Retail's New Reality ft. Scott Lux
"I am big believer in the company that never sleeps. So how do we constantly find ways to have algorithms running while employees may be sleeping or when the employees come up?"
CIO, Palo Alo Networks
CIO, Palo Alto Networks
General Partner, Jump Capital
Host & Head of Operating Platform, Jump Capital
Kathleen Osgood: Welcome to the Jump Off Point, an original podcast by Jump Capital. Today our host, Jason Felger and general partner Saurabh Sharma, sit down with Palo Alto Networks, CIO, Naveen Zutshi to discuss the digital native imperative.
Jason Felger: I'm thrilled to introduce you to our guest today, Naveen Zutshi. Naveen's experience span software development, IT infrastructure, and leading technology organizations from Fortune 500 companies to startups. Currently, Naveen is the Chief Information Officer of Palo Alto Networks. In this role, he is responsible for Palo Alto Networks' solutions to drive growth in SAS security while also building a world- class and global IT organization. Naveen is on the board of advisors for several fast growing technology companies, including Zoom and Rubrik, but he also advises several early stage startups. Welcome, Naveen.
Naveen Zutshi: Thank you.
Jason Felger: And of course my colleague Saurabh Sharma. Saurabh is a general partner at Jump and leads our investments in IT and data infrastructure and application software. He has extensive background in product and operations at high- growth startups and started his career as a high- performance computing researcher. Welcome, Saurabh.
Saurabh Sharma: Thank you, Jason. Glad to be here. And welcome, Naveen.
Naveen Zutshi: Thanks, Saurabh. Looking forward to it.
Jason Felger: Naveen, let's start with a broad perspective, but also get into your background a little bit and how you got to where you are today. So it seems the role of IT in the CIO has just really dramatically evolved. And presently, it's just as much about transforming the organization and not just the technology. Companies are increasingly digital, they're agile, they're software driven. So I'd love to have you take us through your background and your journey to your current position and your current mandate and how that's just evolved over time.
Naveen Zutshi: Yeah. Sure, Jason. Like other first- generation immigrants, I came from a country called India, Kashmir. Actually, from a state in Kashmir. A really small place. I called myself, when I was growing up, foreign language, which is this notion of you're a frog in a small well and you haven't seen the bigger world outside. And so it has been a huge learning in terms of awareness and outlook that has increased throughout both when I went to Bangalore for my studies in computer engineering and then when I came to US, first in Arkansas, in the middle of nowhere, to Walmart. And then later on for the last 20 years, I've been in Bay Area working for both tech companies like Cisco, as well as startups from an engineering standpoint. And then another retailer, GAP. And then now for the last five years, I have been the CIO at Palo Alto Networks. It's been a fantastic journey, one where I've been privileged to get a lot of opportunities as well. When it comes to the role of the CIO, it's an exciting time. And as CIO, your job is never the same each day. What I find especially interesting is you have to go broad and yet you have to deliver with depth. And then it's amazing to problem solve with your peers and you get to solve either small or big problems. And then if you think about... I talked to quite a few CIOs in the Valley, also in Midwest or in South, you see a big mind shift change in CIOs over the last several years as technology has taken a much more central role in a company's strategy. And forward- leaning CIOs are increasingly finding opportunities to say yes, rather than saying no. And that's a big shift that I see. They are not letting risks be a barrier to innovation. The other thing that I'm seeing is this notion of performing under the same level of intensity and speed of delivery that sales abroad teams are expected to deliver at. That's one of the things that we try to hold ourselves at very high levels at Palo Alto Networks, is that are we delivering at the same level of intensity as our peers are?
Saurabh Sharma: Let me just dive in under that a little bit later. We're taking a little bit of a macro view. We have heard this narrative, obviously for many years now, that software is eating the world. And more recently it's obviously transformed into almost every company is becoming a software company. Given that shift in mindset at the strategic executive level, it almost seems that there must be a paradigm shift in how the CIO seat is. So I'm curious how your role has evolved personally, between different organizations as you've seen, or what you're hearing from your counterparts of where... One of the works CIO does is actually the revenue driving engine. Was it just running the rails in some sense?
