I just wanted to share the transcript from this week’s media day at Redmond from Stratechery (I don’t ever share premium member material, but I think this is worth the read).
Satya Nadella: Maybe to orient us perhaps, I can just share a few things about what we are trying to get done, many of you have tracked us in many different areas, have different points of view about us, maybe we will just at least try and talk a little bit about where we are and what we are trying to get done and then we’ll launch into anything that’s top of mind with you. Sound OK?
The one thing that I do think a lot about is, I don’t know if it’s a funny word to use, but it is the word I keep coming back to, is a core sense of identity, which has to obviously translate into a sort of a mission, which then helps us make decisions on products we build, how we build them, categories we enter, markets we enter, pretty much all of what we operate and this notion of when we say we are in the business of empowering people and organizations all over the planet to achieve more, to me, is I think the reason Microsoft exists.
It was Geoffrey Moore, I don’t know, maybe ten years ago, said: “Sometimes it’s sort of important for you to ask yourself the question: say you disappeared, will anybody miss you?” The other thing he also said is that “Sometimes it is important that some customers want you to win, and understand that source – why do they want you to win? Why do they want you to exist?” and the more you are in touch with that the better off you are. And you know, at Microsoft we’ve had our share of successes, our share of failures and challenges. I’ve tried to pattern match where are we at our best, where are we at our worst. I’ve always felt that when we have tried to veer off of why – what are things that come naturally to us? Which, by the way, have unintended consequences. I mean, for example, we think a lot about people in organizations and lament that we started as an end user company and then quickly transformed into an IT company and that made our end users unhappy, because basically we were sort of serving a different master, so there are some unintended consequences of even your core identity, but nevertheless even from a software sensibility, we think about institutions as much as people because we ultimately think that people want to build institutions and there’s complexity there, it’s easy to gloss over it and say it’s all about one versus the other, but the multiple constituency thing is what we have quickly come to realize is such an important balance to strike and not easy to achieve.
I’d also say, today in 2019, I think Ben you write a lot about this thing about Platforms vs Aggregators – we are a platform company. That’s why, and it could be quite frankly, that most of these aggregator companies will be the largest companies in the world and that might be the case, so that’s why I’m not one of those guys who celebrates some market cap measure because I think all of that is – I mean that’s just not stable, at least not with our business model – because our business model fundamentally is about creating more surplus outside us. We will only be long term successful if people are making more money around us. It’s the core of the company I joined, what, 27 years ago, that was the core identity and in today’s world where in fact the opportunity for us is much bigger, I think this if we can keep under embargo, tomorrow we are going to announce a pretty big deal with Walgreens very similar to what we did with Kroger a few weeks ago, but every day we now have many, many big – it’s kind of like an OEM deal of the past, it’s kind of fascinating to think about the magnitude of some of these partnerships is one we have done historically with HP or a Dell in the PC days. When I look at that, it’s great for us, but it better be great for the partner we are working with. In other words they need to become a digital company in their own right, their own digital ecosystem and platform capability and so on.
So that’s one thing that perhaps you should track us for that, you should give us feedback on it, but I have a view that we will only succeed if we live up to it, and by the way, this economic nationalism, whatever you want to call it, is not like some transitory phase which US and China may do a deal next week and then everything is back to normal, it’s done, the globalization phase is over, and at least we know it, and all over the world it is going to be much more complicated. The tech industry at large has been very, very lucky that we have had unfettered access to markets all over the world without necessarily contributing to the local economy. As a guy who grew up in a country that was colonized by a multinational and always is cognizant about showing up and saying I will collect rent and there is no local contribution, no local prosperity. Even today, I learned this was the first week I was in Microsoft, and even today I see that. Whenever I visit a community in the United States or abroad, I look at “Hey how many partners are there? How many people do they provide? What is their revenue growth?” and unless we can point to it, I just don’t see how you have a long-term, stable business all over the world. So that’s a little bit of our orientation, at least as a company.
