The Six Five “On The Road” at Cloudera Evolve NYC. Hosts Daniel Newman and Patrick Moorhead are joined by Sudhir Menon, CPO of Cloudera. In this conversation, Sudhir talks about macro trends that play a significant role in Cloudera’s product roadmap.
You can watch the full video here:
You can listen to the session here:
Disclaimer: The Six Five Webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors and we do not ask that you treat us as such.
Daniel Newman: Hey everyone. Welcome back to another Six Five on the road here at Evolve NYC 2022, brought to you by Cloudera, Intel, and IBM. I’m Daniel Newman, principal analyst, The Futurum Research joined by my esteemed colleague, co-host Patrick Moorhead. Pat…
Patrick Moorhead: How you doing buddy?
Daniel Newman: It’s good to be here. It’s good to be back. It’s been a great day of conversations and we’ve got another great one lined up here in just a few.
Patrick Moorhead: No, it’s super. And it was pretty special doing the kickoff for the event. I learned a lot. My energy is high. I’ve had many cups of coffee and I am super duper excited to talk about products because you know that in the end I’m a product guy.
Daniel Newman: I do. Every once in a while I’ve been on the road with you and you’ve said something like back 20 years ago when I had a real job, by the way, I do the same thing with my kids. I’m like back 20 years ago when I dot, dot, dot in school or when I was learning to drive. Fortunately back 20 years ago, I was still a child, but we’re not going to-
Patrick Moorhead: It’s all good.
Daniel Newman: Age ourselves out here. We’re having fun. It’s been a great event. Data is the theme, but we’ve taken that in a lot of directions because I think what we’ve ultimately discovered here, and it pretty much all the events we attend, is that to get to those business outcomes that every company is trying to get to, it’s going to take an amalgamation, it’s going to take a really great data management strategy. It’s going to be coupled with having the right partners, the right applications, the right infrastructure, architecture, and who better to talk architecture with than the Chief Product Officer at Cloudera? Sudhir, welcome to the show. Excited to have you here. How you doing?
Sudhir Menon: I’m doing good. Thanks for having me.
Daniel Newman: So quick intro and specs for everybody out there.
Sudhir Menon: Sure. So I’m Sudhir Menon and I’m the Chief Product Officer for Cloudera. I’ve been with Cloudera for about three years. I used to run the platform business before that. I’ve been an entrepreneur. I’ve run my own startup in the past and work for companies like VMware and Pivotal and others in the past. I run R&D and support, which means that everything that goes right, I get credit for it and everything that goes wrong is my problem.
Patrick Moorhead: Now wait a second, that that’s changed since I was doing products 20 years ago.
Daniel Newman: Nice.
Patrick Moorhead: Actually it was only 10 years ago, but where typically if something went right, it was the office of the CTO or something like that that got credit. If it went wrong, it was your credit, but no. No, listen, we’re here to talk about… I think what’s on everybody’s mind relating to Cloudera is how do you ultimately make your decisions about your roadmap? And a lot of companies take a, “Hey, tech first. It’s shiny object, let’s go shove it in here.” Which by the way, all due respect for technology, but I’m sure you have a different methodology that’s driven by some trends. And maybe if we could start with some macro trends that are directly impacting decisions that you make on your roadmap.
Sudhir Menon: It’s a very good question. We, as Cloudera, we serve some of the largest enterprises, we actually, I would say have the luxury of learning from our customers, understanding real world problems, understanding what are the business outcomes that they’re trying to get to. And then as an open source company that actually had its start in open source, we still have the grassroots community. We work with developers, we look at trends in the data space and it’s safe to say that almost every improvement or any innovation that has happened in data over the last 10 years has been driven out of open source. So we get both perspectives and we try to amalgamate the two incoming up with the decisions that we make for the product.
