Technology, meet Strategy: The lessons from Microsoft’s Digital Transformation with the Chief
Welcome. Everyone, can hear me perfect. All right welcome to what I hope I know you can't see us because, we're nice and low down here but, hopefully, to be actually a really great discussion I'm thrilled to be here I'm Julia white and I, am the corporate vice president, and asier. And I'm marketing the business side of things and I'm really, pleased to be joined by Kirk Delaney who, I had the pleasure of working with and for a part of my career but, Kurt is actually we're just doing the math up here even the Microsoft for 25, years, joined. A company on. The windows business and then, from there into the, office business and really birth what is you know sharepoint among lots of other things and then we're leading, the helm when we moved into office 365 million. So helping with our cloud journey when I had the pleasure of working, for you during that time what's the pleasure what. Was it was it a pleasure it was a pleasure always a pleasure. I'm. Not just saying that, and. Then you, had the very interesting opportunity, to actually, leave Microsoft, for a short period of time where I remember you retired and we did retirement videos for you but that was all lies because you turned up and worked actually, in the Obama administration, on healthcare.gov. I did and that initiative there and then we, are the great fortune of having you return to Microsoft, in your current job which you, win the longest title ever which. Is the, chief digital officer, and. Executive vice president of, the, core Service, Corps, service engineering, service no wait let me get this right you should say it actually of course core, services engineering, and operations I'm, done corporate strategy not into that whole brevity thing no you're like all those words I only know the acronyms which is why I had to ask what it actually stands for cuz it's too long to speak but that is what you do now so let me turn it over to you for a second to you a little bit about you and your career yeah, so I, started, in product development I've been a product developer most, of my life actually. My engineer and my education is all in operations, research so, I'm a math geek, but. Worked, at Bell Labs and then came to Microsoft, and. Came. Up through development ranks most of my time in office so, I ended up leading office, as the president the office division and then. Healthcare.gov. Crashed. And my, wife who, is actually a member of Congress. We. Were sitting in the coffee tour at the breakfast, table and I was reading New York Times article and I said, it's really clear as to why this thing crashed and she then said well would you be willing to help and. I was leaving Microsoft forgot what else I could do with my life and. I. Said, jerk and literally, a couple within, a couple weeks takes a little time I was sitting in the Oval Office talking to the president and he says I unabashedly. Begged you to do this and. I learned afterwards, that well you don't actually go into the Oval Office unless, you're ready to say yes oh and, so fortunately. I was willing to say yes and so I spent about seven, months with, a very, talented set of people who turned, around healthcare.gov, so, I was kind of the president's guy and briefed him was, a member of the team kind of driving that but briefed him in the Roosevelt, Room once a week, and, so, I spent about seven months doing that when they got it back on track or we got it back on track going. Into the next season, when everything was in track I left, and. Came. Back to Microsoft, so, Sachi originally. Wanted me to or, I came up back. Leading. Corporate strategy because I kind of know the company left-to-right pretty well after 25 years and then, he came to me about a, year year, and a half ago now saying. We're. A pretty old company, in terms, of our core systems, were pretty fresh in terms of being a tech, company I want. To embody, digital, transformation. In how we run in the internal, processes of the company so, the world we can be a model or feel, proud of the way we do digital transformation, and talk about to customers so he, asked. Me to lead both IT and. Business. Operations so how we transact, all the enterprise agreements, worldwide that's, part of my team as well as traditional IT and, so that's what. We now call core services, engineering and operations used, to we, used to call it MS IT and, is kind of a reflection of we.
Think It is core services, there's not a difference between us and the azure team except for we're, they, run. Products to sell externally, we run all the business processes, that drive the success of Microsoft, yes. Now, just, start with you for just second then we're going to talk obviously about the transformation, you're driving across the company you are chief digital officer you're not right CIO. CEO. You have that title what does that mean and what's the mindset you have with that approach well, when, I took over the, roll that. We actually did have a CIO very talented CIO Jim Dubois and, but. I had to figure out what is it you know Sachi basically, said digitally, transform, our processes, and he didn't give me a lot more indeed suggestions. To what to do and so, I, actually, kind, of took a systematic, look at it and said well what are the processes, that we run the company based on and, I. Said okay there's 16 core processes, who run the company based on all, you, if you're going to be cynical about you say all I know is product development so I'm gonna take a product development point of view on it and say. There, ought to be a core vision for each one of those processes that. We think, about it's like what's, this trying to enable what, does it mean for Microsoft, to be, digital. End-to-end in terms of our transaction, again jhin's what are we trying to accomplish so. Define, the, vision define. A release, roadmap, again just like a product team define. Metrics we're gonna measure success on and so, for me the did being chief digital officer is about, taking, the notion that there are new to digital, technologies, which can fundamentally, transform, how you work and to, infuse, those into, a vision, of how those core processes, should run in the company so that's what we've been trying to do got it now, most people know Microsoft, and supposed to hear about our products, and our cutting-edge, innovation, we're releasing out through our customers, but the, area you run is what's 40 years old yeah, the inside guts. Of the organization. And so. You, know wouldn't love to unpack, what you've. Had to do from what we have today the, the org part of it the technology, part of it and you move, us into this digital first approach to that so let's break that down a little bit yeah so the first thing I mean people, think Microsoft's, a modern company we must have modern digital processes, and that is pretty far from the truth. The a bunch of you probably have had enterprise, agreements, from Microsoft. Seen. An enterprise agreement well.