Of course, all of that is great, but the question is: what do you do going forward in terms of technology and innovation? That’s another place where the point of view I’ve always had is look – there’s more computing need in the world, not less computing, because as things get more digitized, that’s probably the new factor of production that’s going to get us back to economic growth that’s going to be sustained – I’m a big believer in that which is in all of this we still need economic growth and our economic growth definitely has to be more equitably distributed for sure, but the source of it has to be innovation and that innovation in at least our times has to be digital tech and our vision for it has been that computing will be very distributed.
I love that Mark Weiser quote from Xerox PARC in the mid 80’s, which it roughly says that computing is just part of the environment, it just blends itself with the environment and becomes invisible. That’s what happened all around us, we describe it as this cloud in the edge and I think of it as three layers. The infrastructure itself being distributed computing, that we all hoped and tried to do in the client-server era and quite frankly got it wrong in terms of truly thinking of it as a distributed computing fabric that is much more pervasive. There was just no way client-server could’ve gotten to such a footprint, whereas now I think it’s much more possible and we can sort of talk about what we are seeing there. But then the real question is this compute is going to be used for what? It’s going to be used to be able to sort of compute close to data and then create intelligence which then will show up in a variety of different experiences. Even on the experience layer, all of the devices you are using today, my hope is that five years from now we are going to be able to transition away from device to device, sense to sense, and have it all be seamless. In other words, where the user context remains, the devices will come and go in a day in your life, so that’s a little bit of a broad vision we have around how computing evolves.
So then the question is: “Ok, that’s also fine, not that that vision is lost on others, there are many others with that similar vision, so that question is what is it that Microsoft can offer?” I know you today will hear from a variety of folks who have some specific areas but the broad picture – let’s start even on this infrastructure, what we are trying to do with Azure and our distributed computer infrastructure is bring that distributed computing to the world. I don’t think it’s a winner-take-all, that’s my belief, it’s a few players. I mean, in the client-server era we succeeded, so did Oracle, so did IBM, so did VMWare, there were many players. The one thing that is complex there is what I’m increasingly hearing, the world over, is it’s just not sovereignty in the traditional sense, but a lot more discussion of I’ll call it “operational sovereignty” – this notion that we will just trust an American multinational company on face value for all time to come is a tough one, so therefore being able to really say “look, we have the flexibility”. One of the unique attributes of the Azure stack, which is doing very well for us, is that, which is that ability to give you operational sovereignty beyond just saying, “oh your data is in your premise” is a super element of it. Everything from Azure [unclear] to Azure’s IoT to Azure’s stack to Azure’s [unclear], one computing fabric is a pretty unique attribute of what we have and that’s clearly something we are very, very focused on when it comes to Azure. The layer above that is all about AI, I don’t think today we are doing that much other than in an applied way with HoloLens you’ll see it, but have a lot going on with neural speech vision and NLP as just building blocks on top of it. It gets used by us in our own services.
The other area, maybe I’ll say, of course that’s one infrastructure identity we have. The other effort for us is what we describe as Microsoft 365, just simply what we are trying to do is bring home that notion that it’s about the user, the user is going to have relationships with other users and other people, they’re going to have a bunch of artifacts, their schedules, their projects, their documents, many other things, their to-do’s, and they are going to use a variety of different devices. That’s what really Microsoft 365 is all about and sometimes I think the new OS is not going to start from the hardware, because the classic OS definition, that Tanenbaum, one of the guys who wrote the book on Operating Systems that I read when I went to school was: “It does two things, it abstracts hardware, and it creates an app model”. Right now the abstraction of hardware has to start by abstracting all of the hardware in your life so the notion that this is one device is interesting and important, it doesn’t mean the kernel that boots your device just goes away, it still exists, but the point of real relevance I think in our lives is “hey, what’s that abstraction of all the hardware in my life that I use?” – some of it is shared, some of it is personal. And then, what’s the app model for it? How do I write an experience that transcends all of that hardware? And that’s really what our pursuit of Microsoft 365 is all about.
