Can you outsource DevOps? And if you can, what are the keys to doing it successfully? In the fifth episode of DevOps Unbound streaming broadcast on TechStrong TV and DevOps.com’s sister site Digital Anarchist, Mitchell Ashley of ASG and Alan Shimel are joined by Ming Gong of UST Global, Chris Rolls of TTC and Robina Laughlin of Guardian Life for a discussion on outsourcing DevOps.
The video is immediately below, followed by the transcript of the conversation. Enjoy!
Transcript
Alan Shimel: Hey, everyone. I’m Alan Shimel, and welcome to another episode of DevOps Unbound. DevOps Unbound is a series, sponsored by our friends at Tricentis, that explores different topics around DevOps every other week, as well as having a monthly roundtable open to the public. Today’s episode is “Outsourcing DevOps: Oxymoron or Truth?” I’m joined, as usually, by my co-host, Mitchell Ashley, CEO, founder of Accelerated Strategies. Mitchell, welcome. Thanks for coming on.
Mitch Ashley: It’s always awesome to be here. I’m excited about our guest today.
Shimel: So am I. Let me introduce you to our guests. We have Robina. And, Robina, I’m gonna ask you to give your full name, title, company, background. And we’re gonna ask you to kick it off.
Robina Laughlin: Sure. Hi. I’m happy to be here. Thank you. My name’s Robina Laughlin, I work for Guardian Life insurance. Actually Guardian Life and I have been working in the quality assurance space now for about 25 years. I’ve been working for Guardian for eight years. We have a great enterprise quality assurance journey. Most recently my department, my quality assurance department’s been working with integration into a DevOps model. So we’re starting that journey. We started about a year ago. And really happy to be part of the conversation.
Shimel: Thank you, and thanks for joining us, Robina. Secondly, I’m gonna introduce my friend, Ming Gong. Ming, a little background?
Ming Gong: Hey, Alan. How are you? It’s always good to be here. My name is Ming Gong and I work for UST Global, I’m vice president. I run a practice called Digital Agility Platform Solutions. So what we do is we provide all kinds of services and even product to our enterprise customers, things such as DevOps transformation, DevSecOps, digital, cloud transformation, as well as latency modelization. So, we have tons of battle scars as we’re implementing DevOps for our customers.
Shimel: Absolutely. And then, last but certainly not least, Chris?
Chris Rolls: Hi, nice to meet everyone. So my name’s Chris Rolls. I’m from a company called TTC. I’m CEO of Americas. And what we do is we help organizations with quality engineering, continuous testing and also Agile and DevOps transformation. So really happy to be here and excited to hear what everyone else has to say.
Shimel: Absolutely. So, guys, when I started – so we launched DevOps.com in March 2014. Working on it all through the second half of 2013. At that time it was sorta dogma in DevOps that DevOps wasn’t about the tools; DevOps was about people; DevOps was about culture. And that therefore, to be successful in your DevOps journey, outsourcing was a no-no. You couldn’t be successful and outsource your DevOps because it would destroy – you never got to the nitty-gritty of the culture.Â
Now, that was seven, eight years ago. A lot has changed. I think we’ve seen some very successful – the world is full of DevOps consultants, large, small and in between, who are helping organizations along their DevOps journey. So I think we have kinda moved beyond the dogma. But have we moved to truth? Can you successfully outsource DevOps? And if you can, what are the keys to doing it successfully? And I’m gonna throw that out to our panel today, and we’re gonna start that as our discussion. Ming, not to pick on you, but you already mentioned you have a lotta scars to prove it.
Gong:Â Absolutely [laughs].
Shimel: Tell us a little bit about your journey or UST’s journey in this space.
Gong: Yeah. I think you mentioned a couple of things that’s really important, which is, ultimately, I think to do DevOps correctly the entire delivery team needs to live and breathe the DevOps – not only the automation, the tool set, but also the processes as well as how you do things or the whole mindset thing. So that’s the absolutely ultimate goal. I think even over the past few years, even though all the companies all moving towards DevOps, but they are still having a lotta challenges in terms how to take some isolated success in certain areas, trying to cascade that throughout the entire organization. And I think this is where, when we talk about, “Hey, is outsourcing a myth or a reality?” I would say it’s actually somewhere in between.Â
Because it really depends on where the continuum – when I think about DevOps, I think about DevOps continuum, ___ maturity, where they are. For the company who’s really mature, they are less likely to benefit from outsourcing before the folks who’s just starting the journey or middle of the journey – they definitely can use some help. And not only outsourcing in general but also what part of DevOps to outsource as well. So typically when I work with enterprise customers, they have a centralized organization, like some kind of DevOps COE. But they also have each individual team that needs to be enabled. So where we found there tends to be more success is when we help the COE to – infuse them with the DevOps expertise and help them kinda build up the practice around DevOps.
