Doug Merritt: Creating a decacorn in Splunk and identifying trends

Welcome to Inside the Network. I'm Sid Trivedi.

I'm Ross Haliluk.

And I am Mahendra Ramsinghani.

We've spent decades building, investing, and researching cybersecurity companies.

On this podcast, we invite you to join us inside the network where we bring the best founders, operators and investors building the future of cyber.

Today we have a special guest, Doug Merritt, who has shaped not a unicorn but a decacorn, a company valued more than $10 billion. As the former CEO of Splunk, Doug steered this big data company to extraordinary heights. In just 6 years under Doug's leadership, Splunk's market cap soared past $25 billion and its annual recurring revenues skyrocketed from $200 million to a staggering $3 billion. At its peak, over 50% of Splunk's revenues came from security applications. Join us today as we dive into the mind of a leader who doesn't just think outside the box, he builds a whole new one. Doug's journey began as a coder and a developer before transitioning to sales leadership and eventually CEO of a publicly traded company. His entrepreneurial roots trace back to founding a company called Icarian, which was acquired by Workstream. He then went on to hold senior management positions at tech giants like Cisco, SAP, and PeopleSoft. Today, Doug is the chairman and CEO of Aviatrix, a secure cloud networking leader that has raised over $350 million in venture funding and serves over 500 global customers. In our conversation today, Sid, Ross, and I will explore Doug's career evolution, insights from his founder experience, and key lessons learned while scaling Splunk into a decacorn.

Hey Doug, welcome to Inside the Network. We have 4 different sections to talk about today, Doug, and we're going to start with your background. And let's go all the way to the beginning. And the beginning was you kind of spent 3 years as a coder at Accenture, and then you diverted your career into sales and you started as an inside sales rep and you moved your way up at some pretty great companies from Oracle to PowerSoft to Patrol. Many founders have this technical background that you started with, but they haven't spent a lot of time with customers. What advice do you have for founders who are trying to build that selling DNA that you have built over multiple decades now?

Yeah, it's really in my 2 years of quasi-semi-retirement, whatever I want to call the time period between Splunk and Aviatrix. I spent a lot of time, uh, investing and advising. I jumped on a couple of boards, and I was surprised that most of my time was actually on seed and Series A companies, because Splunk and some of the larger companies, people, Softest, SAP I've been with, I thought it was gonna be more scale oriented, but it was great to get back to, all right, so what's it look like when you're hiring your very first CRO or you made a mistake and you're hiring your second CRO and second CMO. And this is an absolutely pervasive problem. Like I'll start with tech has shifted a lot over the past 20 years. I think tech oriented founders were a thing, but they weren't dominant. And that has changed where I think with product led growth and a lot of the core lower-level infrastructure enablement technologies, it's now much more typical to have an ex-engineer or a true deep technical founder as leader. And that first time period, months, semesters, years of trying to get the company off the ground before you decide to bring in a professional sales and marketing person, there's so much that the founder learns about customer interaction and how do you get those deals across, across the table. And it is really difficult for everybody to transition from that founder-led sale to more of a professional marketing and sales orientation.. And the two pieces of advice that I've had for those teams is one, give yourself grace and recognize that you have not been a seller. And at this point in time as a founder, you probably are not going to be someone that's carried a bag and been, been on quota. And until you've done that, it's really, you don't have the true understanding, right? They're the same respect that you'd give to the depth of understanding you have to have as a developer. That it's not something you picked up in, in months and you can't just read books. You need to actually do the work for a while. That exists also in sales and marketing. There's some art to it. There's a lot of science, a lot of experience. So my coaching to everybody is be super curious, like that growth mindset thing we talk about. It is really important. Be inquisitive, be curious, be hyper-empathetic. And then two, you've gotta be patient cuz the second mistake that I've seen most early most of these teams make is you're on a growth trajectory. The plan that you've sold to VCs is continued up into the right. Those sales were really hard fought. Now you're dealing in the classic Geoffrey Moore Crossing the Chasm. You're dealing with those leading and early adopters, and it was hand-to-hand combat. And you as the founder, and whoever you're doing that with, were pivoting and learning and adjusting and tuning. There's this setup that I see with most CROs coming in. One, not understand that's gonna take that while you and your mind have perfected the pitch and while you think you've got product market fit, it's probably not scalable yet. And you've gotta give the new person that comes in a couple of months of tagging along with you and listening aggressively to how you're interacting with customers and learning, going back to the few that you have, what really worked so it can be put into something repeatable that, that person and other people can understand.. But the other piece with that patience is that leader typically has 4 very different things that you expect out of them. One, close deals. That's a full-time job. Two, set up all the systems so that we actually have got a reasonable instance of Salesforce and we've got Gong or something like that in place and we're collecting data effectively. And that's a full-time job. Three, create all the processes and enablement and methodologies. Like, how are we gonna sell? Are we medic or med pic? And, Are we gonna have a BDR team, an SDR team, and how they interleave with the core reps? And so a ton of work there is that, that full-time job. And then 4, be, uh, use your network and be a super recruiter, which is also a full-time job. So you've given these poor people 4 full-time jobs and sequencing and prioritization I just find is so important in every leadership role. And you've gotta help them with the prioritization. And I tend to emphasize, let them learn, let them understand and build the materials and put the sales processes in and begin to build the systems without this huge onus of go out and hire a bunch of people and you're responsible for closing all those deals. And the opposite usually happens, which is go close these deals, most important thing, and start to bring people in. But the people, they, they don't have time to really understand how to make it repeatable and they don't have put any of the processes and infrastructure in place. And then you hit a wall and that's why you're usually not too long down the road looking for your second CRO and then your third CRO and So it's just, it, but, but that kind of goes all the way back to setting expectations properly with VCs on when you're ready to bring in a professional VP of sales, director of sales, CRO, whatever the title's going to be, how do you set them up for success?