There Are some that actually if you think about the different ways that you can purchase a sure can be hundreds, of pages long because. Of the way our systems, are built they, don't have the agility to say oh if. You buy a sure, this, way in this geography it's. Just an algorithm, that says how much you should pay for that and you know this world leading marketing, for sure and so, we have these pretty crufty, old systems, and they get embodied. And these you know it's like anytime you want to do an enterprise agreement you crack open a new penny you start writing and so. It's funny Amy hood that when I took the job she says oh you got to go to the rocks the regional operating centers, and see, all the paper and. I'm. Like okay, I'm going to the rocks and we have three of them we have one in Singapore one in Reno and we have one in Dublin and I, went to the rocks and I started, looking for all the paper and it turns out there's no paper anywhere, so I said Wow Amy's, just out of touch she doesn't realize that there's no paper we oh. I'm sorry Amy hood our CFO we vote fully modernized, we're all set well it turns out what we've done is we've, had our licensing, executives, scan. Everything, into PDFs, and everything. Gets transmitted. To us but it's really just a physical document and so. That's, kind of the state of some, of our digital processes. As they. Exist, so, part of what, we've done is to say we're, gonna go digital end-to-end from, the point where you, start interacting with Microsoft, as a customer, to purchase to. Understanding, what you've already purchased, to transacting. From Microsoft, to seeing how everything is doing and going from customer, support that, entire, process is gonna be modern and digital and so we brought folks together across, the company in that particular area and that's one of 16 such places we, said we're gonna have a vision we're, gonna take all of the people that, are stakeholders in that vision and we're gonna bring them all together into a room we're gonna say this is where we're taking this does. That comport. With what your view is okay, and in a lot of cases the interesting thing and I think this is something for IT, overall, getting. All the stakeholders in the one getting everybody in the room on a regular basis and saying this is where Tait we're taking things having. A point of view about not, just saying what, can we do for you but, rather we're, gonna get out in front of it and say this, is how we think we should transact, in the future and, then having, them react, to that it, does you don't you get out of this world where you are you, know an order, taker for this for what you want the services to be you're now telling them this is what we think the right answer is tell, me whether you agree or not in, this particular one it's interesting that we, actually had a modernization, plan, and. In. The very first meeting I still remember that your boss Takeshi. As actually. For, several months we haven't agreed with this plan and nobody. Said anything because there's such momentum, in the industry around or around particular. ATAR saying this is where we're going and so, it, allows people to say that, does not compose, with my worldview, we get together we solve the problem together and everybody is now we're in we have had the longest period of having a single modernization. Plan and commerce as a result so, that's kind of on the business side so imagine, it every Bisby area that, you're working in you, set a vision you bring in the stakeholders, and you say you hyoeun, it out and say this, is what you need and then they come in with their laundry list yeah and you say well let's talk about the relative priorities, of what you want to get done with, the vision of where we want to take things it's just like building a product again yeah so, that and then there's a bunch of stuff on the cultural side that that has to change this result of that as well for sure but getting to that and it wasn't things you and I've talked about over the over the few months of really, running this organization like, a traditional, Microsoft, product team which you have lots of experience doing thankfully, yep but bringing, it in and your stock top of the vision and how you drive requirements.
And Having. Subject matter expertise, versus, just what does the business need to take order so like, I know that's been a big part of the way you have a new leadership you brought so just I'd love to hear more about that aspect, of it yeah, it is it. Is a product, I mean people say it's, funny I came over from office, and everyone's like oh here's the office guy he's, just gonna make us look just like office and I still have this transgression that in fact I was in a meeting earlier today where I said well the way we used to do in an office is and, like. People are just tired of hearing it but it's actually the the parallels. To thinking about the. Business processes, as a product, are very very tight and if. You and the way that I talk about in a lot of cases is one we're gonna be vision led everything. We're gonna do is going to be driven, based on a vision we want to take things if something. Doesn't comport, to the vision then maybe we ought not to be doing it so that is very clarifying, and then the other phrase, that comes up over and again again, is the process, is the product and so if you think about it that way even what we do in operations, is a, product, they're shipping the end-to-end how, I transact. The end and how I stay compliant that's, really, just a product and, launches. A product, even for us the we launched new products, into the market as a product the way we launched business, changes, in like if you guys in Azure decide, that. There's a new way that we should license. Your, org or do compliance, around a sure we, have a launch team and the launch team says we're gonna launch that product and even despite, that it's a business process and, so that getting into that rhythm and it really as I said it's very clarifying, saying we're gonna stop doing this and, we're gonna do this I'll tell you it even goes down to our networking, team so, we. Have one called the core platform team underneath is directly, underneath me and I. Said that with the networking guys and I said and, I, remember, the first conversation, I said how do you think about, when we should change out our routers okay and he said every three years we, change we go through a product cycle we change out of routers I said, well what what's, your aspiration. For our network and what do you try and and nobody, to ask him that question before, really and I said well you must have some vision about where you want to take your network and he, said we'll actually do we, want to actually get we think that the best thing from a security, performance perspective. And. Performance, perspective is to remove the internet that, it's nothing but it slows, things down once, somebody gets into the internet they can go laterally, left and right and it's a so it's a security, vulnerability and, so we have this idea called internet first I'm sure we're not the only people if it got it but, we're actually gonna slowly we want to slowly collapse, our internet, into only the things that can't live without being directly connected, and then, we want to get rid of that too and so, even it's no, matter what area that you thought could not be a product you can still think about it as a product so it's really that mindset. Is. No you, know part of the the, mindset that was there I think because before the product teams like myself and others were kind of like we need this we need this which I'm sure many people in this room can relate to here's my business requirements, yes but how so you can change but how did you also drive change in the stakeholder. System to make, that dynamic different that's, still a work in progress so I'll tell you one area where we continue to be need. To drive faster, as an internal, team as we have these this great tool Dynamics, CRM and, yet, when we the, history, of how it got used at Microsoft, it, everybody came to us and said we want another report we want another report and by the way we're, gonna write the reports, ourselves. And just you, tell us how to bolt them into the infrastructure, that you have that is what our internal, core tools, called MSX and so. That. Created, this world that says we've got you know the, performance, issues, on some of these reports you can create reports that do in power bi that are not particular, performance yeah or and so you.
