 From our studios, in the heart of Silicon Valley, Palo Alto, California, this is a CUBE Conversation. Hello everyone, welcome to this special CUBE Conversation. I'm John Furrier in the CUBE studios in Palo Alto, California. We're excited to have a great tech talk here with good friend, Howie Hsu, who's currently the vice president of machine learning and AI at Zscaler, formerly an entrepreneur which he sold his company at Zscaler before that, entrepreneur and resident in Greylock before that, VMware variety of other endeavors. Howie and I, we've known each other for a while. Great to have you come in and chat about- Great to be here. The Zoom, Zscaler, these are the new breed, modern era company, SaaS, business models, really interesting and it's something that we were talking about on email and over text, it's our topic. Yeah. Thanks for coming in. Great. So you've seen the waves at VMware, you saw the rapid growth there. And now you work for Zscaler which is experiencing rapid growth. You saw Zoom go public and I just interviewed Michael Dell. We were commenting about that on texts as well. He said, these big markets that have big total addressable dollars associated with them are right for disruption. You know, they used to have high barriers to entry in the old ways to look at it. But now with cloud and with SaaS, with data, there's different and innovation speeds. This has become a big deal. Talk about your view on this. Well, to me, you know, when Zoom and then Zscaler founded many years ago, no one believed that they would have become this big, right? When Zoom founded, there were plenty of the conference free software available out there. When Jay founded Zscaler, people thought, well, there was enough security companies, security solutions. Clearly, they defied conventional wisdom and then they just fought on. And they saw something that other people didn't see, which is, you know, precisely what you were talking about. The SaaS is so different, right? The business model, the innovation speed, the data-driven kind of the thing, it's so different. So a lot of people say, hey, you know, what's the difference between SaaS versus the conventional? Isn't that just moving that thing over to the cloud? You know, I actually used to think that way too, right? Isn't that just the virtual clients moving on to Amazon Cloud? You know, after living and breathing in SaaS company and then also observing that in the VC industry as well, you know, it's just totally different. They are nine different. Well, I want to get in this with you, because I think you can bring some good perspective onto these insights into the rocket success of, say, Zoom and Zscaler. But Zoom in particular, recent successful IPO among the recent classes past quarter, you know, Zoom, Lyft, Uber, Zoom is standing out. They're getting profitable. This is video conferencing. You know, in the old days, if someone said, hey, I want to compete with the video conferences, well, the barriers are entry too high, but they took a very innovative approach. Cloud, data, simplicity, and the big 800-pound gorilla was the web axis of the world, who was defined, divine for sharing slides, not so much pure video. They really innovated the focus and the speed of success, unprecedented in my opinion. I think this is a huge success of what the opportunities are for entrepreneurs. Yeah, I think on the surface, right, if you ask Eric, he would tell you that, look, you know, the WebEx was designed for sharing slides, and then the Zoom was designed from ground up for video sharing or the video conferencing. So it's very different. It requires different architecture. So that's very true, but I think that there is a more fundamental to that. The more fundamental for that is, there are a few things, right? One is, you know, the product, the life cycle is very different. Like, you know, how do you approach the customer, right? You know, the release cycle, right? The sort of the feedback loop, right? You know, much, much tighter feedback loop, much faster feedback loop between customer and you, you know, the release cost is much lower now, right? You know, as a SaaS, you know, product. So innovation is just, you know, accelerated because it's SaaS, because it's a true SaaS. And this is a unique thing that you said before. You know, SaaS isn't just lifting a on-premises workload and moving it to the cloud. It's a completely different mindset. Talk about this dynamic, because it forwards new kinds of risk-taking. You know, I was talking about before we came on camera. Share your insight on that. Well, you know, as kind of the traditional software, you know, you have release cycle, you have, you wanted to have, give a release date, right? You know, and then once the product is in customer hand, if you have a bug, if you have something, it's so costly to change it, right? But as a SaaS kind of the form factor, you can take a little bit more risk, right? You can even give, you know, that feature set to 10% of your audience, not the entire set of the audience. You can do those kinds of magic so that you can accelerate the innovation. And while as a stream wrap software, the traditional way, you know, you have one shot. If that software is not good, then you are toast, right? Yeah, so you can move quicker. You can push code. You don't have to have the on-premise dynamic. Yeah, the innovation and then risk-taking are kind of correlated, right? Relatively more risk, you know, the more you are willing to take risk, you know, relatively you can take more innovation. So that's the- Well, you and I were talking, one of the key things that you've been talking about publicly and amongst friends is innovation speed. Everyone wants the innovation fever. Hey, I got to innovate, you know, digital transformation, rah, rah. And easier said than done. Innovation speed is critical with cloud and SaaS. Why? What's the formula there for innovation speed? Well, one thing we discussed, the release cycle, right? You know, for a, not necessarily for Zoom and Zscaler, but you know, for SaaS in general, it's possible for you to have daily, weekly, monthly release, right? Traditional software, there's no way you can do that. But that's just the release cycle of that. The other thing is, you know, the you can actually take a risk and you can say, hey, I want to try to raise 1% of the customer and then see whether they are, how they are going to react to this. But in the