 OK, so let's start. My name is Andreas Constantino. I run a company called VisionMobile. And VisionMobile is a research company. It says mobile, but we actually do far more than mobile. We started by doing developer research in mobile developers six years ago. And today we survey all kinds of developers, including cloud, IoT, desktop, machine learning, and far more. And here I'm going to talk to you about some of the results, our latest results, from our developer research of pass and ES solutions. And it's probably the only talk in this conference that's not talking about Cloud Foundry. But it talks about lots of other solutions in that space. I've also got some of our reports here. I can hand them out to anyone who's interested at the end of the talk. All right, so firstly, can I ask who here is a systems developer who builds Cloud Foundry, who builds with Cloud Foundry? OK, like five people. And who builds on top of Cloud Foundry, like user applications? Maybe half and half, maybe a few more. Everybody else is like product manager, maybe. OK, so a mix of people. In terms of what we do, we basically help the world understand developers. We help the likes of Microsoft, Intel, Google, Amazon, and others understand developers, so who developers are, what tools they use, and where they're going next. And it's basically a two-sided business model. So we connect developers with the people who sell to developers, all the tech companies out there. We run two very large surveys, 30,000 developers annually that we survey. And we survey on anything from languages and frameworks, developer experience, and what features they expect to see next from platforms, and what augmented reality frameworks that might be using machine learning frameworks, and so on. And we offer that as competitive intelligence and to developers, we give back a lot of the research in reports like these. You might have seen actually this one, which is nice graphics on the front page. I would just call the state of the developer nation. This has been going for six years. And we're launching the 12th edition of the survey in a month. So if you have any ideas or suggestions like you should definitely figure out what developers say about X, then let me know, because I can pass the feedback to the analyst team. Now, pass and use developer experience. This is the first time I'm showing the data publicly. This comes from our 11th Global Developer Economic Survey, which reached about 16,000 developers. To understand how this thing, how the data was pulled so you can get some context, this is basically how the survey works. So depending on what type of developer you are, you get asked what type of solutions are you using and then if you're using a pass solution, which of these solutions, which of these pass solutions are you using and so on and so forth. And then we measure all of these things in the gray box. Now, in practice, what we measure for pass and use is between 1,300 and 1,800 developers, pass and use, respectively, across these solutions. Some of these are ES, some of them are pass. I'll show you the different charts. And we measure the developer experience across all of these characteristics, product characteristics. So reasons developers choose a product over another. So this should be a familiar format. The AXIS here differ. And of course, you'll see eight vendors here. This is ES, Infrastructure as a Service. Now, this AXIS is satisfaction. And how satisfaction is produced before I walk you through the chart is we take all of these criteria and we measure which of these are important to developers and how they scored each solution, how they scored AWS, how they scored digital ocean and so on. And we produce that in a composite score, which is a satisfaction index. So the more to your right, the more developers are satisfied with the solution. To the top, the more popular the solution is. So this is Infrastructure as a Service, start with. So Amazon, we found, is a clear leader. This is no surprise, of course. Amazon was the first one to actually coin the Infrastructure as a Service base. Came out 10 years ago. Azure followed in 2008 Google with its pass in 2010. And the top ES vendors up there, so top right quadrant and bottom right quadrant, they more or less appear in the order in which they launched. So maturity comes with age. Lots of bits about these vendors and why they're ranked and why developers choose one over another. So firstly, top right is those who have very high satisfaction, very high use or popularity. Bottom right, we call the fighters. Those with a fighting chance to get to the top. So they have very high satisfaction, but not so high mind share or popularity. And these are the ones that are the least used and the least loved of the solutions. I'll show you a similar chart for pass in a bit. Now, one of the main issues is awareness. I think it's the same issue that Cloud Foundry has. But a lot of the vendors on this side have very low awareness towards developers or by developers. Should be fairly intuitive. And it's the reason they're not being used that often. In terms of what makes a developer choose one solution over another, three reasons. Skillability, pricing, availability. They're different for pass, but for ES, these three are very clear. Skillability, pricing, and availability. Pricing especially has some very extremes, which I'll talk about. So Amazon scores very well on all three. Skillability, pricing, and availability. And features it. Now, especially for availability, we find that Amazon has 92% of developers rating it at four stars or more for availability. And digital lotion is the low end disruptor of sorts because it scores very high for developers choosing it for pricing, but very few developers choosing it because it has a very extensive feature set. In terms of net promoter scores, are you familiar with net promoter scores? So how many people recommend a solution versus how many people do not recommend a solution? It's a standard metric for satisfaction. AWS Digital Lotion have the highest NPS scores. And Rackspace and IBM in our research have the lowest NPS scores. Now, this is one metric. It's not the same for everyone, every developer, depending on the industry, depending on the architecture and so on, we'll have different reasons for choosing that vendor. Let's see. Lots of other data I can walk you through, but I'm trying to pick