We. Got all this legacy, in this debt we need to step back and say what, are the, personas. Of the people that are going to use this tool just like you do you know in the development, teams yeah what's, the overall vision what are the personas, that are going to use the tools what. Is the priority of the workflows, that they need to go through and then, you have to bring it all together there's, like multiple, personas at one point they define 31, different personas, that's going a little extreme, but. There's like six different personas, and you have to kind of balance out what are you trying to accomplish what are the features that will actually provide capability, across multiple, ones that is persona so you get more leverage but. To your question directly, it's, a different, way for the stakeholders to think, about what you're providing them and, they, also when, they come to you and say we just need this one extra report, you, have to step back with them and say wait what are we trying to accomplish where. Is this in our relative, priority, and this even this was what we had at healthcare.gov, as, well we have this exact same issue where. Every. We had 50, different priority, one issues, or things that we needed to do right and here's, a team that actually hasn't shipped and working website and, you're. Still dealing with, you. Know all these different things everything's, p1 and so we instituted, something called radical. Prioritization. Where there's basically can be only one priority. One got, it there can be only one priority, - and, you don't actually work on the second priority until you've shirred that the team is executed, in there right and the first one and so the stakeholders, if they see that. There's a clear prioritization. And you're being consistent, with that will actually give you the space to say, - compromise. Where you don't have it if they just come at you with a bazillion, different priorities, got it I like the radical, prioritization. That use, that on my team when I need it now. There's been in the industry but also I think at Microsoft, to this, kind of where, it is IT centrally. Does it sit with the business teams that, push poll really, interested in your perspective with what you've learned but also what you see in kind of how it works right so, one. Of the things that when I took the job I said let's figure out the model for how other industry, how other organizations, like, us organize, I the. IT function, the operations function and it turns out there are multiple models there's the model that everything, is centralized, and then there's the model that says there is no IT whatsoever, there aren't actually very many companies like that but, there's some that have grown up from a tech perspective, where, they say every single product team just also builds, a nugget of, what our internal, infrastructure, is and so, we saw that you could do it one way or the other we actually centralized.
A Bunch, of it and it. Actually gives you that credibility, to be the partner with. The business but, then you also have to be a good partner basically, saying again. Going proactively. And saying I have a vision does this compose, with you et cetera and so we've. Gone towards more of a centralized, model but, we still have partnerships, like the the new modern stack, has actually done in Scott Guthrie's, world and, so pieces that are close to the business they, end up being in the business, we, have this very tight relationship with them so that we and, we work things out together so it, generally works pretty well, the thing I think is interesting is in in a world where digital transformation. Is core to every business and. Every business is the software is this then what, you wrestle, with and I've heard from a lot of CIOs is well. So-and-so an organization, wants. To build an a software, function within, their organization, weight and even the chief digital officer can be another part of the organization, I think that's tricky, because if. You think about it if you can show credibility. That, your team can build products, deliver products, and actually address the end user the, end user needs and the product needs then you can actually build a very strong software. Development. Capability. Within your organization, that, does both the centralized. Core services, but, also builds, the software, for those products, but, you have to develop that credibility. You have to change the way you're structured, to be much more around, how a dev team would be functionally, aligned there's a dev team there's a program management, team you have to look like a product development team if you're gonna get that product that kind of credibility, I think that's so interesting about all companies, Microsoft we obviously, have our products our technology, so developers, are in every business group we could be, completely distributed from an essential, systems perspective but even at Microsoft, it really does have the affinity, and makes sense to have as more central, systems. Which is mostly where we are which i think is right listing yeah exactly, so this pivot to what you started talking about if you, had to come in and you. Got the you, got as much information as digitally transform us right thanks for that boss great yeah so. You, had to come take your team and you transition. It to run like a product team so let's talk about that part of your team or like the structure or the mindset, let's kind of walk through those pieces of it yes so there's a few things you have to do one again, being vision lettuce is key so you have to have the mentality that you're running like a product team the, second thing is you have to deal with what, how should you be structured, yeah and, we.
Have Decided that we are a very functionally, aligned organization, underneath me and this is the reason that we decided to go without a CIO it's, not that we're not against having a CIO in fact in fact in the future we could have a CIO but, I wanted to have those teams directly, report to me so I have like five major areas that directly report to me and within. Each of those teams they are functionally, aligned which, means there is a software. Development team instead, of software, engineers there's a set of program managers, there, is site. Reliability engineers. As well and, the benefit, there is if. You start as developer in my team it's going to feel just like starting a developer and Azure and by. The way the, levels in the org and how you progress feels, exactly the same okay and so I think part of it is is really getting functionally, aligned part, of it is getting into this vision, discussion. With the various, stakeholders. Part. Of it is getting into the discipline, of like we do site, reliability reviews. And we do we do, services, reviews and we do a mano on a monthly basis, and we walk through all the incidents, and, we have everything. That you'd see in a product development team we have mirrored that because it fits in with the how we're, thinking about it got it the other thing we did and it's a very controversial, thing in IT organizations. Is how you're funded, and, the, way it's worked and I think it's very common for other parts the org is all, your funding comes in from other organizations. Funding, you business. Groups give that right they say okay I'm gonna give you a hundred million dollars or 50 million dollars and I want you to do want, you to build this with it that, is a really, difficult, environment. To live in most people don't get a hundred million dollars I was gonna say it's all a matter of scale. It's. Very it's, the exact same question regardless, of your scale, and. But. The problem you run into one, if you're one of those dev teams they, say well how do i budget, what my how, big I should be next time right, and so, they say I can't actually invest, in people and it actually gets you towards doing more outsourcing.