Many today think of you as a public company CEO, but you've also walked the founder's path. You started Ikarian back in 1996 and served as a CEO for 5 years right through the acquisition by Workstream. What were some of the key learnings for you as a founder, and how have you adopted that as a CEO?

Uh, yeah, great question. I had almost the opposite learnings of what I just talked about, which was I coded for a couple years, very, very mediocre coder, so I was reasonably technical but not deeply technical. And then I sold for a lot of years, and I was really good at selling and pivoting and listening, very consultative. And I, over the course of my career, and I whipsawed the organization pretty dramatically. I didn't really understand what it meant to reliably and scalably and repeatedly deliver software. I didn't really, you intuit it, but you just haven't lived it. Going back to until you've been in that org, it's difficult to really understand it. I didn't understand how much space product managers need to make sure they harden the PRD and the MRD and that they actually get a chance to define. And then engineering needs time to actually go through the full development and release cycle. And so I'd be great at going and talking to a customer and they were a little bit different than, than the last customer. We had sold AT&T and on what Icarian did. And then we talked to Kaiser Permanente, totally different organizations. Like, oh no, no, we can do that for you as well. And so my biggest lesson through Icarian was having walked a mile in different departmental shoes is really important to be a good CEO. And I logically and kind of emotionally learned what it was like to be in engineering and, and where those mistakes were. And we wound up with a reasonable outcome for the company, and, but it wasn't what any of us wanted, which is what led me to take the head of engineering job at PeopleSoft, because that was a huge radical departure. But so core lessons, again, be, being a super aggressive listener and be really empathetic to all the different facets across an organization. And while you're really urgent, you have to be as a founder, and while you know more than most because you have got the image in your head, you really have got to slow down and listen to your folks.

Thank you, Doug. I like that term, super aggressive listener, and of course, being empathetic to not just a few, but all the different facets across the organization. Now, companies like PeopleSoft, SAP, and Cisco are very different from the scrappy startup world that we live in. Help us understand some key differences in being a general manager at a company like SAP or Cisco versus being the CEO of a startup.

I think the number one thing that I took away from first PeopleSoft, but then it carried to SAP and Cisco, is the power of a platform, the power of market share and customer base is insanely powerful. Like you look outside and it's like, man, that's, that's a pretty good thing to have. But, and VCs definitely appreciate go-to-market is really tough. Building that large customer base is really tough. And therefore enterprise sales is one of the key elements to get there. But until you sit inside those companies, you don't really understand the lock that these companies have with that, with the buyers that they serve. And one of the better examples, when I was at SAP, the first division that me and my team kind of came up with, it wasn't really a charter, became the governance, risk, and compliance division. It was right after Sarbanes-Oxley passed and we came up with the whole product suite to help the CFO function with all the chaos and craziness of managing finances in a Sarbanes-Oxley driven world. And we bought this small little company and one of the top SAP customers at the time had bought a competitor and the CIO was actually on the board of advisors. And I think we even was an investor in that competitor. And we had him out for a briefing in Palo Alto as part of the Palo Alto product team and walked through the roadmap of why, who we'd bought and why it was important and where the direction was going. At the end of that meeting, he called his team up and they had already bought and were rolling out the product, said, I mean, he's literally in the conference room, said, cancel the rollout and I'm stepping off the board of advisors because SAP is going this other direction. So it's, Getting to the point where you have the customer loyalty that a large organization has is, it's a goal that most of us want to pursue. It's very difficult to do. And as a startup, the more awareness you can have of how powerful that relationship is and think through how to take advantage of that, like what, what, what can you do on 1 1 3 basis that really makes that large organization look good is one of the more important things you can do to get you across that chasm and to get some of the volume. The other piece I learned is being a general manager of a big and complex division. It was a billion-dollar-plus division at PeopleSoft. We wound up building close to a billion-dollar division at SAP from close to nothing. Cisco is very different. That was kind of a functional leadership role. But as complex as that leadership is, it's still very different than being a CEO. And you can kind of delude yourself to an extent because you've got, for me, especially at SAP, I had full span of control, everything down to sales, specialty sales groups. Uh, but when your neck is on the line every single day as the accountable person reporting to the board and the investors, and every buck truly stops with you, it does change everything. It's great, great training because you have to be good at, you have to understand engineering, you have to understand marketing, you have to understand sales to make these roles work. You do get some of that empathy that I was talking about earlier on how difficult it is, how different the departments are and how difficult it is to manage across them. But it's still not the same as being a founder or a CEO.