So I definitely think it’s somewhere in between.
Shimel: Fair. Chris, your organization is similar to Ming’s in that you are helping organizations with their transformation. Has your experience been similar, different?
Rolls: Yeah. I think I’m not gonna shock the world by being a consultancy who sees consultants can help with DevOps transformation, right? I would be outta business if I thought that was the case.
But if I was to answer your direct question, can you outsource DevOps? I would say no. Because DevOps is about a cultural change within the organization as well as the technical practices. Now, similar to what Ming is saying, it’s sort of in the middle because outsourcing and consultancies and other vendors can be a part of the solution. They can make some things more challenging in certain ways because you are introducing certain barriers, as in contractual sort of challenges around it. But it definitely can be part of it and it definitely can help, particularly when organizations are trying to figure out how to do things better.Â
Shimel: Fair. So, Robina, you look at this sort of 180 degrees different than Ming and Chris do, right? You’re not providing consulting to organizations. You are the organization. What’s your view on this?
Laughlin: Yeah, thanks for having me. This is a really interesting conversation. Some of my thoughts around this is: I think the important thing to understand is: what’s your driver to DevOps? Why are you moving to that model? Why is that model gonna be important to you as an organization or as an enterprise? And when you start to answer that question, you get to different models. For us, we adopted SAFe. We started that journey two or three years ago. So DevOps was not the first thing we tried to conquer out of that transformation. It’s on the tail end of it. And it was always on the strategic road map. However, what we found is that it grew organically based on the improvements from adoption of SAFe and agile. So for the teams to get to the velocity that they wanted to do, it became a necessity as opposed to an aspirational goal, right?
So for us when it became that driver to success, it took on a whole different model. And then when we started figuring out how do we wanna do this with our partners, the vendors that we work with, one of the things that we always hold true to in Guardian is that frameworks, guardrails, models our Guardian culture, and we developed those. We developed them in conjunction with our partners. We’ll seek for collaboration. We’ll seek for guidance. But the model has to come back to fit our culture and how our folks work.
And then once we have the guardrails in place and the operating model, inclusion of an outsourcing model is very successful, I find. But I speak from a QA perspective, and I’ve been on an outsourcing model now for 15 years, right? So I always kinda take the same approach to things in that if you have the operating, the guardrails, and people know what they need and why they’re doing it then outsourcing is a good solution. But you have to have a strong backbone for the company that you’re working for and with to be able to represent that model and the culture.Â
Shimel: Mitch – oh, Ming, go ahead? Whoever, Chris.
Rolls: Oh, sorry. I think just to drill into that as well, Robina mentioned you’ve gotta understand your drivers for DevOps; what are you actually trying to achieve? I think that a lot of times organizations – they want to do DevOps because that’s what the conferences and the experts say they should do. But also you need to understand: what’s your driver for outsourcing or engaging partners and other vendors, right? So if it’s a cost-saving mechanism, which often it is in this economic climate, is something that people are looking at, right? That can be one component. That can be one reason for outsourcing. You could also be outsourcing to get additional expertise that you don’t have internally. You could be outsourcing ’cause you’re scaling up with a whole another set of products and there are sort of HR reasons for wanting to do that. But you need to understand those drivers.Â
And I think another thing that Robina said crucially which I see organizations get wrong is: the organization who’s – the business, the incline for us needs to own the cultural change and needs to own the methodology. I come back to: no, you cannot outsource DevOps. You can’t buy that on a platter. You have to invest. You have to make the changes yourself as well.