And, The, problem without so there's a number of issues with outsourcing okay. There's this reliability, of how much how much can i Bank on yeah and so so. That's one of the issues the other issue is you're beholden, to them they don't just budget, you to build my finance systems, they, budget you to build a particular thing, for them and if, you're going to be vision led you're basically, going to them and saying this is what we think we're going to build and by, the way we've got to be the ones who ultimately, kind. Of bring. Things together and decide what we're gonna build right otherwise you're incredibly short-term exactly, because year to year you're budgeting stuff. You build it's like a random, report, that some people convinced, somebody, in the organization was, important, enough to get funded for that year right the other thing you run into is if the, HR, team has, is going to is a finite. Budget yet, the most important, things to do as an organization are in HR this year well, then you need to actually pull partner, funding, from one part of the organization, over to the other apart and I can't do that if your funding is stratified, that way so, what we did was with we got Amy onboarding, hood to, say we're gonna centralize, all of our funding and we're just gonna be you're, gonna have a C SEO budget, now, she's very hard, core in terms of keeping that lean. And you know, kind. Of as a percentage, of our overall spend getting very disciplined, there yeah but then we get to decide where, it goes. Again in concert with the stakeholders, and that's been a huge thing the, other thing we're doing is we're shifting, the mix of, resources. Away from. Contractors. And more towards full times because. This, whole notion of outsourced development, there are very, good outsourced, team you know sis and I so it's really not a critique of them so, much as you lose connection, to them in terms of what your priorities are and what you're trying to build and, so in fact the contracting, capabilities, are designed, to be, that way if we're gonna outsource, and build this for us yeah the the for, them to actually not be employees, of Microsoft, there has to be a certain distance between, them and us and yet, that means there's a distance, in terms of the processes, that we build, together and we need these folks to, be core developers, we need to have them think as great developers, and so, we're shifting from, being the. More, outsourced. Than in source to be more full-time, developers, and fewer contractors. But all these come together around, being, it's, just the stamp and the way we act is much more like office it's much more like a short cetera got it I mean that's true we don't outsource our core devs from our office. What. Kind of man would ever do that right same mentality in fact we had a Steven. Sinofsky who, I learned a bunch of how to run office, from you, used to say if it's core to what to what you build, you don't outsource it when, the outsource the stuff that, is a peripheral, piece yeah or it's something that's a surge thing that you expect to go away over and and we try to were trying to get that discipline as well yeah but going back to another route that got you there was around funding right, and if you are you, know being product, oriented vision led here's what we're going then you, know going and getting you know 10 cut money each year and then having an inconsistency, or viewer would constantly, undermine, that you, took the approach of to, work directly with the CFO correct, sit there and say this is how we're gonna run it here's why so I assume that took some work but then, you. Know how. Did that exactly happen and guidance of other people who said want to take a similar approach just thought sit on that one well the first thing that happens is well it's you start you get the post the organization, and the thing that came up over and over and over again is that our funding approach was busted okay and then we have our system at Microsoft, we have a pretty, senior, person.
In Amy's team that is correctly my finance, person okay so Victor and I talked, and said this seems fundamentally, broken and, we actually went together and said well maybe we oughta pitch that it should work a different way and then. We wanted to get our controller, on board so and the person who is responsible this. Is one thing if you want to get it sold first, thing to do is get the finance person who, defines. What finance, needs for, software, systems get him on board. And. So we got Frank wrote on board he progressional II said okay I'll ending, model will work and, then and then we talked to Amy about it and got it got it set up and got it you know and then we did it on a fiscal year boundary, it, was too much to actually just kind of recast. Everything, sure but, even then you get a certain confidence. Because, you know that it's going to change the fiscal year boundary, that you can spend you can start spending at a different guy so you started by establishing, with your peers that this was a root problem yes we wanted better internally. We want to do our own digital transformation, this was a necessary change. And then went, through the smart, tactic, so yes, we're. A pretty collaborative, boot Bunch. At Microsoft, and so if, you come to them with a latch logical, argument as to why you need it then you. Also have to deliver for him right you, have to deliver on the priorities, I actually think you deliver better on the priorities, if you're, clear, on what your North Star is so. That's helped as well now, in a key I think you know delivering, on that and, making, sure people see the value added one of the things you guys have embraced is this idea of you, know first party they're probably really nice using, our own technology. And such that we can be our best customers and then give that feedback back to the product team so interested. In that part of what you do and how you think about it yeah, there's a couple of things there first we are and. We know the stats deeply, we are 95 percent on Azure now and, so every. Almost every statement. Of all Microsoft, that is a statement of all the business processes, that run Microsoft are now ninety-five percent of them are running it and that's also includes, our sa P instance, we. Are one of the largest sa P instances, in the world and, so we're running. Wholly. On Azure today, we actually have the system running on premises, but all it is is getting logs shipped to it so, that we can keep we can track the consistency, the one in Azure is our mint is our master, I'm Eric and this is really just kind of a dupe until we decided to shut it off um so, we we deeply, embraced asier we. Actually. By virtue, of doing that we, have helped the azure team build a bunch of the system the bunch of the reporting, tools, etc that, allow us to see are, we spending effectively. How can we reduce our spend we think we're saving, somewhere, around 10 to 20 percent of our overall cost by running on Azure instead, of running on-premises. There's. A different security footprint. As a result because you're not thinking about again you got this massive intranet, that you're trying to guard against, and as I said we will eventually shrink that internet, down to to. Something very small. But. That was one, example it's like everything's, going to be on Azure and we're gonna drive cost savings, from that the. Other thing is we have I, think. By being again focused, on the vision we've. Done that in every area including, what our desktop, looks like ok so we basically said, instead of the windows, guys saying, we, want you to dog food this next we, basically say actually, yeah that's a great thing we need that and some stuff we say no actually, that, doesn't compose, with our view of our, feature set that we need for our internal folks and so we're not going to help you dog through that so go find a different customer from which that's something. That's critical for them so we. Get feedback in itself exactly, well it is true and so another, thing we've done is we've defined our vision, in a big document that says this is where we're heading yeah and now we've flipped that and said what, does it say about our product, vision and, so we're actually taking the learnings, there and and getting, them in a way that's very consumable. By our product teams yeah and so you can think about there's this flywheel, that goes on of we. Define what our vision is we, then get our stakeholders. To talk. With us about refining. That but, we also get the product teams we get the project feedback to the product teams we get super real about for instance we're. Moving from SCCM. To manage our desktops, to moving everything to into ok, we now have a very clear roadmap of how we're gonna do that over a period of time yeah and what that deployment looks like at Microsoft, we've, then given feedback to the the Intune guys Brad Anderson intend of saying we, need these sub features by the way this feature doesn't work as well and so, we get it helps, Microsoft's.