out the most interesting ones. There is a lot of correlation between languages spoken by developers and the stack, or the EES stack they end up choosing. So Java backend developers are 70% likely to be using AWS. And only 17% of Java developers choose Azure. Instead, Azure is much more about C-sharp developers, as you can imagine. And C, C++ developers have significantly above average preference for Google Cloud. Now, moving on to pass. Different chart or same quadrant, but different position for the dots. Here, AWS is not in a leading position. It wasn't the first launch. So firstly, you see how the dots are placed, but then you cannot talk about developers using just one solution or another. They're usually using multiple solutions. So if you're a Red Hat user, you're likely to be using on average 2.2 solutions versus AWS where you use 1.7 solutions, which means that Red Hat is mostly used more often than not used as a gap filler, whereas AWS is mostly more often than not used as the only pass solution. Now, the reasons for selecting pass solution are different to ES. Top three are ease and speed of development, pricing, and feature set. Reliability and scalability don't factor into pass so much. If you're interested in documentation, if you care about documentation, you're more likely to be selecting Google Cloud Engine, or sorry, Google App Engine, and Google App Engine is recommended more than any other vendor for documentation and community. Now moving to Heroku, it's adopted for its ease of speed and development, the adoption reason for over half of Heroku developers, and just over half of Heroku developers who prioritize or adopt Heroku because of price. The same solution scores badly in integration and use of standard platform. And finally, in terms of languages, there's a lot, again, of correlation or matching between language stacks and cloud stacks. So C-Sharp is the language of choice for Azure developers. If you are working on Google App Engine, you're much more likely to be prioritizing PHP and Python than any other language. And Amazon Web Services is the most multilingual of all the stacks, and developers who are using Amazon Web Services are most likely to be speaking Ruby, JavaScript, and Java. So that's the overview of the solutions. And I wanna zoom out and give you a bit of a broader picture in software development. So we looked at cloud development, but there's, of course, mobile desktop and Internet of Things development. Internet of Things is, of course, the newest. Now, mobile apps have been probably the most hyped, but we shouldn't forget desktop legacy. So what the chart shows is out of the entire developer population, actually the entire professional developer population, how many are developing mobile apps versus desktop apps versus internet of things, apps or data or hardware, and cloud and backend services. So cloud is number two in terms of popularity overall by developers. And this data comes from two surveys ago, and we are updating it with every survey so twice a year. Another very interesting observation about what developers are working on is there's no such thing as a mobile developer, a cloud developer, a desktop developer, an IoT developer because they're sort of multi-homing. So only seven, and this is from our ninth wave, so three waves ago, so things have evolved even further since then, 17% do only mobile apps, 4% do only cloud or backend, 4% do only Internet of Things. So whatever application you're building, if you're building a mobile application, you likely have some storage or some session data back in the cloud. If you're building an IoT, you need to communicate with some APIs in the cloud. If you're building a desktop app, for sure there's some APIs that you wanna integrate. So there's no such thing like I'm on developing on this, form factor of this application or I'm just a mobile developer or something else, everyone I think will have to live in a multi-screen world from a development perspective, and this is a clear trend. And I have two more slides showing data on mobile and IoT. So mobile, is there any mobile developers here? Maybe a few shy hands. So this shouldn't be a surprise, Android dominates every country, every region. The breakdown is dark, green is professional developers, light green is hobbies developers, black is professional iOS developers, gray is hobbyist iOS developers. The interesting thing is that Android now surpasses iOS in terms of professional developers prioritizing the platform, which is a change to what we had so far, because so far Android was very popular with the hobbyists, overall it had more mindset, more popularity, but it wasn't preferred by professional developers. But now we're seeing the shift as well. I mean the market share, if you're observing market shares, device market shares now Android is on 85% plus of devices sold, it's an incredible market share. So professional developers are also following that. And last but not least, IoT. These are the different verticals IoT developers are working on. These are snapshots every six months, and this is from our latest 11th wave. Smart home is by far, by far the most popular vertical. And this should be no surprise, of course, because most of the platforms, developing platforms are for smart home. Whether you look at Kickstarter or big tech name platforms, and you also see retail in decline. I think there was a lot of buzz in retail, but it's probably better suited to agencies or system integrators or larger development shops. Industrial IoT is in decline slightly. Connected cars slightly in decline. And also notice that not committed to any verticals is very important because it shows now that the IoT landscape is clearing up. So if you're developing for IoT, you'll have a conscious choice. For a smart watch, am I building for a home router? Am I building for a data application in the cloud for health, for example? So developers are making much more conscious choices. This is the data you can find more data on. So I have this report, I have a copy here, and I have a copy for another report. There's lots of stuff and research you can download for free on visionmobile.com slash reports. And then under services, you'll find more about the data that I talked about earlier about pass and ES. And you can, this is a service whereby you can access data about why developers are making choices to adopt a particular