Laughlin: Yeah. There’s some parallels to that conversation, right? ‘Cause a lot time ago this was the conversation for quality assurance. Like: outsource it; there’s an easy button; all your problems will be solved. I think that, from the experience of that, for me there was never an easy button. It was a lotta work and it was a big partnership. But what we did was, through that partnership, we started bringing in technology that my team hadn’t had a lot of exposure to so our options grew, and I found that there were some easy buttons along the way to build, to run tests against that build. So we’re moving into a model now where the assets of quality assurance can be augmenting what is your DevOps model. And that model has to work for your company and it has to be part of your culture. It is not something that just kinda develop out of a box.
Ashley: We probably need some precision around the term “outsourcing” too. So I think in this context –Â
Laughlin: That’s a good idea [laughs].
Ashley: – we’re not talking about externalizing it. In other words, put it in a box, ship it somewhere, have someone else do it; give it to us when it’s done. Right? Because we’re transforming the organization. I think it can happen a couple of ways. It can happen early in the journey; I think it also can happen – organizations typically get stuck at some point where we’ve gotten 2, 3 teams there but man-scaling it to a 5,000 or 10,000-person IT organization – that’s hard. It’s easy to do it on a startup basis.
So everything to me is always driven by what the company is about, what its DNA is like. Some companies are very data-drive: We don’t make a decision unless it’s got a graph attached to it or a financial analysis attached to it. Or tied to a strategy. Or whatever it is. And different companies often taken on the personality of the founders if it’s a younger company. But it gets established with it. And if you don’t have that figured out and aligned with how you’re implementing anything – agile, DevOps, SAFe, whatever it might be – you either go against the flow and fight it all the way and fail multiple times and never succeed or you figure out what fits into the flow and the momentum, the DNA of the organization.Â
That takes knowing people. It takes knowing the organization. It takes commitment by people in the company to do it. So it’s not an outsourced give away the problem. It’s a: I need help because I need expertise, because I’m stuck, because there’s other – we don’t know what we don’t know. And we’re looking to get smart from other people. And we want you to leave someday and help us operate on our own. Maybe help us with a different problem someday. But that’s kind of our goal as an organization.
Gong: Yeah. Mitch, I fully agree with that. And we certainly have – a couple of points you point out is pretty interesting. One is the challenge that we seeing for large enterprise customers: they usually have problem with cascading kinda DevOps excellence across the organization, right? But there’re so many hurdles, I would say. One of them is organization siloes. We have seen so many times where you have one division, usually on the digital side, front end – they have great success because technology stacks tend to be more modern so it’s easier to get on the DevOps train. But then once you start trying to cascade to a legacy side, it just doesn’t work. Either because they don’t share very well because there’s organization siloes, or because the technology stack they develop doesn’t really work on the other side.
So where we see, again – I know you have term about outsourcing can mean many different things. But where we see a lot of support or at least success is when we help kind of the organization where they have a COE, where they establish standards – and I think Robina kind of mentioned this too. It’s very important for company to really establish their true North, the compass, right? The guardrails and things like that. But it’s also important to not have a very heavy central organization either to trying to drive DevOps. Because then you’re missing out on all the distributed innovation happening on the ground level so you can never respond fast enough.Â
So this is why when we created our model, we usually have a team, or a very small team, help with the COE side of it to help them kinda develop their whole community, the culture, things like that. But then we have very nimble – we call it deep pod. It’s like a five-member crew. They have – consists of DevOps architect and couple engineers. And what they do is they will be kinda parachuted into an area within the enterprise where they need a lotta help or the application that they using is very critical to the business. Will drop in there. Not only helping with engineering, figure out the tech stacks and figure out the DevOps – all the nuances of it. But we also do a lot of coaching to make sure we can train the team to be self-reliant. Once they are, we lift the team up then we drop the deep pod into the next area where they need help. So this is kind of the hybrid model is where we find more success.Â
Ashley: I’m really curious: Who typically brings in your firm, Ming, your firm, Chris? Maybe, Robina, who brought in someone to bring in help with DevOps. Was that at a very high level? Is it a development team level? I’m thinking larger organizations here. What’s the –
[Crosstalk]Â
Gong: I think Chris and I probably will share the same answer, which is: all of the above. We have a large DevOps engagement to where the CIO invites in. And we operate starting from the strategy level all the way down to the individual team level. But then we also have, through relationships or whatever case might be, where we insert ourselves into a particular team to help and then grow from there.Â
Rolls: Yeah. Exactly the same. And if it’s okay I wanna go back to one thing you guys said before, which is having a bit of precision around what we mean by outsourcing. Because I think when I see this topic, I jump to full, functional outsourcing of QA or dev or opportunities. But the reality is: outsourcing also exists on a spectrum.Â
You know, when you hire a maid to clean your house or to service your car or anything like that, you are outsourcing that function of your personal life. And it’s the same with software development, right? So there’s the really far down this end where it’s just you have a consultant who’s embedded in a project team, and then there’s all the way down to the far end where you do that full-scale outsourcing of a function or a department.Â
And what I want to be clear on is: the further you get along that continuum, the harder it is. You have to be far more careful about it. And so that’s something that’s really important. And to go back to the question that was just asked: completely the same. Sometimes – and you’re very happy when the CIO or the CEO says, “Come in and help me do this transformation across the organization.” Sometimes it is a QA or dev manager has a specific problem that they want to solve. But I’d also be really interested to hear Robina’s thoughts.