Products Get better and we, get this clear trajectory for, where, we're going as a result so that again. Is this place where it refines. The product strategy and then the other thing we do is we we. Consume a lot of third-party products, as well yeah and we, make a very clear distinction, about where we do one versus the other okay and we, particular, in the biz app space yeah we talked with James Phillips and say hey we, think this is a new opportunity to you we're spent we're paying a third party a bunch of money for this we, think if you tweak these products, in this way that, could be a first party product is everything you know everybody's trying to push us to say use, less third party use more first party they're places we can't do that but there's also these in-between, places where we can make our products better yeah, and then be, better for you for everybody's, consumer, as well got it can it be that internal, litmus test of yeah it just it again everything, hangs together better, and it did yeah one. Of the one, of the areas that I've been excited about it is your push to get rid of passwords, yes which is another, good example of using our own technically but also pushing things forward that, then we can turn around and help our customers, get to you so I lived it like that one is it you had mentioned that one he ended up here about that right so. Security. Is one of our areas we, think of it of like I mean how many people could articulate, crisply, what, their strategy, is in security, it's, hard because you got inundated, with a bazillion, different product, and it's, a little bit like a belt and suspenders model. Right so if you don't if I have a belt and suspenders my, pants are less likely to fall, down and so let's buy one. Of several things and we'll just make sure that we've got we call it defense and depth but. Right. It for a really a belt and suspenders but. What. We did is we said there are very, specific, things like identity. That. Are a core, to, our. Identity. Telemetry. A number of things that are the core elements, of our security posture yeah and then within identity, how is it that we're going to actually reduce, the threat that comes from identity first it's a recognition, that the, boundary of your internet is not the boundary anymore it's all about who can get in and out things, based on identity, and then, you look and you say what's one of the biggest risks. That you have it's, the fact that people have password, and, people have passwords, that they write down and then, you, know everybody, decides that there's some manifestation, of seahawk that's a password and it's. True if you if you search and you just histogram, of the passwords we in, use in your organization, there are going to be some that are highly, highly used yeah and by the way the culprits, that are trying to get in know what those are and, so there's. And, if that and the attackers, are getting more and more sophisticated. So, Brett. Arsenault said that the goal here is there, should never be a password, yeah and so, we articulated, a vision to get away from passwords, we, then say, okay what are the pieces that actually compose, to get you there and for me at least the penny dropped and said well the purpose of hello is not. To actually make it easier to log in but, to get you closer towards, a point where you don't have passwords, anymore great right. And so we're moving to a world where, right. Now we require, people to change their password, I think every three, months we. Will get everybody, towards, biometrics. As the way you log in we actually track the percentage of our users that log in via some biometric.
Ability, As opposed to their passwords, yeah we will then go to machine, generated, passwords, which will be long and cumbersome, but people have to you know find some secure, place to, have those and then, we'll get to a point where. You. Will only have to change that password once a year, eventually. We'll get to the points where it'll all be certificate, based and the only thing you login is based on biometrics, and at that point there are no passwords, anymore it's all based on certs that. Are passed around and biometric authentication, which. Is the topic we talk to our customers about all the time totally and for me so it really it it allows you to actually say okay all these things that you look at all the products that Microsoft's, shipping, all the stuff, that's in the, security, space step. Back and say what are the core foundations, of what you're trying to do and for us it is, identity. As the perimeter, it's about great telemetry, and what's going on in your network and it's about information protection. And if you those are kind of the stool with a top. Of it being around great policy, in place, as well and when, we look to figure out how can we accomplish each of those in its best way got it so, I want to go back for just a second because I realize didn't touch on this one and he talked about. Going for more of an outsourced model into an in source model so you're hiring a lot of new developers, into, this organization and, I was curious are you looking for the profile that you would have hired into, Windows or office in your past life are, you looking for a different profile, how are you thinking about that in terms of bringing. Devs in because I a lot of customers, I work with they're doing similar like I yeah you. Know for this group of people what are you looking for and now what's the mindset you're looking to hire it's a good question so we. Are hiring at a massive scale right now because of this transformation interesting, we're actually doing it in and, being revenue, neutral or cost neutral okay so everybody, has embraced the notion that I will get rid of to. Outsource, people, it's a higher one insource person and that, that makes it revenue neutral for us got it saving. By running on Azure by the way we are and actually the other great thing it's, a good thing you pointed that out if you own your own budget, you get to plow that that, savings.
Back, Into some other investment that's a great thing as well yes, so. We're, hiring traditional. Debs, we, are also hiring new disciplines, like we now have a centralized, user user, experience, team and user design, team and user research team, we. Have a centralized, data data. Team that basically says how, is our data structured, across the organization, how do we create consistency, how do we build a platform so, that on Azure that we can do machine learning against. Machine, learning we are doing it's infused across, the company of like every. Place that you can think about being better in your decision making is a place that you can use machine, learning so, we, talked about it today in terms of what, you might think as mundane, things of like, what, is the flow of York, of your, support. Instances, for your internal support when, is it likely to peak what, are the particular issues, that you're going to get like it we, find that if you change if somebody's, password, there, is a higher probability that, a week later their. Remote access they're going to call you up about remote access issues, got it and so there's all these predictive. Models. That you can build and so we're hiring a lot of data analysts, and people, who are data scientists, to be able to do that modeling yeah. The. Other thing we're learning is that we are a great, gateway, to, other parts of Microsoft, developers, so we're hi we're also moving much more towards, having, a deep, presence, in university, campuses so, that we can hire people in younger, and career it changes the whole complexion of, what your organization's, like it's got a lot, more young people running, around with new ideas and with. That energy, so we've much we're moving to we still have to hire a lot at all levels but we're trying to build up our university, recruiting and then the final thing we're doing is trying, to figure out alternative. Sources, for, talent, which, we believe we can actually take people with a little west capability. But the right aptitude, okay and, hire. Them with a little bit more training and give, them an on-ramp, towards.