traditional way, you have product manager debating for six months, six years on whether or how to do things right here. Let's not debate. Let's just see. Let's ship it. Right. Yes, ship it. Every Hopin always says, you know, if he's not embarrassed by your first shipment, then you're not doing it properly, which begs the question. I want to get your thoughts on this because you, you know, you've been using you, but VMworld, you saw how early that worked and their transforming cloud is now here unlike when they started the company. What is the right way to do it? And what's the wrong way to do? When you look at an entrepreneur or a friend who's trying to get some business off the ground, SaaS business, when you look at what they're doing, you can look at their, the mechanisms and how they're organizing their team, their code, what's, what jumps out at you as the wrong way and what's the right way? Well, the, I think the culture is really it, right? You know, the kind of the culture of incremental success and the fast, you know, iteration is the culture for SaaS company, right? For the traditional one, you cannot afford to do that because when you, once you make a small mistake, you are toast, right? So, so I think, you know, the culture difference, you really wanted to have faster iteration. And that also comes down to the team, the people, right? Yes. The people selection. Yes. If you, if you are kind of used to this waterfall thing, it's pretty hard to adapt to this kind of the SaaS world. And what's your advice that entrepreneurs reset? Because if you say speed is of the essence, resetting is probably something that's not hard to do then. One, one, one, I would say easy, but not easy way. Hey, these were pivot, but you know, resetting means, okay, stop, rebuild. I think, I think one, you know, way to think about it is actually looking at how to build enterprise software like the consumer sort of product way, right? If you think about Facebook or Google, the traditional Google, right? Of course, Google now has a price product. But the traditional sort of the Google, Facebook, kind of the product, it's more for consumers to consume. I mean, there are faster iterations. How often, you know, what's the criteria to release a product? Enterprise product is getting towards there, right? You know, you need that kind of the thing. So if you don't know how to do it, look at Facebook, Facebook, of course. Facebook need to pull the other way around and they need to care more about the privacy, care about more stability. So I think you are seeing the sort of the two sides of the world that the enterprise side and the consumer side, that they are learning from each other. Well, I want to get to the enterprise talk track in a second because I think you're going to get a lot of incidents. I want to stay on the SaaS, you know, cloud native or cloud specifically because that's where SaaS really shines when you really talk about cloud scale. Data, you're doing AI now and you and I have both talked about data many times. Yes. You know, I'm a data hardcore person. I love data. I think software and data. I wrote a blog post in 2007 that says data is the new developer kit. Maybe the more developer kit was used back then, but you're not now seeing where data is part of the developer's piece of their value creation. Highly addressable, available, usable, not stored in some silo on addressable low late, you know, high latency to get it. How important is the data for the SaaS piece? Because that's where to make these kind of changes you're talking about, you need the data. Data's giving you insights. This is something that's near and dear to your heart. Explain your vision of the role of data. Yeah, I think, you know, you touched upon it, right? If you want to make sense out of something, you need the data, right? And if it's not SaaS, I would go, you know, maybe a more extreme way, but it's not clear to me that data is even useful to you because, you know, the data may be, you know, for some large software company, they may have hundreds of thousands of customers out there, but the data is spread around. I mean, how are you going to train a model with, you know, all the data spread around, you know, thousands of hundreds of thousands of locations? So the real, the correct or the optimal way is actually the SaaS model. You actually have the data with you and then you kind of, you know, leverage the data. So I want to say, this is actually another benefit of the SaaS, why SaaS is going to change the world or either world is it owns the data for real, right? You know, the data may be, maybe not the private data, like, you know, but it's actually could be behavior data, right? How people are reacting to your features. Like, you know, from VMware days, we wanted to know how people, is people even using this feature? How often people use this feature? You know, people are always debating, hey, what's the maximum sort of the policy we need to give this and that? But in the SaaS world, no debate, just look at it. You mean, we always say, don't listen to what customers are wanting you to do, but watch how they do things so that you can sort of understand what product you want to develop, right? Here, you actually can really watch how customers using your product, don't listen to them. If you listen to them, you will give them a faster horse, you know, as we are using. But it was important about the data discussion because, you know, a security person would say, hey, if you put it into one spot, I can hack it. But it's not just people's names, it's other data, it's gesture data, it's usage data. So you're not talking about like, just sign in data. It's data. It could be the behavior, it could be second order data, right, you know, do people use my product? That's my data, right? You know, that's something I wanted to know. I'm not talking necessarily talking about peaking into people's email, no, you know, it's actually the thing surrounding it. It's looking for the good things in the data. All right, let's talk about the customer alignment and customer expectations. You know, customer, user experience is driven by customer's expectations usually, right? As expectations change. I think the Zoom thing jumped out at me, the Zoom IPO and their great success and we're a customer as well. Is that they really nailed the expectation of the user and Cloud certainly helped them get that speed. But this is a key thing. If you