Naveen Zutshi: Being at Palo Alto Networks, we are a digital first company. So it's like an oxymoron to say what is the digital strategy for our company? We are a digital company. When I was at GAP before this, it was a different matter. We were working with merchants. We were working with creative people who were designing high fashion and the business was of building great brand and great set of products. But there was an acknowledgement and an urgency to building and expanding on the online footprint. There was an urgency tying omni-channel experience and there was an urgency to building innovation in terms of how customers consume the products, whether they are in the store, whether they are online, how inventory can be pulled together, how POS can be managed in a more virtual manner, how you can use WiFi as an example, to drive traffic patterns, and understand better about the customer. So all of those innovations came into being while I was there at GAP. So GAP ran through a lot of transformation, not unlike other retailers have gone through. And I would say my experience is it's not enough. If you look at boards today, I still feel the boards don't have enough technologists on the boards. There's definitely not enough CIOs on the boards, for sure. I would forget CIOs. For example, there are not enough CTOs on the board either. And there was also a time when there was this notion of a Chief Digital Officer. I never thought that that would work because now you're set up another leader off on the side, responsible for the digital front- end. All the resources are with the CIO or someone equal into a CIO. That doesn't work either. You have the wrong CIO and you need to replace that person, or you need to make the CTO the CIO. So you have seen that shift back to a front- end, more front- facing CIO. So you'll get fired as a CIO for not having a stable infrastructure, but you wouldn't win in the business without having a CIO who can drive business strategy with their peers.
Saurabh Sharma: That's a great way to put it. That makes a lot of sense.
Jason Felger: Naveen, the future of the workplace is something that we're all talking about, particularly right now as companies are starting to contemplate coming back. And you hear it again and again. Work is something that you do. It's not a place. That's become a little bit more of an actual acceptable terminology and acceptable reality for many of us. How do you think about are we at a permanent shift? Have we actually gone towards a holistically or predominantly remote workplace for employees? Do you think we're coming back? And if you can layer in your perspective of sitting in the Bay Area, working for a technology company. That's one set of expectations, but how does that play out in manufacturing and healthcare and other industries? And obviously, the technologies that we might need to be thinking about to balance this new area that we're in?
Naveen Zutshi: The inaudible answer is yes. Hybrid, especially, is here to stay. And my thinking has significantly evolved and my biases has shifted dramatically in face of data that shows that those biases were incorrect. For example, I was a big believer of having everyone together at work, coming every day, having great relationships with RND, as an example, in the same office, same floors. Worked very closely together and we firmly believed our speed of innovation was directly correlated to having this intimacy of connections in person, rather than anything else. And for the last 16 months now, that hasn't been there. For example, our tech company has probably hired thousands of people and they have never been to an office. They probably have never met any other person from our company in person. And yet base of innovation, speed of innovation, delivery, has not slowed down at all. Let me give you another example. For us, personally, we set up an India COE. I have done India COEs before. We would send people over, we would fly people over from there, we'll fly people from here. We have done none of that. It's been all virtual. And what has been interesting is let's say the engineering manager's in my team. They have had more time to train and bring up to speed the COE people because they have more flexibility and it's not a 9:00 to 5: 00 job anymore. And that has been a huge win, whereas if they had come to work every day, during regular hours and then we are asking them to work either evenings or in the early mornings, it's a huge burden on folks. So that flexibility that has come as a result of being digitally native, wherever you are, has actually delivered a much faster ramp for our folks in India. Humans want connection, so we still want connection. That's why I said hybrid is the place it's going to change to. But at the same time, increasingly whether it's tech related jobs, software related opportunities, even recruiting or HR or legal or finance, and you look at all these different departments, lot of the work can be done all remotely. And in fact, you get more focus time and you can be more productive. And we actually measured productivity quite extensively too. And there's little to no drop- off that we could see.