You. Know towards a full-fledged dev job at Microsoft got, it so because if you do have all this demand you've got to come up with new new sources right speak, about the programming run with the military and they're tweeting people. Who are leaving the military to bring them into it right we're going the, other thing is it helps with your diversity and, inclusiveness. Questions. Around finding places, that are you, don't recruit as much towards we thought wow in are there places in Latin America we can recruit especially, if you think there's a set of a. Kind. Of come, in induced, the base level education that. They may not have gotten at least contextually. And that they need in the United States, like. Particular dev tools that we use for instance if we have smart people we can bring them in we, can we can help there as well yeah that's great I think in general the the growth in demand. We're. Gonna need to be more creative and independent, and if you think about it once we show up the university recruiting, and the university recruiting guys say well. Great I've got the Azure guys that are trying to hire at a record pace they've got the office guys that trying to hire to record business so we got to come up with diverse or different, ways of, getting. It accomplished, got it okay, so now. 95%. Running in Azure your internal, systems that, didn't just happen overnight it didn't it didn't happen in one big chunk so, how did it what did that look like and your prost that that my grace help. Mine for lots of people yeah, so we did it in multiple stages we said one. This is the opportunity, to actually, just get rid of a bunch of systems that you have okay and so the first thing we did is we did not choose to just take everything, that we built and ship. It over and put in a VM in the cloud okay we said this is the opportunity to clean up and we took a set of apps and said we are gonna sunset, these things really fast okay, and so, that cleaned up a bunch of cruft no how did you how did you decide those we're cruft we, found out how many people logged into them. We. Communicated, these folks and said you know you had there are two people who use this system every month we think we can provide you a different way or by the way did you know there's already a different system that does exactly the same thing, that's, right there's such propensity, to, never get. Rid of a system because you sure somebody's gonna squawk ah so, we've a state, yeah it is and so, you basically do the toiletry you find out which ones aren't being used very much we can have alternatives, and you sunset a bunch of stuff what. Was the percentage ballpark, I think it's like 20% of the total set of systems that we ultimately got, to the point where we can basically just Sun Sony Dominion and we're still killing a bunch of them and that's like the 5% part of it is the stuff we're, just trying to get rid of and, it's not eventually going to move, then. We took a, set of things and said which ones are we going to redesign, and, there, are certain things that are massively, cheaper, if you use, a pass service, instead of you just take this VM and ship it to the cloud so, imagine this thing and healthcare.gov, was, like this as well and they did a total on-premises. Deployment, which was wrong, the. Pinkness of that if like and you haven't earnings at Microsoft, that's a big piece of it it just goes like this and so. What you want is the ability to surge, you. Know up and down now you can do that by throttling the size of VMs and sleeping things as well but what about those systems that you just never use, except, for the random transaction. That comes in and that needs to be processed and put in the data needs to be you, know needs to kick off a workflow and the data needs to go in the right place that is a perfect, serverless application. That, you will save a ton of money if you just rebuild.
That Application, as a pass an able application. So. We took a bunch of ones and just said these we were gonna redesign we've always wanted to redesign and we're gonna take this opportunity, to do that we, took and the rest of them we took and said we. These they were probably going to continue to be VMs we're gonna take those VMs we're going to move them up to the cloud and we're gonna do a prioritized. View. Of which ones like we had some there running out at capacity on premises and we're going to go through the Capitol cycle again and I said those are the ones we want to move sooner and so we then just did a prioritization, and moved him over a period of time the longest one to move because. We were early in the cycle of figuring out how to get it done was this ap okay and even, there though and now, we think we could do it in eighteen months the whole thing soup-to-nuts we could move in eighteen months but, at that point we said let's start to move some of the s AP workloads, over and let's, keep the core core. Piece that is a transaction, engine, or in general ledger and we'll keep that for last and. Then we in, February, of this year we moved it over as well. Most, well I think four oh the actual movement of last piece was, about three, months three or four months yeah by the front of the time we started the it the, actual move takes a weekend or. A week actually, but. It's, you know all the planning for the core core ledger nark today it took about three or four months got it okay and, so I, don't. Know if you win that started in the very beginning but just from a ballpark, of taking you, know a lot of systems are Microsoft obviously, to where you are now what was that whole and endless talk okay the, Asscher, has evolved, over a long period of time but we've been added about three years in total but, if you think about it early on we're, in partnership with Scotts teams of saying you, know what does it take to do that okay cetera, and so it's not probably not a fair comparison area, again, I think if you took s AP as one example you, could move a large chunk of your infrastructure, over in an 18-month period got it okay all, right one last area before we can open up for questions.