The one particular thing I am very excited about is what we’re trying to get done with Teams. Brian, who leads that team, no pun intended, was the creator of Project as well as Outlook. So in other words, he’s created many scaffoldings in his life which have been pretty transformative and category-creating. For example, Teams in its core function brings together messaging, meetings, as well as collaboration. So even splitting these three things into three different applications, it’s the scaffolding that brings these three things together. But let’s take NRFwhich is the big retail show happening this week, one of the things that we just did was said – ok, it’s not just those three things even, but it’s the line of business applications that are relevant for retail being brought in the context of your messaging, your meetings, and your collaboration and we are doing this in healthcare, and many, many verticals. So this notion of collaboration and communication, and you can say we can learn a lot from how China works, how WeChat works, or you can even say the prevalence of messaging tools on phones, broadly, whether it’s WhatsApp or what have you and how the blending two business applications teaches us this. But no one’s done a first class job of building something that truly then takes the power of messaging and collaboration but bridges it in a regulated industry in a place where the data is the [unclear] data, in this case when we use something, let’s say even in the case of Walgreen’s, or in Kroger’s, or what have you, or in Walmart they want to use this, and that data better be there, not WhatsApp’s data, or Facebook’s data or what have you. So that’s what we’re trying to get done and maybe Brian, do you want to share a little bit about your think on Teams and what you’ve seen?
Brian McDonald: The big thing that Satya talked to you about, achieve more, but it isn’t about achieve more yourself. For years, our goal in Office was, you know, we’d call our category Personal Productivity. But now it’s really about achieving more together. That’s not just about supporting real-time collaboration and meetings together, it’s about having that kind of work space where we can work together. So the central workspace, and the UX scaffolding for that, that was the key, one of the really key things and drivers of what we wanted to do with Teams, and have that be a hub for Office 365. Before what we had done was just taken all those personal productivity workloads and then moved them to the cloud, but we wanted something that was purpose-built for the cloud that could be a hub across all of Office and frankly across the rest of what we’re doing at Microsoft. A lot of the Power BI Power Apps Dynamics tools that James was building but also third party. So we built a platform for that and the third party platform and the first party platform are actually the same. And the new thing we are doing this week is, as Satya mentioned, is let’s try to take this ability to collaborate and achieve more together and bring it to everyone, because in the past we’ve been very information worker focused, but there’s far more of these first line workers around the world and the desire of companies to want to invest in how much those people get done, how they communicate, how tasks are assigned, how data flows between the information workers and the first line workers, that’s all there. So we’re trying to build that scaffolding that can work across all types of workers. So we have the basic Teams app, but it’s also got shift scheduling, it’s got visual communications, what’s happening right now is that in the absence of this, these workers are kind of self-organizing around consumer chat apps which has a lot of negatives for a company – they don’t control access to the data, there’s a lot of liability associated with that and just makes things harder.
SN: And that is one point by the way, we describe as first-line work, which is actually interesting, I had not seen this because we used to think it’s about knowledge work, and that’s where computing will first be intensely used. You’ll see this even when you listen to the HoloLens folks, I never thought that this first standard issue of a computer for some of the first-line workers will be HoloLens, for example, but it is because the productivity gains for service workers is such that it makes a lot of economic sense where they did not even give them a phone but they’re happy to give them a HoloLens. And that’s happening broadly even with Microsoft 365, essentially we now have subscriptions of MS365 in the enterprise that are not driven by the traditional “Oh I use Office on Windows, therefore I want to continue to do that” they are being driven by “Hey, I use Teams on my phone for my first-line worker” so that’s a very, very new economic shift I’m seeing.
Question: You anticipated one of my questions which is: so Kroger’s out there in the wild already, I assume Kroger’s uses office.
But are you saying that you have major enterprise accounts that don’t?
SN: No, this is about Kroger now saying for the person in the store, on their phone, are now using something like Teams as a way to coordinate the shifts, coordinate the merchandising, all of the stuff, which in the back of the room they had a server, if you think about the traditional retail, it was a server in the back of the room, there was a Point of Sale and some PC that someone used. Whereas now, everything is happening on the phone, but the question is what’s that user experience scaffolding that allows you to interact with the core business and does messaging in a secure way.