Laughlin: So how did we bring – I’m sorry. Can we –?
Ashley: What level did you bring? Was it the CIO brought in somebody to help you with the ____ or was it on a QA development team?
[Crosstalk]Â
Laughlin: Oh, sure.
Ashley: How did you start down the getting external help, outsourcing help?
Laughlin: Yeah. So this is a little bit of a parallel run to a DevOps model. So from a QA perspective, what we did was we decided to centralize our enterprise quality assurance organization. And when we did that, certainly we got economies of scale, but I needed a partner or a vendor that would come and work with me. And when I say partner, I truly mean it. Like we have team meetings. They’re all folks that work for me. I don’t distinguish. So you have to be able to have that relationship I think to be able to get success out of outsourcing, if you wanna call it that. Because it’s really partner shipping, in my mind.
So that was the journey that we took. And that’s been underway now for almost eight years. So we have an incredible wealth of assets behind us. And I started stopping and looking at what my partners were doing on the other side of building and deploying, and how are they verifying that building? We’ve been moving this shift-left journey of discovery of quality issues for a long time now, and we’re at the point where we’re in the build stage. And I’m like, “Well, we have great assets that could help verify that a build’s gonna work. What’s the next solution to that?”
So I reached out to my Cognizant folks under my QA umbrella and said, “Hey, if we were to think about extending our assets use, what would we do?” And they were like, “Turn to your DevOps team. Turn to your build and deployment team. Let’s understand what they’re using. They’re using Jenkins. They’re using Bitbucket, example for us. So when we started looking at being able to scale our assets and use them across the enterprise, we built them in Jenkins. We built our actual Selenium code in Bitbucket. So now we have the hooks there and we can – as these teams start to catch up or we start to catch up and we meet in the middle, we then have a problem we’re trying to solve. We are getting builds. They are broken, or when they pass, there’s quality issues. Can we discover them through automation? So that’s the easy button.
And in this case I actually do have the easy button. I have this regression suite. I can slim it down to a smoke test or something, and say, “If we can deploy this and do all of this then, yes, we can get those results.” And then the next phase of it is accountability of teams of what they’re gonna do with those results.
So none of that is technical. That’s all cultural. Except for the hooks in what we built. So that’s kinda been our journey and that’s where I’m looking to go strategically.
Shimel: Excellent. So, guys, I want to bring this down or – we touched on it already but I wanna clarify it. You know, Ming and Chris, you’re in a unique position in that you guys get to see dozens of organizations and how they approach this problem. Where, Robina, your view is Guardian-based, right? This is what you guys have done at Guardian.
Here at DevOps.com I’m also – I’m in a catbird seat in that I get to speak to many people like Ming and Chris, but I also get to speak to many people like Robina. I get to speak to people like Mitchell. I literally get hundreds of views on this. And I recognize patterns. And what do I do? What’s my superpower, superhero power? I’m good at recognizing patterns. I guess I’m a machine learner. But anyway, there usually is a pattern along DevOps adoption. And it’s similar to what Robina mentioned, which is – when I first launched DevOps.com, our tagline was “DevOps picks up where agile leaves off.” Right?