There's, Definitely a lot of sense of how do you get, the ear of all the businesses. And have that, relationship and that trust and that kind. Of ongoing approach. What. You're the model you've crate sustains, and so just interest in how you've accomplished, that and what that what you have to invest to continue that I think. Again, it's I do think this notion of being vision lead helps, because, then they. See the, roadmap, as opposed to all their people coming to them and saying you know, I want this from you I want this from you and having people come to them to, them as leaders and say here's, all the demands we have on the central organism, right and so it, gives you that air cover to say this, is where we're going and let us slot your priorities, into that context, of where we're going I, can't. Emphasize enough how, important. That is the other thing that there's no you, have to do is you have to just deliver. Form and, right, so you have to set this, product, release roadmap, we've, gone to all a j'l development, ok so basically we run in sprints everybody, understands. You know there's going to be a two or three months sprint we're gonna deliver on that timeframe, we're even working with how we do that in SA p s, AP historically, has been long lead time to get new functionality, out there, more, of a kind of a traditional, model in terms of how they deliver stuff. To you but, we're coming to the point where even in s AP changes, we're doing that on a sprint basis and so they start to see things come in a regular basis, but. I actually think that again, if you if they have that confidence you you start delivering yes, then the, confidence builds on itself and, they. Know, it's important they also know it's not there their strengths and so, if you feel modern you talk in terms of product, as, being the central center, of it yeah they kind of get that credibility I will also say that hats, off to the Saatchi of saying this was a priority for us as a company and such' talks about it internally about the importance, of digital transformation for. Us internally and you have to get that air cover with your with your entire senior leadership team as well goddess fires. Bring layers one of that, executive leadership, and. Vision, for. Then. You having your own vision so people can see the full context, you're working on in the priorities, right and then you had to become, an agile and deliver right you like you had you couldn't say yeah I'll get back to in a year with your requirement, but I have a sprint coming up will slot it in just like a product two minutes yeah and you know the other thing is everybody, in the team can act as a, as. A person, promoting. What you're trying to do as well and so if you can get everybody on board to say this, is where we're going as an organization, we are fundamentally, changing, the nature of the organization then. Everybody, becomes, you the biggest, advocate, for oh let me tell you what's going on and, the other people in the Microsoft example they look and say wow you, know you're just another developer, just like me you just different product that you're building to me yes and so you have to get that grassroots, view.
As Well I will say that's one thing that's been really important. For our transformation, which we haven't talked about is. The coach, the old IT culture, has to go as well this, whole notion that, I have, my systems, I own my systems, I run my systems, and I. Through, the might of my voice, I, am, gonna kind. Of push you into into, this is what we're gonna do that has to go away it, has to be replaced, with the, logic of your answer is the thing that drives every, voice is heard we, spend a bunch of time saying, what is what what, is good hygiene in a meeting how, do you make sure the people who are more introverted actually, get a voice we. Spend a bunch of time at the end of meetings saying has everybody had, their opportunity, you know you don't have to do it in a way that feels like you're following a process you basically look around and say look, and say is there somebody else that's looking to say something here yeah exactly. And, so we coat, a lot of the inclusiveness, aspects, can make its way into traditional, IT as well that's fantastic to see yeah all. Right we will open it up to questions that you might have and. There's a mic in the middle of the room or we can pass it through to there's any topics. Yeah. You mentioned that you are funded, internally by a business units I think you said. And. Just back to that does that mean you within, those units that you don't fund. Based, on strategic, initiatives. Or, do you do and if you do and those, strategic initiatives. Span a year boundary do, you fund multiple year and then, revisit on the annual basis or network, no it's a great question so we. Do we, fund the whole pool and we, do not underneath, my organization, we do not have buckets bucketz, until it's actually allocated. Based, on our priorities, to the different teams we. Keep a list of the priority, projects. In order, again. Radical, prioritization this is the most important thing we're doing part followed by this boat over this and then. We have a cut line and we. Talked a lot about the it's kind of in the product teams it was the last. Yes in the first no and it's, the exact same thing here well heads what's the last thing you chose to prior Tai's and what's the to do and what's the first thing you chose not to and we, compare them across the different teams to make sure there's a consistency, in the overall priorities and a certain level seems right so you're not doing something that's really not that important, over here versus the other in. Terms. Of funding across years Amy's, made it really simple to me she, basically says you're gonna be flat and so. Basically. I know I've got and and she'll probably try to noodle it down and even a little bit more than that but usually I can defend myself to being largely flat and it helps as all organizations, kind of grow in size your, percentage, of your spend actually goes down so that's a good thing but, then we figure out because there's always that cut line by group a multi-year. Thing we, get it funded by saying if it's high enough priority it's one of the first things that gets funded, so. That's that's how we end up will you still revisit that though we revisit, that quarterly I'm. Saying do because again things, it. Doesn't stop asks, from coming in and asks, come in and you have to say where am I going to prioritize, this unless the other thing I think is really, critical is you.
Have To be like everybody, in your team leaders, all of your leaders have to be the always, on the lookout for wasted. Spend because. The place that you get more capacity to, spend, is by, finding the place that you're overspending. Today and we've, we have enough scale that it's it's been fairly easy to find that spend but you'll get tighter and tighter in a number of years if we've got that all really well honed down then, it'll be harder but right now we've, found a lot of that cruft that doesn't have to get spent. Ok one other quick question is that yeah so, you mentioned you're transitioning, from sort. Of contract. Workers to a more. Full-time work for Microsoft entities. One. Of the one of the challenges, organizations, have is that generally, ends. With a longer. Term bloat and yes, of benefits, that are required and retirement, things, like that yep how do you justify, sort. Of that longer, term 10 20 years yeah outcome. When. You're moving in that direction yeah, we call we have an affectionate. Name for that we call it the bow wave and, it. Is real and Amy, says we have to self-fund our bow wave so. We have to find efficiencies, that. Will pay, and it's pretty substantial but. We have to find efficiencies that, will pay it as far as the 20-year thing this. The site the the shape of your organization. Kind of a self-healing because people ultimately decide. That they want to retire and enjoy you know some, some life not working but, you also have to look at what the bit what the curve looks like and if, you aren't Rhian if you're not. Really. And, rebuilding, organization, all the time it can get much more heavy on people. That are more senior and there's lots of look, at my age I'm hopefully. I'm delivering, on the, neon on, priorities. But, you can get to the point where you have a lot of very high paid people and you. Have to again, the university recruiting, building, on that actually, helps a bunch there because, it shifts your cost curve a little down towards, that it has a lot of benefits including they're, less expensive consents. They're earlier in careers, so generally hiring younger people. Helps with the bow wave. But. We we self under a bow wave and we use efficiency, to. Self-fund you're, able to say short-term. Decreases, this year are, more. Than justified because of that curve that it's happening moving, ahead well. Every we take it a year at a time right basically. Victor my my compadre, on. Finance. Tells me this is the bow way if you got a fund and, then we find out where we're gonna fund it okay thanks sure thanks. Hi. Curt hey. Excuse. Me I'm. In a lead to engineering, role, where. Frequently I'm not part, of management, but, I do have the ear of leadership, and one. Of the things that really stood out to me early in the conversation is. More. Than just going, to the stakeholders, and asking. What. Do you need. Saying. You. Know here's some options, I've got for you um, do. You have any, sort of advice for someone in that position. Not. Other than to say there, is I. Think. It's in my blood to actually think in terms of. Prot. Everything, kind. Of accruing, to a product vision and where you want to take things it's the only way of you know in my whole history it's the only way of to build things and so, we get in these meetings where people talk about, oh we, want to build this feature in this feature and I'll actually just stand up in the meeting and go to the whiteboard and say, okay tell me first what's your aspiration. For the customer, what, is it we're trying to accomplish and until, you get anchored on that in a way that actually somebody is able to articulate, it then.