Saurabh Sharma: That's amazing. It's definitely a new world. We are all preparing ourselves both mentally and then with all the resources. One of the things is at this point, pretty much everyone is almost comfortable in using Zoom and Slack. That's the de facto standards. So maybe just thinking about the kind of technology uniform factors that will take shape to what the future workplace and other technologies that would be embedded into that.
Naveen Zutshi: Yeah. So first, let's think about security itself. The attack surface has significantly widened and there is a need for inaudible trust security that is pervasive across your end points, your cloud infrastructure, as well as your data centers, your campuses, or follow assets wherever they are. And we need to think about, do we have pervasive amount of visibility into that? And then can we take effective remediation actions in very fast manner? It's still scary out there in terms of what you can see from solar storm to even the static code analysis tools being hacked. And figuring out a way to have technology that is deployed and can find those issues faster and resolve them is quite important. Second is on the technology side, water cooler conversations like boarding. Those are the things that you miss. And either it's going to be, you come in for a few days a week and do that or you have technology that enables that more and more. It's still a area that's open for a lot of innovation. Also, managers traditionally were never trained to do a great job with hybrid teams. So when you are a fully distributed team, everyone is on level playing field. But when you are hybrid teams, the bad habits can creep in, where those who are digitally remote can be ignored in meetings or in sessions. And so how do you ensure that you continue to pay attention to your entire workforce and make them feel inclusive and make them feel equally productive? So whether those are e- learning capabilities, whether those are technologies that make it easier for hybrid teams to work and collaborate together, there is a seamless transition between what does online and offline.
Saurabh Sharma: Maybe, yeah. Go a little granular on the CIO's day to day ecosystem and talking about digital native imperative of every company. And I'm going to keep Palo Alto a little bit out of it, as you said. But just thinking about traditional company that is making those shifts. They're becoming cloud first now. They're moving to cloud, as a starting point. So maybe give a high level view of what cloud native or digital native stack for a new age company looks like or should look like.
Naveen Zutshi: Yeah. Let me start with the second question first, because the harder one is probably the more urgent one to answer. And that is how do you transform your IT organization or your technology organization to be cloud first. If you look at traditional infrastructure, and I have worked in infrastructure for a long time now, there is a massive evolution in people and the way they work and what they'd work on. Increasingly, there is little differentiation between a software engineer in app development and a software engineer in infrastructure teams. This notion of Site Reliability Engineering is taking on a bigger hold, not just in companies like ours, but even in other companies, they want to adopt this methodology because it tries a right level of SLRs or SLAs between business and IT. In terms of application development, really having this notion of whether it's two pizza boxes sized teams or teams that are agile, whether you use scrum based methodology or whatever the metrology is. Teams are moving towards iterated based approach development because the speed with which you have to drive change is high, and you don't want to build requirements for nine months in advance and then see all your business shift. That technology stack changes your whole pipeline of development changes. And then how you secure that pipeline becomes imperative as well. So you're no longer thinking about security in production only. You're thinking of security all the way through the development life cycle. And then how do you train your developers to be security minded because there is traditionally 100 developers to one security engineer? Then if you look at the technology stack itself, you are in a more microservices services based architecture, and that's not new. We ever had services based architecture since'90s and now inaudible. But having this encapsulation of the database with your application logic and having it in microservices makes a lot of sense, but at the same time, it brings with it tremendous amount of complexity of managing those microservices. Now, you have hundreds and thousands of microservices. They're connecting and talking to each other. They have security risks. They have operational risks in terms of availability. You don't know what your website is down. You don't know what caused the website down. There can be 100 reasons for it. And so there are new observability tools, there are new monitoring tools, there are new ways to manage your applications as a result of a distributed architecture. So it's an interesting and a very fundamental shift. And what I found, giving the right environment where you allow folks to make mistakes, they embrace this because they want to learn, they want to improve their capabilities on a daily basis. And there may be some resistance, but most of the time, it is enthusiasm. There is change. And we need to do a better job as leaders to manage all that change for our employees.
Saurabh Sharma: I can't help notice, but you mentioned a couple of times things that evolved around the developer. So it's almost like the traditional buys and traditional hierarchy always has been top- down, but it seems that there's a lot more decision driven by the faster developer productivity, developer deployment. And then how do you provide the stack that is basically a drive to quick adoption from the developer side, versus necessarily making a large decision at the enterprise level then moved down?