could just deliver a great experience for those customers, you can actually win big part of the market. Yeah, if you Google Eric, right, you know, Eric doesn't speak to media as much, but if you Google Eric, what's sort of the jump? Hopefully I can help you to bring him here too. But what's going to be obvious, you know, if you Google search Eric is, you know, he's sort of the notion of customer success is my success. If customer is happy, you know, I'm going to happy. So my happiness hinges on the customer's happiness. So that's kind of very important because only the SaaS model, you know, made that more natural, right? In the traditional model, like whether traditional on-prem or whatnot, you sort of celebrate when you have customers signing your PO and then you don't hear from, you know, the sales guy for three years, the sales guy may move on to another company because you don't know, right? But for the SaaS, you know, it doesn't stop when you sign a PO. You actually have to earn customers happiness every single day. Adoption is critical. Yeah, customer success is important. And then that's kind of the, so there's a huge alignment, interesting, very interesting alignment between customers, you know, happiness, customer success, customer adoption of your product, and your sort of the success, right? Because, you know, when I came to Zscaler, you know, one of our first meetings is about, okay, we had a lot of customer interest, they signed a PO, how to get them, you know, ramp up the actual, the first use, right? So that kind of conversation doesn't happen in the traditional software company. You sign a PO, if the customer doesn't use your product for another 18 months, which is actually quite normal, no one is going to jump up and say, this is crazy, right? You know, we're going to do that on our part two about the impact of the enterprise, but you made up a good point, everyone, I just close out our last talk point is the data driving the experience isn't like, you know, the old way, throw and get the PO and celebrate, you got to kind of keep that going. The enterprise is changing, and the enterprise has a tsunami of onboarding of new types of developers. In some cases, they grow, we just had Cisco inside here on theCUBE this morning, they're turning network guys into programmers from CL command line prompt dudes to gals to coders. You're seeing developers now enter the enterprise to build the apps. So there's now a digital transformation initiative for enterprises to be, I guess, SAS like, but it's hard. Yeah, I think that's, you know, this is part of the digital transformation. Every company, Fortune 500, Fortune 2000 company need to do it, right? So another key interesting part is, when they do this, you know, on this journey of digitalization, you cannot possibly build all the infrastructure yourself. You will have to consume public cloud, you know, sometimes private and hybrid cloud, and you're actually going to consume lots of the SAS, right? You know, whether Zoom, you know, the Zscaler or the PagerDuty, I mean, you are not going to build those things from scratch, but you want it to have a very good sort of the stack on top of it, and how you're going to take advantage of the SAS is a very interesting aspect. Well, in a part two of our chat, when we come back on our next discussion, I want to get into the enterprise, but to wrap up part one here, innovation speed, you know, leveraging data and the beautiful risk-taking and benefits of SAS, large-scale, fast, high-value target in developing an app or a venture. What is your advice to entrepreneurs out there and or someone who's doing a digital transformation where they want to leverage SAS? What's the playbook? What's the starting point? What's your advice? Well, there are a number of things, right? You know, one, you know, there are so many SAS companies out there, take advantage of them, right? You know, like in the old days, you have to hire email admins, you have to do this. Nowadays, you know, all the SAS, just, you know, that's your kind of the, you only need to worry about the business logic, right? You have some unique insight in the business and then just to have higher programmers to codify that and then the rest will magically happen because of the public cloud because of the SAS, right? So be very mindful about the new environment you are in, right? That's number one. The second thing I want to say is, you know, how do you look at AI technology, right? You know, that the old way is, you know, program something in a definitive way. I think there will be a limit for that. You know, it has taken, you know, the software industry a long way to where we are. But if you look at the next 20 years, I think a lot of the lift is going to be done by the AI side. But it's not going to be easy to be done. You have to think about your data strategy. Where are you going to have the massive sustainable, unique, ideally even label the data? If you don't have the label data, you have to have the strategy. How are you going to have some unique model on the data with the data you have? So the data strategy, right? So essentially how to take advantage of the cloud, how to take advantage of the data. And then on top of that, you are going to do something that's solving a met- Customer problem. Customer problem. A cute landing spot in the marketplace. A met need. In a big market. Well, in a big market, even if there is already a mature solution, I bet since those mature solutions were not developed from the cloud, native cloud era and the native AI era, you have plenty of opportunities. How you and I are on the same page on this, I've been saying it and I truly believe we are living in an entrepreneurial era where with your advice and when you just laid out the better mousetrap can take down a big market. And I'm hopeful that you will also disrupt the media business. Don't tell anyone we're selling them that top secret, silent running. How we, we're going to get part two. We're going to deepen the enterprise because the enterprise now has an opportunity in the first historic time in tech history to use tools and technologies to completely reset and re-architect for this kind of capability. So we'll hit that in part two. I'm super passionate about it too. How we shoot here inside the cube, friend of the cube, legend in the industry, great entrepreneur and technologist here, in cube conversation. I'm John Furrier. Thanks for watching.