I think this, this comment and this point around empathy and humility just keeps coming back. And, and clearly that was something you learned, whether that be, you know, as an individual coder, as an inside sales rep, as a founder, and then finally also as a, as a GM. And I, I think it set you up in a very good position for that, that time at Splunk.

Yeah, definitely a set of life lessons. It's, uh, we're coached early on in school, any of us super, super Type A folks, they You want to sit at the front of the class and be the smartest kid and always answer all the questions. And the, the, the, the right answer is what matters. And once you get outta school and you start to work, you realize answer's important. Like getting to it, the right answer matters, but there's lots of right answers and getting people to understand where you're going and understanding the feedback from them and that kind of, that whole working organism that can evolve if you let it evolve. I, that's the hard part is how do you get 10 oars or 100 oars or 1,000 or 10,000 oars rowing in close to the same direction? And I think a lot of people are writing about empathy and humility and growth mindset. And because that to me is one of the foundational elements of how you make that happen.

Let's move on to the second section. And this section is really around scaling Splunk. And you first joined Splunk to run field operations under Godfrey Sullivan, who had helped to take the company public and created this new darling in the SIEM market. And you became CEO in the fall of 2015. The company at the time had a market cap about, of about $7.5 billion and about $200 million of ARR. And it was back then best known as a log aggregation tool, helping IT ops teams find the root cause of failure and decrease mean time to respond. Over the next 6 years as you as CEO, you grew the business to nearly $3 billion of ARR and the market cap peaked at $35 billion. How did you decide to extend this market opportunity and evolve the platform so that it fully went across IT, DevOps, and cyber?

It was an incredible 7 and a half years 'cause about a year and a half of, of CRO/SPF FieldOps. And there's a lot of, yeah, everyone credits all the brilliance that they have on, these great leadership roles and a massive chunk is luck and timing. Splunk has this incredible product with the Core Index and landed and did a good job of helping the market be aware of the value of logs. And I got, I got to inherit that. But when I came in, SIEM wasn't a thing yet, actually. It really, we were still selling to IT ops, a little bit of cyber. And Godfrey was adamant at that point in time, like, we don't want Gartner talking about us as a SIEM. Or we don't even really want to go after cyber that aggressively. Um, there is really good rationale around that. One, we were playing at a very different level than the SIMs were. They're a structured data source. You refine the data that goes into a SIM so that you can create the incidents and event management alert frameworks. That doesn't generally happen against non-structured indexed data. You want to— we want that in memory and refined. So it's more of an application and we were an index, but two, security at that point in time was, didn't have the power they have today. And they were what we called Splunk huggers, right? We, what we wanted was people to dump a bunch of data into Splunk and then share it so that there was more and more use cases and insights that people would get from this non-structured data store. And cyber was really, really protective of the data, which went against one of our core tenets. Despite Godfrey's insistence and my carry-on of, hey, let's not wander into that SIEM market yet. There's also a strategic element to that. We tend to couple with HP ArcSight and the other SIEMs as the invest— we are the investigative data lake that IT ops use first and then cyber eventually start to use as well. And we complemented the SIEM and I didn't want to go in and tell the people that just bought the SIEM or done a big renewal, hey, that was a bad decision. Like, why'd you do that? Use us as a SIEM. So we were really trying to be cooperative, but what we, what we saw is there's a whole host of IT ops use cases. And then there are a whole host of SecOps use cases. That if we could harden an app on top, it would do a better job of helping people understand what the platform did, because this investigative, better MTTR data lake thing based on logs was still hard for people to conceptualize. So I think there were two elements that really helped get Splunk on that continued, really rapid hypergrowth phase. One was, despite the ubiquity of what the platform could do, constraining the number of use cases to a set of high-value, high-volume, generally horizontal use cases across IT ops and SecOps. And SIEM became the most important of those over time, but we had a whole series of applications, ITSI and, and a bunch of others that tried to crystallize for different points in the IT ops and SecOps landscape, how you could get immediate value from Splunk without having to grow into it based on doing a bunch of queries and finding the answer for yourself. Now it's hard to get across to Salesforce, um, because we were— Godfrey was really insistent we're not a platform, but, but we were a platform sales team in the sense that we went in and said, hey, here's this great tool, how do you want to use it? We could do it here, there, oh, what do you want to do? And let's investigate with you and let's collaborate. And buyers don't generally like that. They want an answer. I find with buyers, the more you make them think, the more frustrated they get. It's like, we're the experts, we've studied this, here is the answer. We've got a lot of customers like you. That's why you're doing business with us. Do it this way, please. So that was a big effort to get to more prescriptive selling and less consultative selling. Um, and it really was use case back. The second piece was making a bet on the cloud, which sounds stupid in 2024. And even in 2014, '15, it's like, what do you mean you're not in the cloud? But data is, as we're seeing, it's finicky. Like there's people are worried about it. Where does it go? And how many people see it? And is it safe in the cloud or not? And what's happening with residency and what's happening with privacy? And so So it was a bit of a gamble to go to cloud. We really got there, there's lots of good business model reasons, but the core was we had seen our bigger customers get to petabyte scale and then 10+ petabyte scale, and they're on their way to get to 100 petabyte and potentially exabyte scale. When I looked at the complexity and the technical skill necessary when you got to hundreds of terabytes and petabytes per day of ingestion, it's like, wow, yeah, a good tech company can do that. The Googles, the Apples, the Facebooks, but man, the average company, there's no way. So how do we help them? We've got to do it for them, which meant going to the cloud. Yeah, that was a great tailwind. And it also was— made life incredibly complicated because at the same time that we're trying— that we're in hypergrowth and we're trying to move to a use case orientation and we're going through a complete replatformization of the technology. It was not horizontal scale out. It wasn't microservices oriented. We had a lot of static relationship between the compute and storage layers. So insanely difficult for the company. You're now taking the street through this crazy journey of, hey, nothing's going to look good. Cash is going to go upside down. Revenue is going to go upside down. Op margins go upside, but trust us, it's all going to work out. It's not a journey for the faint of heart, but if you have customer success as a top metric, it was what we needed to do. And, and it wound up being, uh, I think a big, big lever for the growth of the company.