And, yeah, because what happened was people were moving the agile development processes probably going on 20 years now. Agile’s been around 20 years. And it was great. It makes you feel old. I’m sorry, Robina. But, yeah, it’s 20 years [laughs]. Don’t feel bad. I watched this “Halt and Catch Fire” about the birth of the PC industry. Yes, that’s what it is. Einstein’s theory of relativity [laughs].Â
Laughlin: I remember my agile training with Ken. I can’t imagine 20 years –
Shimel: Yeah. Ken, who now of course is Scrum.org, right? Versus Scrum – the other one. Not here to get into that one. That’s another rat’s nest.
Gong: Different podcast [laughs].Â
Shimel: Yeah. But the thing is: the normal course, if you will, is: yeah, we do some agile. Whether it’s Scrum or SAFe or whatever we’re seeing out there, scale. We wanna start getting into some lean IT stuff. And of course when you start marrying agile and lean, it’s DevOps, right? That’s kinda what DevOps is. And now we’re gonna get into CI/CD pipelines. And of course it’s grown since then. We’ve added SRE on. Continuous testing has become huge. Right? It’s a huge piece of this. So we’ve seen the sprawl, the urban sprawl of DevOps and agile over the last 20 years. Along with it, I think it’s become all but impossible to do at a large organizational scale, at a large enterprise scale, without some outside help.Â
I think we’ve recognized the enemy. It’s us. We’ve created the need in our large organizations, the way we go about adopting these things, that we can’t do it without outside help.Â
Laughlin: That’s very true, Alan. I think that’s why you really need the outside help and you need to go to the experts. Because no one size fits all. No one tool solves everything. That’s how I got involved with Tricentis: I went out really early looking for a vendor that would grow with me and was on the same aspirational journey that I was, right? So that’s where I think it really helps in terms of consulting. You need to pick the partner that’s gonna work with you and grow with you. And sees you for your uniqueness, not an industry standard.
Gong: That’s right.
Shimel: Chris, Ming?Â
Rolls: I think you can’t take a single ___ of technical practices or cultural adoption and apply it to every single organization. So that point, Robina, about making sure that you tailor it for the organizations that you’re working with – or even within an organization, the different departments, the different products that you’re working with, right? There’s a different sort of cultural mode within the digital section versus the enterprise technology section, right? So that’s really important.Â
And I think that also comes to a thing that I see that often people get wrong when they do outsourcing is that they don’t align the incentives properly. Because ____, we all have different incentives. Even, Robina, if I was partnering with you and we want to work as really good partners, then I have a slightly different set of incentives to grow my company as Guardian does, right? And so we have to find ways to align those incentives. Otherwise you will create a cultural conflict which prevents you from really going as fast as you can. So that’s just something that I see often people get wrong about moving to DevOps, because they have to use outsourcing in particular, engaging firms to help them with that.
Gong: I wanna go back to the point Alan was talking about in terms of the massive talent shortage in industry. And it’s not just DevOps. It’s cloud, everything else, right? Just in general I think we all wish there’s more qualified folks who has those technical skills. And I think that’s one of the reason why a lot of our customers want some help from us. Because they simply cannot find enough talent out there that not only knows the DevOps technology but also has the ability to keep up with a crazy changing industry, right? I mean, you see a massive change even within the last couple years where used to be all the siloed best-in-breed product like Jenkins or whatever else, and then artifactory, whatever.Â
But now you see a trend where industry’s moving towards more of a kinda end to end all in one type of tools, right? For example, you have the Azure DevOps, GitLab, and the whole – they start offering the whole suite of things. Because they recognize the massive talent short. You can’t take care of the problem until you can also solve the problem on technology side as well. So that’s kind of approach that we’re taking as well, where not only we have on people side we have our deep pods kind of coaching the team so we can get lots and lots of teams to train up to be self-reliant. But on the other side we also develop on platform to kind of try to extract all the complexity of DevOps from the user so the user doesn’t have to be an expert. So we call it democratizing DevOps for everyone, right?
So I think for a company to be successful they need to kinda take approach of kinda both people and the tools combination.Â
[Crosstalk]Â
Ashley: ___ kinda just share a personal story in my additional adoption of DevOps when I was a CIO at a company. And it was initiated by me ’cause actually Alan had reached out to me and said, “Hey, Mitchell, there’s this thing called DevOps. You gotta go read about.” I’m like, “Okay, what is it? I’ll go figure it out.” I got hooked onto it.