BM: There’s over twenty airlines using it, airlines have another example of that problem where they’ll take seven or eight cockpit crew and stewards and they’ll have five or six flights together over a period of days and those people want to come together and have an environment where they can interact and share data, the company wants to share data with them if there’s a change in their schedule or change —
SN: It’s actually pretty cool, like every flight is a channel. It gets provision just in time because the crew is different for every flight but all the apps that need to be brought together for every flight are all the same and that level of dynamic construction of a group – but not just a group to communicate but a group that gets provision with all the things you interact with.
BM: But all the workers of the manufacturing plant, and a lot with healthcare so doctors and nurses in an ER, all these people need to communicate and there’s other traditional information workers that need to communicate with them.
SN: So that’s a bit of what’s happening, this first line work is different, this notion of really doing vertical workflows is very different, but that transitions into the next piece which is around Dynamics. The thing that is happening in Biz Apps is pretty unprecedented. It kind of comes naturally, as you digitize the world, that means there is the need for more business process, not less business process. So we have made some big, big changes in even just the packaging, the pricing, the modularity of Dynamics, it’s unlike anything that we did. This notion of I’m going to buy a big ERP or a big CRM, those days are gone, I mean people are just going to essentially buy a module, integrate it with the rest of what they have, and so we restructured the entire business and the product around Dynamics and 365.
That by itself is disruptive, but the big shift is if you look at the history of business applications, there were systems of record, things like ERP, then the systems of engagement like CRM or even on the supply side they came about, and now what are seeing is two other new systems. I would say systems of observation, what I mean by that, take an IoT project, an IoT project is essentially taking a thing and then being able to observe it in the real world continuously and that digital trail of that thing, whether it’s a refrigerator or a factory or anything is all schematized in its own way, it can even be a simple log format, but it’s all there. And it can be related to the actual BOM, so literally I can have the system of record where the Bill of Materials is there, the system of engagement where the customer who bought it and their information is there, the system of observation which is the actual product in the house of the customer and it’s digital trail. Once you have these three things, you now can say, oh let me do a system of intelligence which really triangulates across these three and adds new value. One great scenario is in Field Service, it’s pretty stunning to think about how field service has become the most important, interesting thing, which is because now anybody who’s got this piece of equipment knows how to make that digital trail, do the analytics, do predictive maintenance, now the fact that something bad is going to happen doesn’t stop you, it’s the start of a truck roll in many cases so you can get the person who can fix it before it breaks and that’s a place where you have to coordinate, who is the right person to show up at the right place to fix the thing. That entire coordination is a great example of this phenomena and James Phillips, founder of Couchbase, I forget now James, how long you started at Microsoft?
James Phillips: 7 years ago
SN: I can’t believe you’re still here . James leads all of our Dynamics business, and maybe you want to say a little bit about what you’re seeing?
JP: Sure, yeah so just extending from what you’ve said. Business applications are what embody the business processes of our customers, it’s the way you automate your business and I think if you look back over really all time, business applications have been unchanged. You go back to 1978 and you look at the call center, just to extend from Satya’s example where you’ve got a customer service organization taking calls from customers who purchased something from you and it’s not working, they need support, the business application that was used in that environment was, call center representative opens a screen, takes notes about the problem, schedules a field service worker to go out and make a repair whether it was consumer or commercial. You look at that system in ’78, ’82, ’93, 2012, and you’re literally looking at the same screen. One was running in green the other was in a Windows client the other one was in a browser but it’s the same thing. For the first time, in the history of our industry and the results are for the history of all industries, there’s a fundamental change occurring. That fundamental change is that data is now coming first. That sensor in the piece of equipment that you sold to a consumer or in the factory is sending a continuous stream of information about how it’s being used, about it’s health, that data lands in the cloud, an anomaly detection model can predict it’s going to break, not that it has broken, but that it’s on its way to trouble, that can proactively result in a field service call and keep the customer up and running. So the systems that you need, the business applications that one needs to enable that very proactive versus extremely reactive model of business are fundamentally different. You don’t start from a screen somebody types into, you start from data arriving. That is a complete flip which sort of turns it all on its head and as you think about what we’re doing, just to tie it up with what Brian said, in a world where things are proactive versus reactive, you need a channel into which you can notify the team that something’s up and that they need to respond. So all these things really come together and if you think about what we’ve built at the company it really starts at the ground from this infrastructure that can collect this amount of signal coming from everything that we can then reason over and then deliver transformative experiences, so it really composes incredibly well.