Naveen Zutshi: Yeah. The way I think about it is ultimately, what is IT responsible for? We have responsibility of building things, whether that is building through our own development or bringing in a SAS company and integrating that into our fold. Alternatively, we're providing a set of services through either products or through a service like HelpDesk. And who is providing that service? It's primarily the engineers. Everyone else is overhead. So having that notion and understanding that it's an inverted pyramid and we are ultimately helping them become productive.
Jason Felger: Naveen, you've gone into a lot around where your priorities are and where they've shifted. I want to see if we can dive in a bit for everyone. Just how you think about the prioritization of the spectrum of responsibilities you've had. You've touched on everything from one end of executing and delivering against a cloud native infrastructure to you've mentioned just being more customer centric and being with customers. It is a extremely broad set of responsibilities from being a customer and outward- facing, all the way to, as you said, building the things and deploying what needs to be done. Would be great to give people a sense of how broad is that mandate? And how do you just think about the prioritization of those massive responsibilities, both internally and externally?
Naveen Zutshi: It's probably four to five. It's actually four to five different strategies coming together. There's a strategy around customer journey and making the customer journey amazing, beautiful, simple, great experience for our customers. There is the employee journey and making the employees as productive as they can be. There is the data element, using data as a strategic weapon. There's the automation journey, this notion of the company is running all the time on algorithms. And there is a few other things. But if you think of broadly speaking, those are four or five things that every IT organization is focused on. Now, there will be specific strategies for a company that is in retail or manufacturing or others, but these four are fairly interchangeable and probably, in some form or shape, in a strategy for an IT organization, regardless of where they are. So if you think about beyond security, which always continues to be in top one or two priority for CIOs, the two others that I often hear and most often hear, is data and what are we doing with data and how can we turn data into a strategic weapon for the company? And then second, is there an automation? That has even accelerated further under this pandemic.
Saurabh Sharma: That's a great segue into something that I've been thinking about. The world of automation. It's pretty fashionable to talk about automation and then the transformation for every company. Every annual report, big narrative is around that. What is your opinion on the digital transformation phase you're in today?
Naveen Zutshi: Yeah. So while it is true that software systems or SAS companies are... If you're building your own software, that's the core of what your automation is around. I'm increasingly impressed by how little it touches beyond the core functionality. We were doing an analysis of a day- to- day financial analyst in our company, as an example. How much time are they spending on spreadsheets versus in systems?
Saurabh Sharma: That's right.
Naveen Zutshi: How many decisions are they making using spreadsheets? What does their day- to- day look like? And whether it's Microsoft productivity tools or Google productivity tools, they're probably using those more than anything else. And so you would say automation hasn't touched them as much as you would think that it has. I'm not deriding. A lot of the great applications can be built on spreadsheets and people have. But if you think beyond that, every time we want to forecast our business, you want to have some level of automation associated with it rather than just having humans do it every time. So I feel like process automation, if I look at IT back office automation, this tremendous amount of opportunity is there. So for example, we use our SOR tools like inaudible for our IT back office automation. We use RPA for front office automation. And talked to many CIOs in different companies and they are big in this journey because they see immediate value. This has been one of the revelations also. The value curve is much faster in terms of the businesses themselves. And then if you look at automation beyond that, whether it's low code, no code, that is starting to take off. Not taken off as much yet, but it will. And then if you think about other ways where we can find ways to reduce the number of manual touches. However, it is activation of a product, creating an opportunity by a sales rep, taking an order. And then you go into proactive and predictive business use cases, which are traditionally done by employees, can be supplemented or complemented by machine learning.
Jason Felger: Naveen, has the pandemic changed the priority of automation initiatives for you? Many technology priorities got shifted over the last year and a half. And automation is one that's uniquely centered around people in many respects. Has it accelerated, decelerated, stayed the same as a category in and of itself?