But the rest of the team hadn’t really caught fire yet of getting onto DevOps. And so as a leader of an organization, it’s tough to say, “Okay, I want us to go this direction; I think this would really be helpful for us” when the team’s not there yet. And really understanding what it’s gonna take to get the organization going and get behind it and how to be successful at it. This was earlier, so you reach out more to individuals that’ve started to build some brand and some experience in doing things like DevOps.
Today I would go to a different kind of a team to say, “Here’s where we are. Here’s the projects that we have going on. Here’s kinda what we’re being asked to do,” right? I think every – maybe not true, but it seems like most IT organizations are always being asked to transform some way, somehow. And sometimes pretty radically. And if you can sort of set where we’re headed and tie it to the why, as you talked about, Robina, the why for the business though, not for IT, that seemed to me to be the biggest thing that everybody could rally around. Not that it was DevOps or agile or it was this. It was: “We wanna go after this market. And our ability to get there is really constrained today. And let’s figure out some ways we can go after that.” And that tended to rally around.
So I think today – I guess my point is: Today we’re at a point where we can hire more expertise. And we wanna de-risk our journey and learn from others. We don’t have to figure it all out ourselves and be the smartest person in the room. Certainly I don’t. And that’s a great environment to be able to bring in help.
Shimel: Robina, I’d like to throw out – there was a time when we used to look at quote/unquote “outsourcing” and consultants, and one of the terms that was thrown around was “body shop,” right? Like, “Bring in some bodies. I’m gonna rent some bodies.” I remember when Mitchell and I were at StillSecure. We’d go down to – we had a big federal practice, and we’d go to a federal facility, whether it’s a lab or a knock or what have you. Eighty percent of the people there were not federal employees. They worked for Grumman Northrop or SAIC – take your pick of the beltway bandits. But it was all body shop. And when you’d speak to the salespeople at these consultants, you’d say, “How’s business?” “I got 50 people placed.” Not what the mission was, what the tools were. They judged their success by how many people they had placed.
I like to think – and maybe I’m naïve; Robina, I’m gonna ask you – that we don’t judge the success of our DevOps outsourcing by how many bodies I’ve got buried at a given location. ‘Cause I think that’s a terrible metric to determine the success of a transformation.
Laughlin: Yeah. That’s a great observation, Alan. You just kinda gave me shivers when you talked about number of bodies. I remember those days. They’re painful to look back on, honestly. There was a time when you needed an army of folks to do things and I don’t necessarily think that’s the case now. For me, our journey was: I needed a partner and I needed someone that was gonna come here and work with me. And through that partnership I also needed to know that I was gonna have a 24-7 shop, and that’s gonna necessitate overlap on both sides, right? Your onshore and your offshore. So I don’t think you can ever just kinda set it, forget it, and throw it away to an offshore team. That’s just not how it works.
And I think the other piece of it is: at the very source of it, you have to remember that these are human beings. They have families. They have lives. They wanna do good work just as much as we wanna do good work. And the way that they’re gonna feel that contribution is being part of the team. And the way they’re gonna feel that sense of that is by being present. And we actually flipped the dynamics. So we’re no longer a headcount shop. We’re an outcome shop. So I measure the success of my organization by the outcome of our testing. And for me, with my partner, if they do it with 2 people or 200, that’s their discretion. So I don’t have to get into any more of these conversations of: “We need two more bodies here. We need eight more bodies there. I’m anticipating 12 bodies.” That just turns into I think an empty use of really valuable strategic time that you have with your partner. And you should be using it a little bit more productively, right?Â
So that’s how I flipped that dynamic. And I’d be very curious to see how Mitch and Ming feel about it. And Chris.
Ashley: Well, my first reaction was I thought you were gonna talk about “Sopranos” or something. A whole different kind of body [laughing].
Shimel: Bada bing.
[Laughter]Â
Ashley: You know, I really like how Robina described it. Because it isn’t about budget size and all of those kind of backwards metrics. And I’m sure they’re still used today. To me that’s a measure of how aligned you are with the business. If you’ve shed the financial size, the IT-only-centric metrics to “What are we making an impact?” – there’re a lot of ways we can do a digital transformation project. We could get the contact list services let’s say. Something very relevant today. And it doesn’t have to be with hiring 50 more people ’cause we’re probably never gonna find 50 people to hire anyway. So we’ve gotta find other ways to do it.Â
But there’s also – there are times where I need these people to fill these roles. And I’m sure the body shop approach still applies in some situations. I doubt it on strategic projects though. That would be my supposition.
