt. And so in my head Kurt was always going to lunch with salespeople, and even though he didn’t have budget, he had that was valuable was information, right?
So you can imagine him playing a very good game of trading information for information in a way that serves each other and also help serve the organization. So it’s not that he’s like a double agent, he’s actually advancing everyone’s goals included for Parts Unlimited. Another piece of information that might be less obvious is salespeople find out who other champions are. And salespeople live or die by their ability to judge who can drive purchases, at least if they get paid on quota. All right. Next question. This is so fun, these questions are so good. I’m hoping you’re getting as much out of it as I am.
We just took a quick pulse a lot of positive reactions.
Okay, good. And I would love this transcribed and let’s post that, publicly posted that’d be great. By the way we use rev.com to phenomenal success they’re a great transcribing company and it can be very affordable. And mostly fast, when we needed them once it took them actually five hours, which is very atypical.
Question twenty-one, David, “What is your perspective on How the Product Drive Org model applies to Infrastructure Services. Basically convincing a CTO group this product stuff applies to them and they are critical.”
Basically convincing a CTO group, this products applies to them and they are critical? Oh, yeah. To me, it’s obvious that a lot of these internal platform, those are products. So if you look at the… the Disney presents… No, I can’t still not allowed to publish videos. Even better, the Adidas presentation from DevOps Enterprise Vegas. Right. Fernando Cornago of which Adidas is served as a model for “The Unicorn Project.” He’s a director. He’s responsible for all the internal platforms.
And he has some exciting news to share about some big responsibility he’s been asked to lead. But that’s a product team. In fact, in the previous year’s presentation that he did with his VP, I think that was in 2017. Essentially, they created this internal platform team at the CIO level, and I think they recruited a 60 or 300 people, they basically took 60 or 300, or like whatever the best number, they took the best engineers from the product areas, from the business units to form the centralized platform group out of self interest, right? Because they would all be better served by creating this kind of center of greatness and platforms that people could use without having to deal with opening tickets and so forth, to get what they need done. And so everything was self service. So that is a product team, that they compete in an internal marketplace. You don’t have to use it, but their goal was to be so good that they will win a Bake Off with any sort of external vendor.
And his teams keep growing because they’re that good at winning in the internal marketplace. So absolutely, platform teams are products and they get funded. I think they did watch the presentation, but I think they got a year or two of funding. And then they were going to move to a chargeback model or whatever, where basically you use it, you pay for it. Just making sure that they were held accountable to building a product that internal development teams loved. So I would say that would be a phenomenal basis for an example of a model that is clearly working. And they took over not only internal dev productivity, but all the internal data platforms as well. So post a link to… Alex, please link to Fernando’s presentation Vegas 2019.
And the one that he gave with his boss in London, year before. Thank you. 14 minutes, okay, good. 22. I think we’re doing great. Pacing is a… Wow, this is takes a lot of energy.
Question twenty-two, Richard, “As an agile based enterprise recently acquired by a monolity/waterfall based company, how can we prevent their practices from changing (regressing) ours. Or better, changing theirs?”
Great question, Richard. So one of the people I admire the most is Scott Prugh CSG. He’s spoken every year at the DevOps, he’s spoken more than any other person. In fact, we’ve started moving… Last year, we kind of moved to a rule where repeat speakers speak every other year. But I wanted Scott to speak so much he was the exception. So he’s now spoken every year of DevOps Enterprise. And his talk was one of the most amazing in the history conference. Okay, so why am I bringing it up? How did he end up a CSG, the nation’s largest bill printing company they’re publicly traded. It was born out of first data in 1977, it was spun out.
Anyway, his company was acquired by CSG. Then he became chief architect, then chief architect and VP of R&D for all of CSG and then he became SVP of all of R&D and operations. So it’s just his ascendancy that is, I think spectacular. But he came in through an acquisition, and so I give that to you as an example of a trajectory I would love to see for you and your team is they ended up taking over the mothership, to a marvelous effect. And I think it’s just an inspirational story. In fact, if you email me, there’s a talk that he gave at the scaled Agile Summit about his kind of personal history that describes how he entered CSG through an acquisition. That might be worth a conversation. We can see if that video is publicly available, and secondly, if you want to talk with him, and maybe get some tips on how to reproduce those same outcomes for you and your Unicorn-y start company acquired by the monolith-ly acquire, I think that would be a very worthy thing.