And some of the trends that actually have shaped our thinking over the last couple of years, last two or three years, as hard as it has been with COVID, to actually get out there and talk to people, some of the things that have become real is that we are about 15 years into cloud being a thing. And I think 15 years, just your financial advisor, you’ve been with your financial advisor for one year, you don’t know what their performance is. Five years later there’s a track record 10 years later you know how much money you paid for what you got. So we are in that phase of cloud adoption. So people are starting to take a look at, I am in steady state, what does my cost profile look like? That is one trend. The other thing that we are seeing is initially it was all about who has the skills to get on the cloud. So the old school companies were like, “No, no, we’re not going to do it.” Some of the startups were like, “Hey, we’re going to go crush this.”
Now what has happened is that you are starting to see CIOs worry about technology lock in, right? Do I have the right skills? Am I actually locking myself up? So that’s another thing which is actually driving the conversation around hybrid. The third thing, which is actually more of the grassroots thing is that, look, we were the beneficiaries of the innovation that led to this whole big data ecosystem. All of those Apache projects that we worked on that became part of our platform, were base level innovations that would signal the start of an industry. 15 years later what we are finding is that innovations now actually have to satisfy multiple cohorts of people. If you come up with a new open source innovation, it actually has to work with things that are there in the ecosystem today in order for it to be picked up.
Similarly, things like enterprise readiness, backwards compatibility, these things are becoming very real for customers. And so the open source community actually has to innovate. And this was the theme of one of my keynotes recently, that open source itself has to become much more industrialized. So that’s one macro theme. Hybrid is another thing. And the third thing is cost. Because back when you were doing products, you remember we talk about user kernel crossings. Every time a piece of data goes into user mode, down into kernel mode, there is an… I don’t think they talk about that… things like that anymore.
Daniel Newman: Thank goodness.
Sudhir Menon: Here’s the deal though. When you use five products to actually build your application, you’re incurring the cost for all of those user kernel crossings. So the TCO argument becomes actually very real because now you’re paying by the minute. You’re paying by the second actually. So those are some of the trends that we see driving. So the other thing is this ecosystem it’s become very varied. So observability becomes a key part of how customers are able to keep their systems running. So those are some of the themes that actually drive our thinking both top down, bottom up and that’s how I make decisions.
Daniel Newman: I appreciate that. So maybe we just drill down on a one by one basis. I feel like you spent a little time on open source, you dug and then you got to the next two and you went quickly with hybrid and then cost and then you observability, and maybe Pat, we can just ping pong.
Sudhir Menon: Sure.
Daniel Newman: Because I’d love to… Watching the opening keynotes, I found out that my cloud sucks. Thank you, Patrick. But in all serious, I did see theme from speaker to speaker to speaker and hybrid was a theme. Cloudera seems to be doubling, tripling, quadrupling down on hybrid. Talk a bit about that theme as you’re seeing and how it’s driving your product.
Sudhir Menon: Three years ago when we talked about hybrid, people would look at us… They were bit polite, but people were like, “Well, these guys didn’t cut it in the cloud. So they want to talk about hybrid.” That’s no longer the case. Now what we knew at the time… And common sense ultimately wins out, but what we knew at the time was if you are running workloads on hundreds of parabytes of data and you’re burning up thousands of CPUs and you’re running them 24/7, it is only a matter of time before you figure out that the cost doesn’t work out. So that happened over the last three years as workloads matured. This is true of as you get more experience, you learn, you have the track record. So when that happened, people were like, “Look…” There is absolutely no argument that building new applications is faster and easier if you go to the cloud because who wants to wait three weeks to get hardware provision?
Who wants to wait six months to get through legal approval to actually buy software? The cloud simplifies all of that and there is no way that it’s ever going to go away. But once you build these applications and once you get them into production, what happens is you’re now successful. Let’s say you are running a data warehouse, you’re running some workloads that produce dashboards. Your dashboard is popular. More and more people actually ask for it. So now these are custom slices on that data, which means you’re burning CPU, you’re burning memory and networking and so on. So the cost starts to increase. If you’re running workloads that need to run 24/7, if there comes a point in time where it doesn’t make sense to run those workloads in the cloud, especially for organizations that can amortize that cost over with hardware that they have on-prem. So we are starting to see customers and there are 57%… There was a slide up there. I think 57% of customers have actually repatriated workloads back on-prem.