Shimel: Chris and Ming, I’d like to hear from you guys on it though. Because you’re on the other side of the equation here. Chris?
Rolls: I hate the term “body shopping.” It’s a terrible, terrible term. And I talk to the team and I say, “We have to have stamped on our forehead, ‘We are not a body shop.'” Because if you’re going in there and your goal as a account manager is to get 50 people on an engagement, you’re not thinking about what that organization needs. You’re not thinking about what the business outcomes are. You’re just thinking, “I can make $10.00, $15.00,” whatever how many dollars it is off each person, “and what’s my commission for that?” That’s a terrible way to think about it. And what’s more, it shows an outdated way of thinking about the people who are going to do this work. People, process, technology – people, in my mind, being the most important part.Â
If you’re thinking of them and you’re calling them “resources,” which sometimes we all slip into, or you’re calling them “bodies” – that’s not just a warm body in a seat typing some code or testing some software or monitoring the systems. They’re people who have skills who are thinking about things who are trying to do the best work they can to deliver outcomes for the company. And so I hate that term “body shop.” I hope that no one associates that with me and my company. And I think it’s a terrible way to think about delivering software.
Gong: Yeah. I definitely second that. I mean, what Robina just articulated pretty much echoes what pretty much all of our enterprise customers have said. They should absolutely transition from where it’s just TNM type of model before now to it’s all about outcome-based. And that’s what we prefer as well. To Robina’s point, why the customer should care how we deliver it as long the outcome is what they want, right?Â
We actually created a digital agility transformation manifesto [laughs] in honor of agile manifesto. And one of the top one is actually called “Value over volume,” which is exactly what, Alan, you just talked about in terms of, hey, you should not be measuring success with transformation by the volume of things got done. But it’s more about the value, right? So in this case, instead of measuring, for example, how many DevOps ____ did we create – that’s not important. What’s important is: how much faster are we able to help the team to deliver software to the production faster? What’s the frequency increase? So those things are what matters. So that’s what we absolutely focus on.
Shimel: Excellent. Guys, we’ve got maybe seven or eight minutes and there’s one area we haven’t hit on yet that we’d be kinda negligent if we didn’t hit on it, and that is: what effect has COVID had on all of this? On one level, we could say, “Well, look, everyone’s looking remotely anyway. What difference does it make if they’re an employee or an outsourcer? We’re all remote. We’re all equal in that regard.” The other way is: “How do I graft these people into our culture? How do we make them part of the team when we’ve never shook hands in person, when the only way I ever see you is on a “Brady Bunch“ Zoom window here and you take the top left?” Right? “Can we really integrate in just remote-only settings?”
Ming, Chris, you guys deal with it. You’re dealing with it over and over again. Robina, you are too. Right? Everyone’s remote. How has COVID helped, hurt, hindered, accelerated – how’s it affected this?
Rolls: So, I think one thing is that if we asked people five years ago, “How do you do agile? How do you do DevOps?” you’d say, “Everyone has to be in the same room. They have to be collocated. You’ve gotta have stickies up on the wall.”Â
Shimel: Chris is kinda sticking on us here. Okay. Go ahead, Chris, I’m sorry. You kinda hung on us. We got stickies on the wall.
Rolls: Yeah. So I was saying in the past people would say, “You have to have everyone in the same room, stickies on the wall, and that’s the only way you can do agile and DevOps.” And I think one thing that has been good with COVID is that it has dissuaded us of that illusion that that is the only way to do things because we have been forced to do it differently. That being said, it is so much harder to create culture, whether it’s in-house or whether it’s outsourced, when you don’t have that face-to-face collaboration. There’s only so many times people can do a virtual happy hour and still really get into it. So you have to be – I’m sure first time everyone’s like, “We’re having five happy hours a week. I’ve got my bottle right next to me.” And you’re two months in and you’re saying, “Geez, I cannot do this anymore. I just wanna go to bed or watch “Sopranos,” right?Â
But I think that we really need to focus on building that culture and come up with creative ways to create connections between people.Â
Shimel: Agreed. Ming?Â
Gong: Well, let me see. Hold on. Give me a sec.Â
Shimel: Okay [laughs]. You going to your happy hour background?