SN: And just to complete that picture, we talked a little bit about this notion of building for the user and all of their devices and all of their artifacts with Microsoft 365, we talked a little about this transformation of business applications to really accommodate way beyond systems of records, systems of engagement with the systems of observation and intelligence. The thing that still is unsolved in all of this is there is lots more new code that needs to be written to make all of these things come together. We have attempted this in the client-server era, achieved actually quite a bit of success, and it’s always been my dream of saying “Ok, completely SaaS world, completely cloud world, how are we able to change that curve of being able to allow increased digitization?”. Because you can’t say this is going to have the same expense burden as was there in the client-server, so one of the things that is probably not well understood but at scale business at Microsoft already is what we call the Power Platform. For the first time at least in the twenty-seven years I’ve been at Microsoft, even inside Microsoft, we never had an extensibility model that was consistent quite frankly across all of our own SaaS applications.
This Power Platform is the extensibility model for Microsoft 365, it’s the extensibility model for Dynamics 365, and by the way, it can integrate with Salesforce, it can integrate with Workday, it can integrate with SAP or what have you, so what is this thing? It’s got a workflow engine called Flow, it’s got an analytics engine called Power BI, and it’s got something called Power Apps, which is the ability to build apps. Sometimes maybe you’ve heard – you’ve ask the VC’s, they’ll say “hey, the hottest category now is robotic process automation” – that’s what this is. Robotic process automation brought together as a suite of applications so that any business can start building, literally a knowledge worker or an Excel analyst type person can start building an app, a workflow, where none exists, and that level of I would say automation would be very [unclear]. If you want to change the productivity curve, or you wanted to drive the advantage home of all this digital technology deployment without something like this without something like this it would be hard, because you can’t just rely on the apps that are built which definitely don’t accommodate for the variability in business process. So it’s these three things: Microsoft 365, Dynamics 365, and Power platform on top of what we’re doing with Azure is the core of what we are doing as a company viz a viz I would say our commercial customers, businesses of all sizes, whether it’s small business, large business, whether it’s in an emerging market or in a developed market.
We are in, quite frankly, the early days in all of this, but good momentum and we do have ambitions outside of this which I don’t think we covered particularly today with you all. Some things that I’m really invested in is gaming, and the simple idea there is we have as much of a shot to build a subscription business as anybody else. So we describe it as shorthand, “Netflix for games”, we have a structural position in that we have both a console business as well as a PC business which happens to be in fact bigger than the console business when it comes to gaming and the idea is to aggregate those sockets with a subscription service, we won’t be the only ones, there will be competition just like with other content, there may be a few subscriptions that will be successful, so we are going to go after it. The good news is, we have a huge back catalog, which is we have our own games, we bring not only users, we bring already a social network in Xbox Live, we bring content, and we’re going to go after it. We’re going to go after it with current sockets, and of course streaming is another element to it, because then can you plug in an Xbox controller into even an Android phone and play a Triple A game? We know that’s feasible, now the question is when does the cost curve on it going to be viable. So those are at a high level what we are trying to get done.
Let me just stop there and turn it over to questions, comments.
Question: So you focus a lot on enterprise and business, I think one of the areas that’s harder maybe for me to understand, is your aspirations in consumer, if you put gaming aside. So you’re talking about this pervasive computing, what is Microsoft’s business in consumer looking a few years out? It seems like it’s hard to separate consumer and business uses of technology like Apple is a device, the iPhone, resonates with consumers but yet for most people it’s actually a business device as well.