Now, the key to doing that is three things. You’re not going to build an application that uses certain APIs on-prem, that uses different APIs in the cloud. You need a common layer of security governance. You don’t want to have a security model here and then have a different security model with Amazon and a different one with Google. Those things cause a lot of friction-
Daniel Newman: Or ain’t compatible.
Sudhir Menon: and they’re completely incompatible. So you want to have security, you want to have lineage of where the data… and then you need a policy driven way of actually moving data and metadata and the workload itself. So this is where the whole open thing becomes important. If you are using products that are built on top of open APIs, that are built on top of open source where innovation actually guarantees interoperability, you have the ability to take it from on-prem to the cloud or cloud to another cloud or back on-prem.
And that is the end state. Now why do we talk about the end state? Because we have some of the largest customers on earth, the end state does matter to them. So that for us, it wasn’t very revolutionary talking about hybrid. We just getting to a point where two things are happening. First is the market’s understanding of hybrid is actually getting to be more mature as you pointed out. The other thing that’s happening is that our tech is finally getting to the point where everything that I talked about, being able to move data services, being able to actually have security governance and lineage that spans cloud, adding support for things like spot instances in different clouds, all of those actual… You need real software to make that happen, that’s all coming together, which is what makes hybrid real for our customers. And the other thing is, look, people are ultimately driven by personal motivations. So what happens is that initially there was a push, if I’m not in the cloud, my career isn’t going anywhere.
Daniel Newman: Exactly.
Sudhir Menon: So today what we are doing is we are giving them skills that. We are building cloud native products that they can move to the cloud so they look smarter when, and that makes them stickier. So that’s our whole hybrid thesis there.
Patrick Moorhead: It all kind of came together when entries and Horowitz came out on a 16z site talking about repatriating workloads. I almost fell over in my chair when I saw it because if there’s a company that has made hundreds and billions of dollars of profit off the public cloud as Sentries and Horowitz and their premise was exactly what you talked about, which was listen, in startup mode and in scale mode and for super seasonal it makes total sense. But if you’re not there, and quite frankly I don’t know any of Andrews and Horowitz’s investments that are mature… Well, maybe 10 years ago with something like Facebook, but it was pretty big. It was the first cloud native VC that came out and did this. And by the way, to me it was an inevitability anyways, like I discussed in my talk, but we didn’t do the industry any favors by stubbing our toe on OpenStack. So in our preface up here, you had used a term that I’ve never heard anybody from Cloudera make that was observability and [inaudible 00:14:17] our companies both track observability companies. How does that relate to Cloudera?
Sudhir Menon: So when I took over the Chief Product Officer realm, I looked for both continuity and change. You don’t want to be the guy who says, everything’s fine, I’m not going to change anything. At the same time, you don’t want to be the guy who disrupted everything and then caused the company to go under for another 12 months while we worked through that change. So one of the things that I saw fairly clearly was… two things, that if we actually had standards based observability in our platform, we would be able to tell our customers about how to more efficiently use our platform. So that’s one. The second thing which is slightly further out is we live in a world that actually where Cloudera is not the only player anymore. Best of breed includes Cloudera, but includes other products as well. So if a customer is going to actually have to work with an ecosystem of products, then building a product that actually gives you observability across the stack reduces the total cost of ownership.
My initial instinct was not to go and build the product. We actually had a product called Workload Manager. But then we looked at our options and we said anything that we build needs to work, needs to span on-prem and the cloud, needs to be available in PaaS and SaaS form factors, needs to be available and it needs to use open standards. So we felt we had a healthy start on that. So we think that that’s going to pay off big time. So everything that we are doing, all of these data services now exposed metrics that allow us to look at performance, one of the things that is becoming super clear to customers is look, when you buy your hardware and it’s accounted for and the IT guy no longer has any bills to pay on that one, performance is like, hey, if I’m not getting performance, it’s an annoyance.