All, That you're talking about is actually just you, know blather, in some sense I I need to be anchored on a very clear vision of what I'm trying to deliver and then, I get down into what's the product ization plan, that I have to deliver that and I think as a lead developer I. Don't whether you have whether you're structure so they have program managers is or product managers is similar. And, we, call program managers what most the industry calls product managers but the idea the people that define, what the feature what, the product should be don't. Let them be. To, come at you with a litany of features and if they do and, you like. Say well I don't understand, how that composes. To our overall, priorities, have them prioritize, it for you tell me where that fits in your relative, priorities of what or if you want to go even further saying tell, me how your set of features that you've demanded, from me fit, into some vision of where you want to take things and so. Having tools like that and bring it all back to first principles I'll. Pee, Pillai always try to step back until. The the, problem doesn't change my, view of the problem doesn't change because then I feel like I've got the whole thing in my sights and until, you can articulate, what that overarching. Thing is all the stuff underneath won't compose. Thank. You thanks. Guys. Hey. I'm. Just curious is, there any precedent for using. Like a product development, strategy, to apply. To digital, transformation. Or is that just I mean have any of your customers, use that same kind of model or is it just because you're a product dev guy like it it made sense and it works I haven't, heard anybody who truly, did it just this way but I'll bet there are so, I don't. Think that corner. You, could just say that Curt only knows how to build products so that they this just felt like a product afternoon but I've been shocked at how well it actually. Composes. Really well and it is, Clee the same you, just have to change your lens a little bit the one company and I don't have a perfect view of everybody in the industry the one company that I've had a long relationship with, that, thinks, a, bunch. Like this is is royal, dutch/shell because. Those guys say, they actually have a different team in IT, that. Is all about where, our what's our vision for where we're going longer term in terms. Of our IT investments. And. So. He that's probably the team that's been closes it's interesting it mirrors their whole business because. The other thing they do is they actually went in an incredibly, difficult industry, like oil and gas they. Say where is the world going in oil and gas moving, forward and they, actually make bets many, years out that say we. Believe it's moving and. They're not massive, things except for they align the whole company around it of saying, we think fossil, fuels are going to go way over a period of time an imaginary land, gas companies saying that as a primary. Premise of what they do and then, saying so the investments, we need to make that, that will, keep oil Dutch Shell relevant, or. Look like this and so I think they probably since.
They're Very visioned, lead they, then drive it down. To their teams as well I'll also tell you actually getting crisp on the definition, of a vision, is really, really hard I was talking to one of the teams just this week of saying, and I felt bad about saying but I said you realize what you've defined there is a mission, not a vision, because, what you want to end what they end up wanting to talk about is we're, going to build the most agile. Up-to-date. Financial. Transaction. You. Know set, of services, etc so, that we can create, the most agile, experience. For our sellers, I said, that actually you didn't say in any of that what you want to build and the vision should be what you want to build and so it's hard and this gets down to the people involved as well you have to con today aren't at. Least in our work they weren't typically used to set into thinking as a product, team and so, you there's a different mindset that has to come out of as well cool. Thank you thanks, I. Think. We have one minute if there's a last, quick question we, can break - we'll take two bow tie. So. You mentioned, that you weighed across the top, and try. To, drive. Consistencies. It's, my question is how do you drive reusing, consistency, across a huge organization. Without, demanding, you must thou, must do, this and be. Able to have allow your team to own, their own destiny so how do you guys try. To reuse and reuse reuse, and basically, build, it to control. Your destiny I. Don't. Know how much well it's interesting for us the biggest thing has been getting to an architecture. That makes sense and so, we. Have what, we try to do is say. Looking. Forward, what is our data architecture. How, do I think about the, relationship. One of the big initiatives we have is a customer 360, which, says that we should be able to within this modern, and and digital system we should be able to tell our sellers, or tell our customers and our partners every. Shred of information relevant. For them to be better at delivering, to, the customer, value and the only way you're going to be able to do that is to, have a very, consistent master, data strategy followed. By a catalog, of what data exists, where a structure. For how you're going to create an azure data lake that you're going to do machine learning on and and, have that be a platform. That everybody, bets, on and so, there's a bunch of it that says if you get the architecture right, then, you, it's very clear, when, you're doing something that's off strategy, or that is it as duplicate, if the, second thing we've tried to do we're just doing it now this transition, is not done for us by a long shot but.
We're Doing a list of saying every, place we make an architectural, we, make a product, investment. Is, it first party is it, bespoke, or is, it third party