SN: One of the things on the consumer category, it goes back to a lot of people say “hey, you used to be a consumer company and now you’re an enterprise company” the reality is what is considered consumer today is very broad and what is considered consumer tech is very broad. Uber is a consumer tech company, Amazon (the retail side) is a consumer tech company, and so 1) we have to be careful that just because we’re a tech company doesn’t mean we get to participate in every consumer category – what is it that we uniquely do? One thing that I feel like at Microsoft we made a bunch of mistakes by just saying let’s just enter every category, just because we’re a software company. Sometimes, it’s sensible to do so because there are big TAMs [total addressable markets], but at the same time, if you don’t have some unique thing to contribute, you usually fail. That said there are two, two and a half, three things, which are purely consumer. One, you brought up Xbox, that’s purely consumer, although by the way, in all of the businesses that we are, there is a nexus between our consumer side and our enterprise side. For example, now Azure is becoming a fairly big destination for a lot of game development houses because we are taking in fact all of the higher level services that are part of Xbox and making them available as PaaS services on Azure, so that’s one other side of how we are connecting the dots.
The other one – when I think of Microsoft 365, I think about Microsoft 365 as a two-sided market. It’s very fair to say as I said in the very beginning that we started after all on the consumer side and then over-indexed to the IT side, and we definitely are very focused on bringing that back. A lot of what we are trying to do with our devices, Surface is doing well because ultimately people do have a special connection with the device they use. It needs a brand, it needs that last fit and finish, the design ethos, therefore we are committed to our Surface business in the categories we can compete in whether it’s the large screens, the laptop form factor, the 2-in-1, I don’t think this session we are doing a lot more on hardware, you’ll see HoloLens, but we will be in the hunt even of what’s the next big turn in the device form factor which is mass market, even if it starts niche, we will go after those. So I would say Surface is a brand, what we are doing with Office 365 or what we will soon be talking about as Microsoft 365 consumer subscriptions, those would be again completely consumer businesses. And don’t forget, we actually have a big ad business, the ad business that we have which is in the context of both News and Bing, one of the things that we want to do is to say “What’s a way to change the game?”. We’ve tried in the past a couple of different things with rewards and economics because I fundamentally believe that – definitely the guys, the big guys are way ahead of us – that we have all kinds of strategic flexibility to do things that would perhaps be more amenable to both the end users as well as the advertisers and publishers so we expect to do things even in that space. So we have this collection of businesses, we have significant strength in our commercial side, but we have a sizable footprint on our consumer side where we will double down on those but be very mindful of categories we enter where we can do something unique. Like a good one is speakers, to me – one of the biggest challenges was exactly what is it that we can do in that category that’s going to be unique? Would we be better off for example to make Cortana a valuable skill that somebody who is using Alexa can call? Or should we try and compete with Alexa? And we decided we would do the former which is because Cortana needs to be that skill for anybody who’s a Microsoft 365 subscriber. You should be able to use it on Google Assistant, you should be able to use it on Alexa, just like how you use our apps on Android and iOS so that’s at least how we want to think about where it’ll go.
Question: Satya, you mentioned economic nationalism and if the era of globalization is over, can you just play that out a little bit more, looking at kind of a bipolar world with US and China or maybe also India? What will it mean for Microsoft and others do you think?
SN: Yeah, I obviously think about this perhaps a little narrowly, because it’s such a complicated topic and people, you should opine on what you see, but here’s how I think about it, simply put for a multinational company we have to accept that unless you are really added an economic service in every country you are operating in, truly, that is measured by employment, measured by taxes you pay, measured by essentially all of the prosperity around your activity that gets created locally, if that is not true, I just don’t see how the world just kind of says “let’s just go back to this thing”. I mean basically, look, globalization was fantastic except that it hollowed out the middle in many parts, not just in the United States, but all over the world and given that, I think everybody is going to be looking to say, “Ok how do I get back?” and the equation between innovation and democracy or whatever form of government they have and economic prosperity that’s broadspread because that’s the only way to stay in power and so given that we as a business community better be sure we are in harmony with those goals.