How much do you think the success of scaling that ARR 15 times from $200 to $3 billion was based on execution versus vision? Was it, you know, was the vision obvious, what you just described, and really what was hard to do was make executing on that vision, or was the vision the hard part? No one else had thought through those things, and the execution was, you know, as good as you could do with all of the changes that you are making.

I am firmly in the camp, maybe too rotated, that execution's a premium, not the idea or the insight. Especially in our industry, but I think in general, I think humans are really smart and in tech, I, you've, you've gotta be super creative and smart. And any idea you have, I just assume there's 100 or 1,000 or 10,000 people that have the exact same idea. The premium is how do you actually bring that thing to life? So log data, and we weren't the only ones, I just think we did it better than anybody else. And then we out, out-executed them. Getting that platform to cloud, if anything, we were late. Making this platform use case-centric, that's a tried and true orientation that you can go back to Microsoft and, and IBM at the beginning of tech, and then that's been a play that they've run. So I don't, you know, putting that all together and making it digestible for people and selling the people on the vision, like, there's definitely skill to that. But the hard part is, so we want to get to the cloud, how? What's the sequencing? How do you prioritize stuff? How do you deal with the unexpected turbulence that always comes up? What happens when you got to change personnel because you've got the wrong people? And that's just, that's just a grind. Of these jobs. And I think all the premium's in the grind.

Talking about execution, a big part of that is always about the people. Throughout your time at Splunk, you hired, mentored, and promoted well-known executives both across infra and cyber. From people like Hyun Song, who now runs a big part of the business at NetApp, to Jason Child, who is now the CFO at Arm, to Sneel Antani, who went on to start his own company, Horizon3. What are the one, two top traits you look for in executives and how do you motivate them to do their best work?

Yeah, we had a great team at Splunk and some of those folks Godfrey brought in, some I brought in, and we were lucky again to have this incredible brand that people actually wanted to work for. The way that I have learned to filter is when you're dealing with SVP or EVP or C-level hire, what we've been taught in most of our hiring through our careers is competence is really important. And most of us started in engineering or sales. And so you do a classic engineering hiring, coding tests and thinking tests, and it's all about competence. You know, can you actually do this job? At that level, what I learned, and had to learn it the hard way a few times, is to get to be a candidate at that level, your competence has to be high. You still have to test for it. Is it relevant? Is your competence relevant to what we're trying to drive? But I think it all becomes grit, resiliency, empathy, creativity, operational efficacy. There are more of the intangible elements that come with these execs. It's so interesting to hear people talk about the past because they make it generally sound so predictive and linear, and anyone that's gone through it knows it's so non-deterministic and non-linear and non-predictive. And, and there, therefore the skills really come back to how aware can you be of signals, which why I think empathy and listening is so important. How quickly can you process? How quickly can you make decisions often without information and without the comfort of, of assurance? And then how humble can you be on admitting you're wrong and pivoting quickly? And then what kind of grit and tenacity do you have to be okay being wrong and picking yourself up after you face planted for the third time and still being excited about moving forward? And those are harder to test for. I think, you know, I've spent a lot of time with folks and asked them much more of those situational questions of how is, how'd your life progress? And tell me about the hard issues you went through and describe those in more detail and tell me all about your failure. There's a lot of standard interview questions there, but that is where ultimately I've kind of pivoted to being the most important thing for the exec team hire.

