Using New Tech Foundries for large digital transformation, with Regis Allegre

Episode 139 · 18:21 · SRT · archive.org · all Tanzu Talk

Hey there, this is another interview from Pivotal Paris. And just to remind you, if you're interested in these kind of topics and things like Spring and DevOps and kind of leading change in an organization, digital transformation, if you will, you should come to our annual conference Spring One Platform. It'll be in Austin, my hometown, October 7th and 10th. That's a great time of year to come to Austin, not too hot. And if you register before August 20th, you'll get $200 off. So go over to springoneplatform.io. Chances are, when you hear this, there'll be a whole bunch of sessions that have been accepted and listed. So you can get a sense of what it looks like. And with that, enjoy the interview. If someone like myself or someone else was coming to Paris, like what would, what would you recommend they do? That's just like, you know, not necessarily a touristy thing, but they just like, I don't know, I got, I got a couple hours to kill. What's something relaxing and nice to do? There's some very nice parks. So yeah, go, go to the park next to the Louvre, actually. That, that's pretty cool. Otherwise, a Jardin du Luxembourg is pretty nice as well. Or if you want a bit more hipster type thing, go to the Canal Saint-Martin site. So Canal Saint-Martin, that will be the super hipster place. And otherwise you have the riverbanks of the sand that are pretty cool as well with lots of open terrace music and so on. Yeah. It's kind of an outdoor sort of thing. Exactly. Yeah, exactly. Yeah, that makes sense. Well, why don't you introduce yourself? So I'm Regis Aleg. I work for Accenture while I'm a director in the Intelligent Engineering Services group, which is basically the group we've built around the new type of custom development. So it covers custom, IGL, DevOps, and everything that is cloud native as well. Right, right. And we were talking earlier in the day about how, well, there's two things. One, this notion of people call it various things like a lab or, but the organization they set up that's, to be overly vague, the new way of doing software, of doing innovation. And you actually have been working on that kind of thing

for a while. Absolutely. So what do you call that kind of thing? What's the word, the phrase you like? So the name we call it in Accenture Technologies is New Tech Foundries. And the idea behind that is it's because it can deliver many, many types of actual technology artifacts. You can deliver applications with them, but sometimes clients will want them to build, visualize this in use case on top of a data lake. So it's, so the tape factory comes sometime as well. We like to come. Because it's not just a stamp. It actually makes things. Yes, it makes things. And so, so what is that? Like, what is a foundry? What does it look like? So it's the, it's the tool that you can use to help on the digital transformation of a group to actually show an example on how you build things in a digital native way. Ah, right, right. So very often our clients have spent lots of energy and time trying to transform the organization from within. So deploying agile transformation initiatives or development transformation initiatives or building cloud capabilities. But they don't reach this dream of a continuous delivery ID to a real product with actual users in less than three months, et cetera. And sometimes the best way of doing that is actually to create a new organization on the side and build all of the best of breed, both from a technology perspective, but also from an organizational and culture perspective. Yeah. I mean, that's definitely what I see at most of the successful large organizations I talk with is they just, I always think of it as not always, but I think of it as the problem with successful organizations is they're immutable. Like you can't really change them. And so if you do want to change them, it's, it's a lot harder to actually change the organization than it is to just like start a new one. Absolutely. And then I tend to see that over and over again. So I think like in your experience, like what, I mean, it was difficult for the organization to change in the first place. Yeah. And so like, what have you learned about how you get them to

actually change? I mean, you know, you can set up a new organization, but you want to make sure you don't have all the same cultures and processes in place. So actually, so one of the things actually, you want the culture in that new organization to be new. So one of the things that you cannot import everybody from your existing organization into that, because then people will just get back to their whole habits. So actually you need very significant diversity when you incubate the thing. Yeah. And that's why actually they're quite cool stuff to build with a partner because you have at certain moments, lots of new people going into that. And you have a group of people that don't know each other that are very interested in building something new. Oh, right, right. So they have no expectations about what they should be doing. Or I should say how they should be doing it, right? Like they wouldn't say like, Oh, we've always done things this way. And like, you know, this is the only good way to do things. So they're more, uh, I was going to say malleable. They're more open to new ideas. Yeah. And especially because even if there's a template into how such organization should work. So like we have a pretty clear, uh, let's say coverage of the functions that should happen within one of those foundry. So typically there's, there will be something where you identify the good products for that type of development. You will need to scope them. So what happens when the guy with the business ID meets up with the delivery team, they decide in two weeks what they should be doing from a technical perspective, roadmap perspective, functional perspective, et cetera. Then there's a delivery phase where they'll iterate over five or eight, two weeks prints to actually build the product. And then there's a transfer phase, which is how do you actually give that product back into the existing organization? And that's where the catalyst aspect of the foundry works. When you teach the existing organization to operate in the digital mode. So the functions are those. However, how they're actually implemented will very much depend on the existing context and type of product that you need to build. Yeah. So tell me more about that last phase, because I think since doing things, uh, I'm being kind of vague, but since doing things in this way is relatively new,