Question twenty-three, Ryan, “Speaking of dashboards, I completely agree that they are fantastic, but you can easily go overboard on the amount of dashboards. How do you balance the quantity of dashboards you are showing verses the quality to not have information overload?”
Yeah, I don’t know. But luckily, this is not my area of expertise, but this is the areas of expertise for a bunch of people. And this is actually addressed in that presentation that I had posted from Steve Pool at IBM. And there are a bunch of presentations that I’ve seen, but I don’t remember which ones they are about how to build good dashboards. I remember a bunch of them being by like folks from the dashboard companies like Geckoboard, which was acquired by New Relic. And so I would say, hey, vendors are great, right? They educate at least the really good ones do. So I would recommend going there. And if you tweet out that question, and mention me, I’ll retweet it. And just ask questions like, what are the best resources to create great dashboards? And I’ll retweet that and let’s see if we can find some resources and let us know and write about it. And we’ll cite that, that would be great.
Question twenty-four, anonymous, “How do you go from showing interest/getting more involved in the platform while not dropping the ball in your day-to-day feature work?”
Yeah, that’s a great question. Yeah, so this is like another one of those things about like, what do you do when your work systems are just kind of terrible that don’t recognize the value of paint on technical debt and like when you have to do improvement work on the side, which is a fact of life. Why did they have to meet after work hours? Well, only because it wasn’t a sanctioned activity, they could have gone to lunch, I suppose. But I did love the idea of them meeting after hours, and I did love the idea that Eric showed up. Not as a bartender, he owns the bar, right? So I guess I would go from least desirable to most desirable, right? And ideally, kind of march up this scale. One is you do it on the side, right?
And sometimes it means like, working, you’re doing it like literally overtime or you’re so good at implementing your features. You just serve high to excess capacity and work on your improvement work, you do that at the individual level. It’s a little one part you can say it’s dishonest, but if you’re doing it for the best long term interests of your organization, right It’s absolutely is the right thing to do. The next better thing is you do that at the team level. The development manager knows what’s best for the organization, they actually hide a pool of capacity, they create a 20% pool that is taken off the table. It’s not eligible for use by product owners, and the use for improvement days, hackathons and so forth. Then the ideal is, hey, the entire engineering organization understands the value of that. And that’s what actually Scott Prugh CSG moved to, and resulted in these incredible productivity gains. And in my mind, they create a world class engineering culture, and that’s the way it typically works. So, yeah. And eventually, it just becomes a cultural norm and you’re not taking any risk at all. It’s just the way the engineering organization works.
Twenty-five, Alex, this is the last question I have in my queue. So if there’s more can you put them in the DM. Twenty-five, and I don’t even know what the time is eight minutes. So we probably have time for two more or maybe one more. Okay.
Question twenty-five, Steve, “Do you have personal experience you can share when you decided it was better for you to move on rather than continue fighting the good fight?”
Yes. But let me just carefully edit this to not get anyone in trouble. Okay, not speaking in the… I’m laughing just because of all the things I’m going to edit out. If you want the unedited story, just reach out to me in DM and we’ll schedule some time to talk, and I can show you all the gory detail. But yeah, my decision to leave Tripwire I was there for 13 years. I was a technical founder of the CTO. I’m very proud of what we achieved there.
I learned so much, and I don’t think I could have written the Phoenix Project without my experiences there. I have nothing but the utmost gratitude for my Tripwire experiences. But yeah, I always say it became less fun after we… I left when there were about 350 people, we had just filed for our initial public offering, at least filed our S1, we never made it to the public markets. It was eventually bought by a private equity firm Thoma Bravo. But I had left by then. And I think it just got less fun after we crossed 180 people, and I think the reason is that’s when the senior director showed up. And kind of in my role, I chose not to build an organization and have a lot of line responsibility. So I would use influence as a way to get things done. And looking back that was probably not so smart because what it didn’t allow me to do is do things without a lot of buy in. And so when the senior director showed up, the surface area of influence that I had to exert was so much bigger, 10 times bigger.