Doug, one thing you've done very well is you also identified when the right person would be great at that point in time, as well as assessing when they weren't likely the right person for that point in time and swapping them out. What advice do you have for founders on how to assess when, hey, we need to upscale this role, we need somebody else? What's the, the thought process in your head that you think through?

I'll do that. I'll echo the trite advice of never ever can look back and say, geez, I wish I'd taken more time and given someone more chances before I had a hard conversation. And I can 100% look back, like literally 100%, say, oh my God, I knew. In my heart, like my, all my alarm bells are going off and I just, I was way too slow to act. And everyone will say that, like, that's trite advice. For me, what it means is I really like collaborating with my team. I learn from them as well as hopefully they, they get some learning from me. So I want to be, I want to be involved. The signal for me is when I feel like I'm repeating myself over and over. I keep asking for the same thing. Like we've agreed that this, this one deliverable is really important. You've got to get more granular on this aspect of operations, or I need you to create something in this area that's strategic. Or, and I'm back on my third or fourth meeting and I'm asking for the same thing and it's still not being delivered. That is usually that first early warning of, okay, something's wrong because like they've got the IQ, like they've got lots of different facilities to make it happen. They probably don't know how to get it done. And I'm trying to collapse the time even now, still after so many years doing this. How do I help them see they're having difficulty? What quick mentoring can I give to help them get over the hump that I'm trying to shorten time of? If you don't rush to the water and take the drink, I'm trying to show you the water. I'm not going to force you to drink. Then we've got to have a quick conversation of, I think, you know, maybe I need to hire above you. Maybe we need to move you somewhere. Maybe you need to move out of the company, but there's just no, it's really difficult, I think, to drag someone through that given the speed of what's happening in our industry and the expectations that. That we all have to live with.

Thank you, Doug. You've given us some great insights on market dynamics and talent acquisition. Let us shift our focus to another crucial aspect of your leadership at Splunk—strategic acquisitions. You led several key acquisitions like SignalFX and Phantom Cyber. Can you walk us through the acquisition process from the buyer's perspective? As you were the CEO of a public company, what factors did you consider when evaluating potential startups for acquisition. And let's flip that script over. What advice would you give to founders who might find themselves in discussions with acquirers? How can they best navigate these interactions and negotiations?

So one, whenever I hear a founder or CEO of a small company say, you know, what's your goal? They say to be bought. Yeah, we're targeting to be an acquisition target or something. It's like that. I I lose my mind. It's like, do you understand how difficult it is to get bought? Like that, that's the worst strategy in the world. You can never build a company to be acquirable. You've got to be open. You've got a fiduciary duty to yourself and your investors. You've got to be open to M&A if it occurs, but M&A happens because you're building something really special and unique that has the opportunity to become an independent public company. And if not, it's going to be very difficult to get bought, I believe. As a buyer, I've only been in high-growth companies. So my view is very skewed. I know that people buy for market share. I just try to take out the competition. I've never done that in my roles. I mean, I know companies like Broadcom can buy more for operational efficiency and margin enhancement and it's almost a great, you know, it's a, it's a PE company to an extent. I've never, never been there. I'm always trying to figure out how do we accelerate growth. And to me, that's two different things that again, you have to have done this well to be interesting. I'm generally not buying marketing and sales expertise because at PeopleSoft, at SAP, at Splunk, we had a lot of that and it's nice to get some additional great marketing and sales folks, but we already have got a great system in place. I'm trying to figure out a way to either fill out and round out aspects of the buyer's journey for what we do for that buying center. And it's just a build by partner approach. And we generally have determined like with Phantom, hey, something like security orchestration automation and response is really important. And we, we could build software. You can always build it eventually, but we looked at it's like, it's at least 2 years to build this, maybe more. You always underestimate the market's moving pretty quickly. Let's go out and we could partner, but we think it's pretty strategic. We want to actually, we think it's very tightly coupled with the core SIEM market. We want to own that asset. So let's find the best property we can, which was the same thinking with SignalFX, but there are a bunch of smaller tuck-ins around SignalFX that were trying to highlight different elements of the observability and DevOps world. But with that orientation, it's kind of a tech team tuck-in and a technology augmentation piece, which goes back to, to build a great company, you know, laser focus on the buyer, what the problem is, how do you take a unique approach to really solve that problem and ultimately disrupt the market because you're creating a brand new category, you're creating brand new capability, you're not doing product market fit of I'm just going to build a better, better mousetrap than the person behind me. I'm actually reimagining the way it gets done. And if you do that super well, you'll for sure grow within your buying center. And then you'll get potentially the opportunity of someone that already is bigger than you in that buying center saying, I want that. And if you do that right, you're in the catbird seat of, yeah, yeah, that combination makes sense. Or no, I really want to continue to try and go on my own. We've talked a lot about Splunk.