pass or ES solution if you're a product manager. And my contact details. And I can take any questions, if there are any. Oh, semantically, we consider, so if you are deploying a pure server-side application, then you would call it a cloud application. If you're a mobile developer and you have some stuff sitting in a data center, but you don't care where it is, like with a mobile backend as a service, then you would call it a backend application. We also called it backend and cloud because to make the distinction between front end and backend. So semantically, we grouped everyone who is developing applications, not sitting on a local machine, but sitting somewhere, you know, in an infrastructure. Yes, yeah. So this is like the stuff in the reports that we sell, but the developer information is segmented by B2B versus B2C, region, languages used, segments, you know, revenues and much more. It's part of our, what we call the control variables. Any other questions? Yeah. So we had an issue when we launched the survey because Cloud Foundry has very low awareness with application developers. And so we could not compare Cloud Foundry side by side because this is like measuring the developer experience. So if you don't have developers experiencing a solution, you'll get no developers to kind of tell you what they feel about it. In the next one, we'll probably include Pivotal because it's probably the purest or, I guess, closest to the base implementation solution out there. IBM was there, yes. So IBM also includes an implementation of Cloud Foundry among everything else they do. Yes. So, yeah, so we did, in past, in past the choice was for IBM Cloud Platform, it was not specifically Bluemix. So, I'm not sure exactly why you made that choice. I don't know why you made that choice. I can ask, but I cannot, like I don't have it off the top of my head. Here, in the past, we specifically included soft layer. In AWS, for example, in past we included Beanstalk, which is like a pure pass, whereas in ES they have a number of products. So we didn't name one because we wanted to be closer to how the developer experiences ES. So there's a trade-off in how you measure the developer experience against a specific product. If you have an input on how we should run it in the next wave, then let me know. It's starting in a month. Yeah, yeah. Is it like, is that vertical axis there, zero satisfaction, or is it some other number? No. How do you calculate this in the way? Yeah, so satisfaction is a composite. We take these product attributes. These are specifically for ES infrastructure as a service. We take these product attributes. We ask developers to tell us which are the most important reasons for selecting a solution. And from that we extract overall popularity of selection reasons. And from that then we say, if this is the weighting matrix, the set of factors that are used for weighting, then how do developers score each of the solutions weighted by these weighting factors? So it's a composite score. It doesn't attribute a particular importance to how important is scalability versus reliability. It's not like other analysts who come out and say, well, we think reliability has a 0.8 factor. We're not here to say what we think. We interpret what developers say or users of these services. So we figure out how important they said these criteria are, and then we multiply these weighting factors with how they score each vendor to come up with a composite satisfaction index. Like, what's the, like... The actual range. Is everything that you were showing me between zero and two, or is it zero and five, or what's the scale? I think this is, I'm guessing, so I wasn't the one that drew this. But I'm guessing it's like minimum and maximum over these two ranges. I don't know if we have, like, for example, I don't know if this minimum and maximum is the same as this minimum and maximum. I'm not sure. This one? I'm not here for this by region. We have, yes, we have the... So if you're interested in like the drill down, then come and talk to me after this. There is like 10 variables we have on each of these experiences, meaning like, so you can say developers who are based in Europe or who are based in, like North Europe or South Europe or whatever, how do they score these solutions? Or pro versus consumer, or those who speak JavaScript versus those who don't, and so on. Yeah. Especially the Eastern Europe, central Eastern. Yeah. No, there's no private clouds there. We had, if I remember correctly, the breakdown from between private and public is about 45% private and 55% public. So the biggest cloud is the private cloud. All right, because if you consider like private hosting as one choice, then there's more developers doing private hosting rather than hosting with any of these guys. Does that make sense? No. No, so we didn't, all of the experiences we benchmarked were on clouds essentially sitting on an operator, meaning that we also tested for reliability, scalability, so aspects of the developer experience that have to do with the deployment, not just the tooling or the language or the frameworks that you use. So if you ask anyone here what's their experience with their own private cloud, they'll probably, you probably have as many experiences as there are clouds, right? So it didn't make sense to measure private clouds unless you can compare the experience across multiple developers. I can explain more offline. I guess I'm understanding the question well enough. I differentiate somehow with the developers, so let's call it native developers in particular region or developers for, let's call it development, for example, for US company, but located in Eastern. No, no, so we don't look at the location of the employer. We only look at the location of the developer. We look at the company size and the team size. Because in that case, the developers that belong to, for example, for US company, they have some standards enforced from US. Yes, so good point. I'll point you to one data point that's relevant. Azure has the highest number of developers moving away because of a management decision. 25% for, pass in 30% for ES, VMware has the highest number of developers selecting it because of a management decision, because it's corporate policy. Anyone else? Okay, thank you. I have three reports if anyone wants them and if you wanna talk about the daytime here for another half hour.