Naveen Zutshi: It has significantly accelerated for us. We started the year before with 20, 000 hours of total FTE automation that we measured. And we are happy and patted ourselves on the back for that. And we said," Why not do it 5X more?" And we are on target to do a 100, 000 hours this year. And even then, I feel like we have done little. There is so much more we can do. And that's just one metric of measurement. You can measure in many ways. I feel like whether it is, how do you look at virtual Genius Bars? Because give you an example of HelpDesk, We would have folks come into Genius Bars and they can get the help anytime they wanted. Now, you can't do that. What do you compliment? How do you replace it and still maintain a high NPS score? And so having virtual chat bots that deflect some of the using an unused, and then compliment that with having multiple channels of communications from a user to HelpDesk, and then having tons of automation in the backend to have software assets. If you want to get a new software asset, it's just all automated for you. You drive more persona- based profiling of users and drive change in terms of what tools they need as a result of that. So there's a ton of innovation, actually. Very interesting innovation that can be done in this area.
Saurabh Sharma: Yeah. Maybe just expanding on that, so in the automation realm, we talked about cloud, which is obviously linked to scale, RPA, which is around process automation. There's obviously a lot of machine learning technologies coming out. Could you share your successes with which of these that you felt that you've tested and have obviously proven it out that these are all there and maybe the efficacy on driving value? You already referred to that, but just maybe calling out things that you've personally tested and felt that these are at the maturity level for everyone to go and try it out.
Naveen Zutshi: Yeah. RPA for sure has been there. SOR, for sure. Both in stock automation as well as IT automation. I would say on the NLU front, natural language processing and using that to drug derive, not just context and meaning from data. Increasingly better technologies are coming out, both on customer support side, HR support side, as well as IT support side, and really using bots that are not annoying, but actually are helpful.
Saurabh Sharma: Right.
Naveen Zutshi: Right?
Saurabh Sharma: Yes.
Naveen Zutshi: I used to find bots incredibly annoying. Vision- based ML is there. And whether it is physical security and complimenting the guards, or whether this other technology that uses vision, whether manufacturing or IOT or other traditional companies that require that, there is increasingly construction space. All sorts of interesting examples that I've seen. And while some may be in POC stage, some may be in early development stage, some are actually in full- blown deployment stages. The other areas of ML is still work in progress. For example, a lot of proactive and predictive use cases, you have to prove it to the users that these are effective. And I know you have a strong sense on this as well because we have gone through this example. For example, take next best action for sales. Yeah. That's a common use case. Many companies have tried it. Unless you don't show how you came up with that, and even the first example of where a probabilistic model was wrong, folks will use it as a way to say that it doesn't work, whereas it's a probabilistic model. Other learning for us has been being probabilistic model of the error rate is let's say 5%, and the added expectation is less than 1%. That's probably not a good use case for it because our many algorithms we try, we can't get to that trait. And data is a huge issue as well. How clean is your data? How much data do you have? And then finally, the other point I felt is it takes time. It takes significant amount of time. It takes a lot of experimentation to get even one use case right.
Saurabh Sharma: Yeah. You touched a really good point. One of the biggest challenge for most companies built around the value proposition of machine learning and setting it in enterprises is getting the trust factors internally. You might have a champion, initial champion, but for the champion to then get broader adoption is very tough because the expectation from the machine is significantly higher than the expectation from a human. How do you think about it? What are the things that companies can do to set the expectations right? And of course, we talk about within the AI adoption curve, there are elements that are now coming out, which it's more of a question that would those help? And so do you think, one, are you getting more comfortable, these technologies, as you go and then what is the best way to build that trust and then get comfortable and drive adoption?
Naveen Zutshi: Yeah. It's less about me being comfortable. It's more about my teams, my employees...
Saurabh Sharma: That's right.