Gong: Absolutely.
[Laughter]Â
What you’re seeing here is the – my family and I, we took a hike recently to a Malibu beach, and I took a 30-second video of Malibu and use that as my background. So this is one of the things that I started asking my team to do: trying to personalize this kinda virtual experience. I’m a absolutely people person. So I miss the in-person bonding dearly. But in lieu of that, we basically, every time we have team meetings with our customers or with our own team, we pretty much strongly encourage everyone to get on the video so at least we can see each other. But to Chris’ point, there’s always a silver lining to everything. Because the COVID certainly forced or speed up a lotta things that was gonna happen anyways. So, to that, I think it’s a positive thing. Robina? I’d love to hear your thoughts [laughs].Â
Laughlin: Sure. What a year we’re all having, huh? Surprised. So, I think in terms of working with your partners through COVID – and we’ve been talking a lot about culture – I think one of the foundation things is to pick a company or a partner that has a very similar culture to you. So then you’re not trying to find middle ground in that. And that should really be a big part of your selection process. You need to be comfortable with the folks that you work with and you need to be living kinda to the same standards I think.Â
So for me, how COVID affected me was: while we were all scrambling to get ourselves home and reaping the benefits of a long-term plan to be able to have a remote workforce – we learned that through Sandy a long time ago and we had done a lotta changes – I was also working really closely with Cognizant in helping to get all the resources that they needed to get my folks home at offshore, right? So I was as concerned about offshore as I was concerned about onshore.Â
And now to this day my job is actually pretty high-touch now. And the reason being that we’re so far along, I have people that’re really overworking themselves. And I think that – and I don’t know that it’s deadlines as it is: take pride in craftsmanship, and they’re home and they wanna – this is something for them to do. Or it’s a stress release in a really challenging time where everything’s unpredictable. But I spend more of my time saying, “When was the last time you took a day off? When was the last time you took an afternoon off? You need to recharge. You need to gather yourself.” And that’s for offshore too. So that’s been my challenge this year.
Shimel: Fair. Mitchell, I know you guys at Accelerated Strategies Group did a study on the whole remote thing and COVID. Thoughts?Â
Ashley: Boy, I could talk for an hour on this. But I’ll just point out a couple things and I’ll point folks to the report. We looked at – software developers already know how to work remotely, right? A lot of them do. And so how has that affected them? How has COVID and the work from home affected them and how have they kind of affected the organization?
So, interestingly enough, across the board – there’s about 350 participants. Across the board, everyone is above 60 percent of who have accelerated digital transformation projects, contact – new forms the delivery and automation. But what I really found interesting was how influential DevOps practices, whether you called it DevOps or not, had started to infiltrate into the organization. So I’m talking about things like doing daily standup meetings or more cross-functional teams, increased use of that, asynchronous communications like a Slack or a team as opposed to being on a phone call or in a meeting. So there’s all these ways of working that, as we were talking before, sort of forced everybody into it. And the teams that had really good software teams working remotely jumped on board. And you saw an increase in the use of that kind of style of work, sometimes facilitated by tools, sometimes just process.Â
So if you wanna read the whole report – and reading is minimal; it’s all infographics; it’s super easy to consume – go to ACCELST.com. And it’s free to download. And enjoy. And let me know if you have questions. Love to talk to you about it.
Gong: Awesome.Â
Shimel: Excellent. Guys, we are outta time. We’re actually over time. Robina, Ming, Chris, I wanna thank all of you for joining with us today. This was a great discussion. In the show notes, production notes, we will have links to your websites and any other links you wanna put in there. But all three of you are dealing in the real world with real problems that we’re seeing today and helping organizations solve ’em, whether it’s your own organization or other organizations. So keep up the great work. We may revisit this topic on DevOps Unbound someday, and we’d love to have you back on here. Okay?
Gong: Pleasure.
Shimel: Thank you so much.Â
[Crosstalk]Â
Laughlin: Nice meeting you guys. Good luck. Thank you.
Shimel: All right. Mitch, thanks as always, man. This is Alan Shimel for DevOps Unbound. Thanks for joining us. We’ll see you on another DevOps Unbound real soon.