Let's talk a little bit about new beginnings and Aviatrix in that topic. You left Splunk at the end of 2021, and I remember sitting down with you and grabbing drinks in Palo Alto, and you spent a year and a half doing various different things. And then in July of 2023, you joined Aviatrix, where you are today as CEO. You had quite a few different options. I remember having a few of these conversations with you to be CEO at a whole different set of companies. You also had other options, you know, you could have joined a venture firm, you could have gone and, you know, started your own firm, you could have just taken some time off and served on board roles. Why did you choose Aviatrix, which at the time and still is, is a private growth company, a few hundred employees, very different from Splunk? Very different, yeah.

So I toyed with all that you talked about. I did a bunch of sidecar investing, I partnered with a couple VC firms, I toyed with do I want to be an investment partner. Versus venture or operating partner, but really cross the transom and be on your side of the table. And do I want to be retired? And through all of that, it's like, I'm just way too much energy to be retired. Investing is awesome and advising is great, but I like being on the playing field with the rest of the team and be in the middle of stuff. So I did get to like 2 years of back and forth. It's like, no, no, I need to be on a team. I love the purpose, the vision. Purpose, mission, everyone's in it together. I was not targeting Aviatrix in any way. I, I did a little bit of networking at Cisco, but I would put myself as a very, very lightweight, I guess, advanced beginner on networking. And it's a tough category, but we had one, a couple of board members I'm close to that I deeply respect. And we have a long lasting relationship with, and ultimately we all have that maxim of you join a company and you leave a boss, like you're board is your boss and you've gotta love your board. So that was, that was a strong draw. I wanted something with a decent— so I determined private would be interesting versus public. If I wanna be back on the playing field with everybody, that means you're in the trenches and you're interacting with customers deeply. You're going through product review cycles. You're deeply meshed in QBRs. You're in the day-to-day fabric of the business. As a public company CEO, you get that, but there's a reasonable chunk of your time. Yeah. Today's blue screen of death is a good example. Like that's a huge distraction from what that CEO would normally be doing. There's a reasonable chunk of your time that is not that. And the bigger you get, the more time you spend not inside the company, just by the nature of the job. So that led to private. And then within private, there were a bunch of criteria, but at the core is who do you serve and how unique is what you do. And, um, what Aviatrix does— so Aviatrix, uh, I didn't know this before— Aviatrix is a female pilot. Um, I guess you call it an aviator, but if you're a female, it's an aviatrix. And we had, we had a female founder. And that metaphor was, hey, we're in, we're cloud networking. Um, and her idea was like, it's like an air traffic controller or a pilot. Like you're trying to make sure that things go from one spot to the other in the air. Um, which makes sense from once I realized that it's like, oh, okay. It's kind of a cool name that describes what we do. Um, but, uh, what I found in my couple of years off is the landscape is so, so crowded right now with so many companies that are well-funded that are all lookalike. And for a whole series of reasons, networking has got— been overlooked to a large extent. And what we certainly found is we reconfigured Splunk completely to be a cloud-native service. To work well in the cloud, everything has got to be rethought. And right now, Aviatrix is the only company on the planet that has a cloud-neutral data plane, the actual packet mover. We touch, filter, route, optimize packets at the packet layer across all the major clouds and now onto edge providers, the Equinixes and Megaports of the world, and into some of the next generation LLM providers. And I looked at that and said that it's like Splunk. You're serving a super technical buyer, network engineer, network ops. It's kind of like IT ops, IT ops engineer, underappreciated, right? They've got an incredibly difficult job. The entire weight of the world sits on their shoulders because as we're seeing today, when systems crash, like the world stops. And networking couldn't be more of a ring-zero activity. Like, if you can't deliver packets, nothing works, other than maybe if an OS crashes there. I guess they're similar. And it's not that crowded. The alternative is cloud-native networking services. And the clouds, the whole message was, developers, you're awesome. All those other people, like, don't you want to get rid of them? Like, don't you just want to code? Like, forget networking and storage and infrastructure and QA, like, just go code and we'll take care of the rest of it for you. And that's just not true. And networking is really, really difficult. And what happens on that last mile, and what happens as you traverse multiple different hops, that you need visibility, you need control, you need resiliency, and you need effective and efficient developed services. And so that combination of board that I know and like, a good balance sheet with ability to maneuver and a really compelling problem for a difficult buyer that I think is hard to replicate without 10 other people that have something that looks incredibly similar. For me, it was a good concoction of, okay, get back on the playing field and see what you can do.