there's not all that many multi-year cases of, and here's how we, you know, here's how we incubate this team basically or this product and then move it back, kind of like spread it back. And so what, how do you make sure people don't revert back to the old way of doing things? So actually you cannot really, so none of it. So the thing is that at some point as so the way it happens is that at the end of the product, during the, during the product delivery phase, actually, we've helped that organization recruit the team that will inherit the product so that they have the right skills. They have the same organization principles. They are agile minded and so on. So there will be some compatibility. Then we take them with us for a while. So we do one sprint together. Okay. The last sprint of the product where they learn about our continuous integration pipelines and so on. Uh, then they, uh, we do the scoping of the next phase together as well. So now we have the two teams defining what the future of the product will be. And then after that, we go to their location and operate in that mode with them for an entire sprint. So you sort of stick with them and kind of like give some coaching and input. And to a certain extent, um, what would you call it? Kind of guard railing. Yeah. Right. Like to sort of remind them that like, this is the habit you need to form. Absolutely. Yes. And also at that moment, we've agreed on what are the value KPIs that we need to measure to keep track of the successful products. And also we've put in place a reporting mechanism so that we can validate that they keep on working in the right way. Still, uh, we believe ourselves in the autonomy of teams. Okay. Right. And so we can only be prescriptive on how they should work up to a certain point as well. Right. Right. So they need to pave their own way at some point. Yeah. I mean, I guess, I guess it's kind of a cliche way of putting it, but you can instill the principles that they should be operating at. But at a certain point, you don't want to describe what every single person does on a daily basis, right? You just more want to say like, this is what it

should look like and figure out how to get there. Absolutely. Absolutely. Right. So things that remain important is that they should have a fast iteration and should not kill the continuous delivery and so on. And we have ways to monitor that. But if at some point it's more suitable for them to move to one week sprint instead of two, that's their, that's their product. That's their team. They've determined that that's the valuable thing to do. So, so, uh, conceptually, like with a couple of teams, like it seems to make sense, but you know, I mean the, the, the types of organizations, both of our companies work with are huge. Yes. And so there's a certain challenge of how do you, how do you approach like, you know, uh, a place that has like 20,000 people in the IT department or like, you know, 500 applications that eventually need to move over. Like even starting up with thinking about what you do, I imagine must be kind of different. So how do you all wrap your hands around that? So one, one, one of the things is that we really see actually the tech fundraise themselves as a minimum viable product itself. So actually you need to validate its success when things are still small and still in control. So actually we'd be more in, um, um, at my client, we started with only three squads and ensure that those were successful before when paying up. Right. So that's, so that you have something actually to base yourself, uh, to base yourself upon. So that's how you validate. Then after how you, um, you know, the, um, uh, elephant, uh, the feeling or the eating? No, no, no. Yeah. How do you eat an elephant? Right. One slice at a time. So it's something very similar happens. So, so the idea we, we followed at our, at our, at my main client was really that, um, every time you transfer one product back to his business line, you create a small bubble of digital mating way of working there. Right. Then with the second product, there'll be another one. And progressively there'll be three, four, they'll start talking with each other. And that's how you convert. Yeah. Yeah. Yeah. That's one of the, when that's, it's a catalyst. It's that's, that's really the way for sure. Yeah. And

that, that pacing is something I've learned over the past three or four years is, and I, I kind of have forgotten to remind people of this is like, if you're at a large organization, you should set your expectations. Cause this is going to take a long time, right? Like you can't just, it's not, it's not like some things where you just install some software or you have a new type of dashboard, right? Like it's more of, it's almost as if your chances of being successful will be very small if you don't start small, right? Like you say, start with one and then two and then so forth and so on. And so, uh, yeah, I mean, especially at large organizations, I think that takes a lot of persuasion because everyone in a large organization, if there's any new budget being spent, they're like, I need to have immediate impact in ROI. Yeah. So let's approach the biggest thing possible. And like, but actually the, the, the budget story is, is a key one. Oh yeah. How does that play? Yeah. And it, it comes to the type of product that you will select for your transformation where you actually need the mix, meaning you will need some, uh, really cool and shiny thing, pure player type thing. Okay. Where you will try to have a nearly B2C play that will grab customer and show the perfect example of a startup like initiative. Right. The thing is that you won't justify a multimillion dollar investment with those type of thing, which if you're lucky, they'll generate 500K revenues a year, you know? Right, right, right. So you need to have a part of that to set the clear examples. You also need to select a few of existing products that are, that drive much larger revenues that you will either re-platform or convert a ways of working so that you can actually tag significant revenues to the, the thing the factory is doing. Right, right, right. So that you have some scale that actually matches investment. So you will need the mix. You need the pure players to actually show the best example and go as far as you can into pure greenfield environment. But you also need to have some more like kind of technology conversion type place where you will actually re-platform some larger revenues