Question: By your framing, Apple should be in great shape for example, and Microsoft through no fault of its own, won’t be.
SN: Actually, interestingly enough, given that our footprint in China is actually pretty significant, even though we don’t get paid for it. Quite frankly, I think that China is a great example of places where basically you’ve got to remember Apple’s done superbly in China, it’s the only multinational company that has done well in China, probably because they build a product that the Chinese customer will pay with the business model, interestingly enough they got the market fit in China and also the fact that they employ a lot of people in China, so that is probably to their advantage, whether there is some issue in the short term, that’s something that could be short term, but for me at least I do believe that – you don’t have to go to China, go to Europe, let’s start in Europe, let’s start in the UK, what’s happening in UK? What’s the tech sector in UK? How is innovation, how is employment around the tech community improving in that country? Those are all top of mind considerations everywhere.
Question: What does it mean practically for Microsoft?
SN: It means structuring simply a business model that allows you to create for every dollar that you make multiple orders of dollars beyond that in our channel broadly speaking. What are the local startups? What are the localized feeds? Who are the local SI’s? What’s the total employment digital skills inside of the companies?
The great opportunity is if the tech sector is not just about the tech companies producing some gadgets or even cloud services, British Petroleum is a tech company because they are going to do a lot of stuff going forward that is digitally driven, they’re going to employ, believe me, a lot of AI. In fact if you look at the competition for AI talent, it’s no longer just in the tech community or even with the financial services, it’s with everybody. The Github acquisition to me was fundamentally driven by that, when I sort of said “what’s the next big community of users in every business?” it’s going to be developers. And in fact, the LinkedIn data was what showed us that the number of developers outside of the tech sector is growing faster than in the tech sector. This is developers, software engineers, being hired by anybody that is not in the tech sector. They are the biggest growth market, so you then extrapolate that. So it seems like it’s impossible to get into a computer science class as a computer science major in 2019, everybody wants to be a software developer. So that’s the supply problem which will get solved and then it’s all [unclear].
Question: Can you provide a concrete example of how your behavior has changed going into some new market adapting to economic realities?
SN: Quite frankly, if anything, it’s not the change in behavior, it’s sort of like really reinforcing the behavior we’ve always had. What I mean by that is take the service providers we’ve had or the cloud solutions specialists we’ve had, all of that by making sure you are investing in building that footprint out so that platform sensibility whether we added with MS365 or with Azure, we added with the client-server era it’s becoming more important for us to invest in it, that’s kind of it. And it shows up in practical terms, like even the investments we make in our training, of people whether it’s in the company or in the channel, it’s pretty significant.
Ben Thompson: I have a follow up on that. You’re referencing the idea that Microsoft has had where Microsoft is going to take a percentage of the ecosystem’s total value and it’s going to be less than 50%, I think the goal is always like between 25 and 30%, but it seems like there is a marked difference in that previously that other 75% was other tech companies mostly, or consultancies or ISV’s or whatever it might be. Whereas now, so many of the things you’ve presented and talked about, whether it be Teams, Dynamics, it’s a much more Microsoft doing the entire job, much more of integrating all the technical pieces and bringing it all into Azure or whatever it might be and now you are using the same language as before, but all that extra money in the ecosystem is actually not tech companies, it’s retail stores or airlines, or something along those lines. It seems to me that it’s the same language, but who benefits is different.
SN: Of course all of the others who are not traditional partners have to make the transition to this new world, just like we have had to do. It’s a hard medicine, you’ve got to take that, which is take a reseller of Office in the past versus a CSP for Microsoft 365 in today’s world, change management of trying to get Teams deployed inside of a company with all of the business applications integrated into that scaffolding and so on is a job that Microsoft’s not doing – some local partner is doing. So that exists, even, so in fact I would say there is an expansion, new skills, but an expansion of what we traditionally thought of as the integration business. It’s not just the global SI’s it’s the local service providers, a lot of emphasis is given to global SI’s [system integrators], but actually there are a mound of local service providers in the world I just described is huge. Local ISV’s, for all of my love of business applications and what we are going to do, the one thing is, business apps is the most fragmented business by country, by vertical, by size of business. In fact, high share in these apps would be 20% for the largest businesses in one category. So you go to Turkey, do you want to sell to a small business in manufacturing, you’re going to get a local ERP system, these are all old guys we participated in the past with SQL servers, guess what? They’re all moving to the cloud. That’s how I see even the tech sector, but you are absolutely right, it’s also about whether it’s Kroger, whether it’s BP, whether it’s Walgreens, whether it’s Walmart, they themselves are building digital ecosystems of their own.