You've talked a lot about the reasons why cloud networking is such an intriguing market to go after. How are you looking to evolve the business Aviatrix is in? I would love to understand what your vision and your priorities are. Continue with that Splunk parallel.

Two of the major initiatives, there are a few others, are literally replicas of Splunk. One, right now the software is completely customer self-managed, which is an advantage. It's all native in the cloud, but because we're literally touching packets and we don't control the underlay, all the physical infrastructure is controlled by the cloud providers themselves, then the customer's got to retain control of the data plane because they are accountable and they are responsible for the data inside those packets. But there's big chunks of what we provide that could be delivered as a service. And we can make the data plane feel much more like as a service. So one, like Splunk, one of the big initiatives is, all right, let's make it easier on people. Networking is a very difficult area, much like trying to manage complex and super high volume data environments. There's only so many people with skills to do that. Let's shorten that for the customers so we can actually provide them the skills that they might have a difficult time finding and procuring and give them a better experience. The second piece is Aviatrix has built this really broad-based platform, networking platform. We do all different combinations of routing use cases, load balancing use cases, network address translation, NATting use cases, firewalling use cases, and we're continuing to wander into other areas of networking. That's awesome, but like with Splunk, customers need crisp use case. And so we're pivoting the company to be much more use case-centric of, yep, you can do anything with this thing. It's a box of Legos for companies like you. Here's the two areas that we've seen them start with over and over. And then here's where they go next and next. So it sounds trivial, really, really difficult to execute across a country, a company, especially one that is pretty firmly entrenched in let's be consultative, come in, see what you need. Like, well, lots of investigation and we'll work together to figure out what you need. The two areas I'm really excited about is long-term what the world is wanting with networking, given how complex it is and how much more complex it's becoming. Like we thought everyone said cloud would make it the world simpler. Cloud made the world much more complex. There's hundreds of services per cloud. Those services are not defined the same across clouds. So, and you still are holding onto your last generation technologies and you should for a while, because not all those things belong in the cloud. Now we've got generative AI and the, uh, invasion of a whole new set of data centers. So the complexity of the world, I think, is going to be exponentially increasing and having a consistent networking visibility, manageability, cost optimization, reliability layer, I think becomes even more profound. So we're expanding what the current things we do across even a broader, broader landscape. And then two, the power of GenAI has been to do some of the more difficult and high pattern matching thought work for human beings, so we can continue to kind of go up the stack. Um, and networking is one of those really, really difficult areas, going back to that complexity I talked about. Um, and where I think we eventually need to get to is autonomous, self-healing, self-managing networks, and because we're at such a low layer and we see so much with the way that we implement, we've been working for the past quarter, a couple quarters on how do we take this data that we're accumulating and a lot more data we're going to start to accumulate as we begin to move pieces of what we do to as a service so that we can actually go down that path. And again, take the load off the customers wherever possible on intent-based networking. I'm trying to get this thing done, render the network for me. Self-adapting networking, self-healing networking. And so those are the two, like we don't really fully play there. We've expanded to Equinix and Megaport on the edge. We've got more work to do across these next generation data centers. And then brand new ground on what we're doing with the GenAI work.

It definitely sounds like an exciting time when GenAI meets networking. And we completely agree that when we think about networking, we often do not realize its complexity. And so what you're building at Aviatrix sounds really cool. Let's talk about leadership changes as you bring your own fingerprints to Aviatrix. How do you see the leadership team evolving? How do you see its vision playing out over the next 5 years? Both Sid and I know John Donnelly, who used to be at CloudKnock Security, a company that both Foundation Capital and my firm SecureOctane had invested in, and Jon clearly is the kind of a guy that can go from 0 to 10. Who are some of the other people that you have identified to come in and help you at Adiatrix?