applications to actually bring into the new color of money some large initiatives. Right, right. Yeah, yeah. And so what, what do you, what do you call that process like of gardening the portfolio? Demand, so we call it demand management actually. Yeah, yeah, yeah. Because that's how you manage your pipeline of work that enters the factory, the foundry. Oh yeah, yeah, yeah, yeah. So demand management and, and the, the way we've named the process is what we call framing. Uh-huh. Where we actually analyze, analyze the business ideas and validate that it can be a proper minimal viable product to put into the factory. Oh, right, right. So there's validation to show you can do validation, which, which makes sense. And, and so like, what's the ongoing, um, when you do the, uh, the, the, the demand management, like, what does that look like? Is that redone every like quarter or every year? Like what's the ongoing process of it? It's a continuous process where you need to have scouts in a various business, uh, uh, domains that will spot ideas and promote them. And then after that becomes your pipeline. So actually it's your pipeline of work entering the foundry. Right, right. So, so that, that's a whole other role that you have is to basically be generating demand. Yeah. Actually when, when you look at the, uh, and that's actually for us at Accenture, the, uh, these new tech foundries, they're really a multi-brand play, meaning that on a successful one, we will have people from Accenture strategy that will be analyzing ideas and doing the proper kind of business cases around which products should enter the factory. Right. There will be some, uh, industry experts coming from our consulting groups that will help on the shaping of those products. Also, for instance, help redesign how we do the HR function to be successful in the type of profile. Right, right, right, right. Um, and we will have technology people that will actually man the delivery teams in the first phase before the organization had time to recruit their own people. So there's, there's sort of a coordination of multiple groups that are involved in this.

So there are not, people are coming from many things, but there's one single culture. And actually one thing that is absolutely key for success is that if you want to be successful, the loyalty of the people you've assembled, they should come to the foundry first before looking at what badges people are wearing. Yeah. So actually for us, they're really pure partnership where we cannot really differentiate when you enter the engagement, who's coming from Accenture and who's coming from the customer. Right, right. So, so, I mean, you were mentioning a client that you have and having worked on this, but like going forward, uh, like are, are you all assembling kind of like a repeatable process for this? Absolutely. Like, like, like, and, and well, are you validating that idea? There's almost like this recursion of like, you know, lean feedback loops that goes up and up and up, but how, how are you building out that practice so you can take it to other organizations? So, uh, first of all, so I'm mentioning one client, but actually we are probably a dozen or so operating ones worldwide at the moment of different size and shapes. Okay. But that's already a thing. Okay. Uh, what we're doing at the moment is really, uh, let's say capitalizing on our various lessons learned so that we have a bit of a guiding boot on when we start up one, we actually can implement the, I mean, we can get the best from the lessons learned we've had from the various, uh, various engagements. So that's, uh, that's one. And then after we do what, uh, let's say, uh, uh, services organization do, which is really, uh, making sure that all the people on our account are equipped about the information, about the offering to start promoting it. And actually, uh, if they meet the client that wants to be successful in the digital transformation, they understand that there's such an offering available, right? You need a foundry for a foundry to sort of productize it. Yeah. Well, that's interesting. Well, great. Well, thanks for going over that. I mean, like I've mentioned, uh, a couple of times when we were talking, like I see that pattern playing over and over again and it definitely, I mean, I don't want to be like a hundred percent like true in the statement, but I'm pretty sure it's pretty much

always what you have to do to improve the way you're doing software and products. Cause it's just very difficult to do otherwise. And I think, I think, you know, several of the key things you outlined, I see as well, where, uh, you want to take enthusiastic people, right? Like more or less volunteers like at first, and then you need to start small and pick a few applications. But then importantly, those applications have to be, I don't know, to use a highfalutin business term material, right? Like they need to be actual business facing things that can have results. Yeah. And there's another thing that's key actually that's quite close from, uh, the way people think as well is that you need opinionated people on that. Right. Meaning you want to bet on the fact that you know what, if you want to be a product in the foundry, you need to be cloud native. Okay. Right, right, right. We only do agile things. We only for us an initiative that allows us to eight sprints. Actually, we're going to slice it. Our maximum team squad is eight people. You know, no, you cannot be nine. Yeah. That type of things, you know? And, and, and that's really something on which our, our respective organizations actually share. Yeah. Yeah. No, like you're saying, there's a certain, there must be some phrase for this, but there's a certain, uh, performance advantage to saying this is a bucket of things we're not going to think about. Yes. Right. Which, and, and what, which are essentially opinions, right? Like we have, we always do pair programming. We have a maximum of 12 people and just like, let's not spend a week considering that, right? Like we could, we could reinvent this stuff everywhere, but yeah, let's just go with this for now. And that way we can focus on this whole other set of things to train ourselves with. Absolutely. And the thing as well is that this also makes you super powerful when you talk actually to business departments that need to be converted because whoever they talk to, they face that, those shared beliefs. That's right. Right. You have kind of like a unified system. Exactly. Well, great. Well, if people wanted to find out more about it, where, uh, where would they go look? Um, so it's not a public website type thing yet. So at the moment it's more something we communicate

on internally. Uh, however, uh, if people are engaged with some of our accountants, they can definitely get a view on the offer. Right. All right. Sounds good. Well, thanks for going over all that. That was really interesting.

Tanzu Talk · 🍕 Get more pizza.