And so kind of on a good year I was working on exactly what I wanted to work on, and I was getting resources from sales, marketing, engineering to work on what I thought was good for Tripwire and me. The worst years, it was like nobody knew what I was working on, and when they did, they were like why is Gene working on that? And they would take away funding, that’s not so fun. And so I eventually did leave to work full time on the Phoenix Project. I could not have done that without having been able to do some private share sales and create some runway. I actually work full-time on the Phoenix Project for three years from 2010 to 2013. So that’s the reason why I’m so grateful for Tripwire, the Phoenix Project would not have been possible without it. But yeah, when it just became too hard to do the things that I thought were important for the good of the organization. And in reflection I think I could have pulled it off with nine years more experience.
I mean, I can even imagine like how those conversations would have gone, I would have talked more about the highest level goals of the organization. And how I understand that we’re doing X, Y, and Z, and yet they’re not working. Right, the category size is too small, and we shouldn’t go into a bigger category of security, we should go into a larger category in operations and developments. I think there are a lot more tactics and techniques I could have used to be more convincing. But at that point, I think that would be on my abilities. And just the opportunity cost of staying was just too high, ad I think that was a calculation that I think was correct. So I think there’s two things I’m trying to convey. One is given my level of skill and experience at that time, I was just not effective at convincing who I need to convince in order to get things done that I felt need to get done. And then there’s an economic calculation of like what are the other things you could be doing, and that would be appreciated, would be valued, advance your own personal goals.
And I think that is absolutely an economic calculation of opportunity cost, and I think the decision I made was, the right one. Besides, given the fact I’ve never had as much fun as the work I’m doing now, I think folks will enjoy is not only the right economic one, but also the right one for what I want to be doing every day. Okay. Three minutes, one more question.
Twenty-six, Steve, “Along the same lines, do you have any advice to those feeling burnout that look to your books for inspiration and fuel?”
Yeah, just find the right allies recognize kind of… find the right allies, you want to find the rebellion that kind of has the complementary set of skills required to take on a very ancient powerful enemy. And then also be cognizant of what the skill levels are, if you’re a bunch of level four characters trying to take on a level 30 death mage maybe the outcomes aren’t going to be so good. But in combination, if you have the right quantity and the right skill levels maybe you can beat that level 30 mage.
And in the case of my Tripwire experience, I just didn’t have powerful enough allies, and I didn’t have the right skill level, so I think that informed my decision to leave.
The last question and two minutes,
Question twenty-seven, David Levine, “Is there a link to the Capital One and CSG experience reports from DevOps Enterprise?”
Absolutely. Alex, you will post those, and I’ll say this right now, with very few exceptions, I would say all the DevOps Enterprise plenary videos are worth watching. I mean, and certainly unqualified opinion, like without qualification, without exception every one in the last three years are worth watching is me. I think the quality of the talks are that good. And then Alex will send you to the YouTube channel and the playlist.
All right, Alex, thank you so much for the help and to everyone who made this Ask Me Anything possible thank you to you. And again, I appreciate all your enthusiasm, your support, helping get the word out, your curiosity, your teachings, I genuinely appreciate it and when I say that I’ve never had as much fun as I’m having during the work I’m doing now including today that is so true. And that wouldn’t be possible without all the things that you all do. So, my most profound sense of appreciation and gratitude to all of you, so with that I’m going to turn over to Alex because I have a sense of incredible tunnel vision, right now. I have no idea what’s going on. So Alex, can you help close this out?
Yeah. So thank you, everybody for joining today. I think we had about 108 of you live at the peak. That’s incredible. The questions were great, we have the same schedule for next week for the following AMA. So it’ll be same time.
We will post a new link to join for that. Afterwards, we’re going to get this video processed, and we will get a recording out to everybody. We’re also going to get it transcribed and we’ll get it just made available in any format that you like. So thank you all for joining.
And we will sign off.
And I’m actually light-headed from like not eating enough and like the energy consumed. I will actually eat breakfast before next time, so that I can survive the mental ordeal of going through, this is great. Thank you. I’ll see you in a week then. Take care all.