Question: So you described, you said, “the era of globalization is over”. And then you’re describing fairly pervasive data collection, many of the businesses you’re describing will require a massive ingestion of data and continuous monitoring so that is customer data, they’re handing it over so that is consumer data – those regimes around the world are changing, how are you managing this?
SN: It’s a complex world and it’s only going to get more complicated. The fundamental posture at least that we have taken is for three sets of, and I think Brad is going to come and he’ll talk more about it, but the fundamental posture we have taken whether it’s on cybersecurity, whether it’s on privacy, or on AI ethics, just assume that these three things from our perspective we better take them as core responsibilities as say we did with security. We got a massive wake up call, you know, 15 years ago, we need to really ensure the products we build are trustworthy, and have to change everything from our culture, to our tool chain, and how we build software products, and we realize that there was going to be a continuous journey. It’s not like ok, let’s do a stand down for three months and we’re done. But we have to re-engineer ourselves to start building – we still have security audits. So that is exactly what we now believe we are going to do across all of these areas, and we think the regulatory environment around the world is going to get fairly complex and Brad can talk about it. In fact, in some of the places where we are asking, in fact, take this notion of facial recognition, I think that right now it’s just terrible. Right now it’s just absolutely a race to the bottom, and it’s all because in the name of competition whoever wins a deal can do anything – that’s not going to end well for us as an industry or us as a society. Therefore it’s better for us to have some modicum of rules by which all of us play so that if we protect what actually matters most.
Question: I’ve asked Brad this before, but I’m asking this to you too. You are now in a world where you have an ongoing relationship with customers, not by yourself walking away, you have an ability to control what those customers can do, you can take data off of Azure, law enforcement can come and look into Azure. Who won’t you sell to? What use case won’t you enable? Robotic process automation is great, but if you’re going to sell that to some health care vendor, a bunch of people will go out of work. Does that ring your alarm bell? Or is it, we’re not going to do facial recognition for drones.
SN: Well I think the criteria for us for who will we not sell to I think has got multiple layers. There are certain people we can not sell to, because there are laws against it. To the extent to which in first place where we invest quite a bit, which by the way has become quite complicated, is compliance on existing “do not sell” lists and so that’s kind of the first place to start. Then, I think it becomes a question of where are we passing editorial judgment, it’s a pretty scary line. Where I want us to be is what are the principles we have? For example, the platform versus our own people, let’s say consulting, are two very distinct things. Where we are directly involved in some consulting project, where we have a lot of data on what the use case is, we will, in fact in many cases we have walked away from things are not conducive to our principles. If there was a project that came in and said “hey we’re going to do weaponized AI that is not going to be in control of some democratic process” then we are not going to engage in that. Brad can go into that, but there are regimes that come to us and say “hey, give us your latest technology”, and if we don’t have the assurance that that’s a regime has the democratic principles of how that technology is controlled, we won’t sell to that. But at the same time, arbitrarily trying to censor the use of technology platforms, we want to be super careful. So it’s ultimately going to come down to have a set of principles and relying hopefully on especially liberal democracies to come up with the rules, even if those rules are different by country. It’s OK, we can wish sometimes for an equilibrium, but we do believe even in world where it’s not as globalized as we had, maybe the global norms around some of these are very important. I was actually, on that context, I was very pleased with the reaction in China to what happened with that CRISPr gene editing piece was one of the most pieces to watch. The fact that even in China people were worried about what that means is probably the positive sign in all of this whole thing for us.