I need better performance for my customers, but the costs already are already baked in. I pay it for the 100 machines or the 1,000 machines. But in the cloud it’s not like that. If the damn thing isn’t running fast enough, what that means is it will run longer, which means the meter is running right longer. You’re paying more for EC2. So performance equals cost in the public cloud and you need to have solid observability on performance. There’s another area where Cloudera is actually looking at performance, performance optimization for the right reasons, I want better performance, cost optimization because performance optimization allows me to control costs. And the third thing is something that we don’t have in the product. What we are starting to work through is ultimately the software industry, the cloud industry uses a lot of electricity and we can only ignore climate change for so long, we will drive electric cars and all that, but at the end of the day, we need to actually figure out how to reduce our footprint for the workloads that we are running.
So we are going from performance to cost… At some point the observability product will actually tell you how sustainable your workloads have been, are becoming and give you the opportunity to actually do something about it, telling you all the way up to how much carbon have you burnt up, how much carbon have you produced as a result of running these workloads over the last three weeks, two weeks, and so on. So that’s a big deal for us. We’re doing observability the right way.
Daniel Newman: Yeah, you absolutely tied a lot of things together. You even brought the cost structure into that. But it sounds to me like if we really want to submit it, you’ve got four or so really big factors that are playing apart, Sudhir, right now in driving your product strategy, which by the way is the culmination of what you’re doing with Iceberg, it’s the culmination of CDP One, it’s that we are going to have these driving forces being brought together. We all know the kind of data proliferation and exponential data growth, but we also know that heterogeneous environments are the way of the future.
Sudhir Menon: [inaudible 00:19:13]
Daniel Newman: And that means open, that means hybrid and companies especially in a market that’s going to be very likely recessionary, at the very least, were going to see a major slowdown, tech is going to be part of the solution though for this. So you got cost and you dribbled in some sustainability in there, which I think every company needs to be thinking about responsibly approaching sustainability and that’s not consistent. There’s a marketing sustainability and then there’s a doing which…
Sudhir Menon: I think we are focused on the doing side of it because this sustainability thing came about from a grassroots effort internally where we have folks who are highly committed to making the right choices when it comes to instant selection and so on, making our product more sustainable, but one of the things that we did hackathon recently internally and one of the things that came out was, hey, what can we do to educate and inform our customers on how they can be more sustainable using our product? And in many cases, actually using less of my product to be more sustainable. But my product needs is getting better and better as we go, but that ultimately gives people one more reason to use my product.
Patrick Moorhead: And I think whether people are on the sustainability bandwagon or not, when it comes to public companies, they are being held account from board of directors. And it’s funny, no matter what side I found the people, nobody is arguing with using less energy to do the same thing. I see that as efficiency. I think the rub gets into where maybe you went a little bit too far and there was a counterbalance. But I think any type of thing, we’ll find our static point here. Sudhir, I want to thank you for coming on the show. I feel like I have a lot better idea of why you make the decisions that you make and I’ve been accused of not being the best listener in the world, but I did hear a few future forward things that you’re looking at and whether it’s observability, whether it’s sustainability, different levels of openness, it’s really exciting to see.
I will tell you in the future, I am looking forward to CDP One and where that goes. I feel like that’s a force multiplier for you that nobody is counting into anything. But listen, I know you have a lot of… There’s a lot execution work that needs to be done in moving customers to CDP cloud. So I don’t want to get the horse before the cart here. But anyways, thank you so much for being in the show and we hope we’ll have you on again.
Sudhir Menon: Oh, thanks. Thanks for having me. It was a pleasure.
Patrick Moorhead: I appreciate that. So this is Pat Moorhead at Evolve 2022 here with my co-host Daniel Newman, Futurum Research, and Chief Product Officer of Cloudera, Sudhir. Thank you so much. We hope you like the show. Hit that subscribe button if you’d liked it. Give us feedback on Twitter. You know where we live. Have a great morning, afternoon, evening, wherever you are in the globe. Take care.
Daniel Newman is the Principal Analyst of Futurum Research and the CEO of Broadsuite Media Group. Living his life at the intersection of people and technology, Daniel works with the world’s largest technology brands exploring Digital Transformation and how it is influencing the enterprise. Read Full Bio