Naveen Zutshi: ...Our employees, our customers being comfortable with this. Explainability helps. Explainability frameworks help, but it's difficult and it's tricky. So for example, if you take an opportunity, let's say in sales cases, we want to derive new space opportunities and you want to predict," Here is a new space where you should focus and here's the reason why we have said that." If you look at behind the scenes, the development probably is using 150 different features. You're not going to show all the 150 features completely. Folks don't understand it. The level of complexity and the level of interaction between an algorithm is so high. And so you're dumbing it down and saying," Okay. Here are the top three factors that are the reasons why you should expect this opportunity to work." And in many other cases, it does work. So for example, there are companies that are coming out that will explain to you,"Here is the opportunity score. Here is your forecast score." And increasingly, sales reps are becoming more and more comfortable using that data. And I think that evolution is going to take time. The ML models are also going to improve and our ability to explain those ML models to our employees is going to improve and evolve as well.
Saurabh Sharma: I think we've talked quite a bit about the infrastructure upgrade, the automation technologies. To the extent you can share, what is top of mind for you, Naveen, from here on? From both technology perspective and things that you think are still somewhat unsolved or high priority things that you're thinking about, and to the extent, your counterparts that you talk to are thinking about for next three to four years?
Naveen Zutshi: The first and foremost is as employees start coming back in, it's interesting to say," How do we bring an amazing experience to those employees who want to come back?" Everything related to real- time access to data to drive decisions continues to be a very high priority for companies, including ours. How do we drive better decisions using access to data that is at our fingertips? And there is a lot of innovation here that's going to continue to happen. And then finally, I am big believer in the company that never sleeps. So how do we constantly find ways to have algorithms running while employees may be sleeping or when the employees come up?
Jason Felger: Naveen, we always like to end with the entrepreneur and the founder in mind. And you work with many startups. So would love your perspective of one, is we've talked about trends. Are there trends and technologies that you're thinking of that you seek out and would think would be really ripe opportunities for entrepreneurs and founders? And then also, you are a CIO. You have been in this seat for a while, and just your advice of expectations and things to maybe do and not to as entrepreneurs and founders who are thinking about selling into the CIO?
Naveen Zutshi: Let's start with the second one. Founders' passion clearly comes out. Their commitment to their company, their commitment to the idea, their commitment to what they are trying to solve. My advice has been it's often you come to large companies and you get swayed by orthogonal use cases, which will take your company in the wrong direction and say no to that. You will gain more credibility. And the founders are effectively the best salespeople into CIOs, into CSOs. And if you use that as a mechanism to open doors, it will. In terms of problem spaces, there are, I'm sure, people a lot smarter than me who are working on tremendous model ideas. And I hear ideas all the time where they've taken simple existing problems and transform them. From insurance, to construction, to inaudible. Everything else has been transformed. Anytime you see friction, there's an opportunity to change that and simplify that, streamline it, automate it. And so in a B2B company, there are many areas where we have still have friction. I talked about FPNA, how we do planning. Seems like unsexy area, but there are very traditional companies solving that problem today. And then you get into other use cases which are security related use cases. There's a ton of opportunity there as well.
Saurabh Sharma: Awesome. Well, this has been great, Naveen. Just on a lighter note, I'd like to ask some different questions than our normal thought processes. And thinking about the old school world as we're moving to this digital world, what are the things you think you might miss from your analog era, for lack of better term? Because we are so automated, we are moving to hyperautomation, and things are moving so fast and rapid. Then obviously, the metric baskets. So I'm curious, things that you think could have stayed the same, but obviously are going to change?
Naveen Zutshi: Well, I love vinyl and they had taken a huge comeback. They had made a huge comeback. I think books. I still love holding a book in hand and while they have been transformed and you can read on Kindle and the conveniences there, the quality of shuffling paper, I love it. And so I'm sure a lot of folks feel that way as well.
Jason Felger: Well, Naveen, this was an excellent conversation. And Saurabh, thank you both for joining. It's been wonderful to chat with you both.
Saurabh Sharma: Thank you, Naveen.
Naveen Zutshi: Thank you, Saurabh. Thank you, Jason.
Kathleen Osgood: Thank you so much for listening to this month's episode of the Jump- Off Point, an original podcast by Jump Capital. If you have an idea for the show or know of someone who would make a great guest, please contact podcast @ jumpcap.com.