The interesting environment that I came into is, like so many companies in fall of '22 and spring of '23, we went through a ton of changes. The funding environment had changed, buying environment had changed, and through all those changes, a bunch of downsizings and restructurings, by the time I got there, the, what the normal executive team, the C-suite, whatever you want to call them, had all left, almost all left. We had a CFO still. And so I walked in and I forget the exact account. It was somewhere between 19 and 23 direct reports that were a combination of directors and VPs. And I did a very contrarian thing. The usual approach, and I fought my, I went back and forth with myself. I still go back and forth with myself on what's the right approach, is immediately open up a bunch of recs. I know I was going to need a CMO, a CRO, an SVP of engineering, a chief customer officer, like just a CHRO, just go across and open them all up. The quicker you can fill out that team, the better. But I had an appreciation coming in. Networking is really, really complex. That's one of the reasons why it's not overfunded right now. And this company had been in business now, we're close to 10 years. Like they've been at this for a while. Take the time and learn. You've got this rare opportunity to really understand what does the landscape look like? What's the culture of the company? Where do you want to take the company over time? And then hopefully you'll do a better job of bringing in the right exec with the right skills and the right personality and the right cultural fit for what, what you need. So Jon is a good example. He started as a consultant for us and he's added a ton of value. But it was, I was trying in every way possible to be, be thoughtful so we could guide this company in, in the right direction. We're at the point now we've, basically filled out that slate, which is great. Um, we— the latest hire, which I'm really excited about, is our, uh, Senior Vice President of Engineering and CTO, uh, a gentleman named Anirban Sengupta, who most recently was running all the Kubernetes landscape at, uh, at Google, at GCP. So, but he had been NSX, the SDN layer at VMware that Nicera found his way into, or I guess Nicera helped start. And he was at Cisco before that and a kernel developer for that. So, you know, really good background. And I may not have gone there right away way if I hadn't spent more time understanding the technical landscape of— there's this really interesting combination of knowing last generation network networking, knowing software-defined networking, which never really got all the traction it should have, knowing cloud networking, and knowing next generation compute and ephemeral workloads. But I feel like, you know, we've got a, we've got a really good team now, and, and we'll continue to evolve that team. But a journey of patience, like, there are many months it's like, oh my gosh, if I just would have accelerated this a little bit more, I wouldn't have— yeah, I wouldn't be up at midnight and, and worried about this topic right now.

So we'll, we'll see. Talking, talking about hiring in this time when it's starting to get competitive once again to hire high-quality talent, what advice do you have for founders on being able to go and get the people they really want on the team? How do you go and do that? How do you go and sell this, this vision when you're at a startup?

I think it's one of the best times to be recruiting. Right now, there are a lot of people in flight, right? There were a lot of layoffs. There's a lot of churn. There's a lot of missed expectations out there. I think the hardest, hardest piece is spending the time to get truly clear on what is it you need and, and being patient, like having, maintaining your bar and not caving, even though the board is probably gonna be on your back, you're gonna be, you know, your own worst enemy. And I've gotta get this thing done to make sure that you, that you bring in the right person that you've spent the time on the role, you know what you're looking for, both quantitatively and qualitatively, going back to some of those characteristics that we were talking about. And then just don't cave. And it's, and again, so trite, everyone says it, but when you're in the moment, it's hard to actually live the advice. There's one-way doors and two-way doors. Hiring is a one-way door decision. It's reversible. It just is really expensive and difficult to reverse.. And so giving yourself enough support, enough multiple pairs of eyes, um, rigorous interview frameworks, and then patience to, to make sure that, that it's okay. Like you might even lose a candidate or two because there's a process you want to go through and there's assurance that you want to bring in the right person. Um, and it takes a lot of courage to do that properly, but there's so much talent out there right now. It's, I've, I'm blown away. At, at how enthusiastic— I mean, there's a whole new love of tech. I think Gen AI is completely— people were trying to figure out blockchain and was it gonna be real or wasn't gonna be real, and, and it's got its value and it— but it wasn't a revolution like Gen AI is. I think Gen AI is the entire industry back leaning forward, and people are coming out of retirement because of it. And so I think there's— it's a great time to be hiring.

You already mentioned Gen AI, and we have to talk about it. And then our last section, we're going to talk a little bit about catching trends. I think For the listeners who've listened so far, they will have recognized that you're particularly good at catching trends and figuring out how to build out a vision to really scale a business. And one of those trends is GenAI. What are your thoughts on this GenAI trend and where do you think there are opportunities for innovation that founders aren't thinking about today?

One, so I came in, it came in, I was retired when ChatGPT, 3.0 and 3.5 were released and spent a lot of time on YouTube videos and playing with both with ChatGPT and then the many that popped up, Claude and Mistral. And I know that I'm out of sequence here, but I came into Aviatrix with the kind of Chamath perspective of this is going to change everything and I can change it now. So July and August of last year, it's like we are 100% GenAI based. We brought in a whole tool stack. Uh, whole abstraction layers. We could play with multiple LLMs, low-code/no-code framework, non-NoSQL-based data repository that we had control over, like everything I thought we needed. It's like, and we're going to rebuild all these applications and we're only going to, we're going to be able to get to $500 million in sales with 20 enterprise sales reps and couldn't be more excited about the potential of GenAI. And what I've found a year in, and we've got a good tech team, is wow, it's really hard. Like the promise is there for sure, and I have no doubt that we're going to get there. But it reminds me of trying to build, uh, multi-tiered internet apps in 1997, '98, '99. Doable, doable. Just a lot of baling wire and bobby pins to try and get everything from the web interface before web app servers existed all the way back to the database. And I feel like that's where we are right now at GenAI. So we're still on it, investing like crazy, but I think we're still down at that lower infrastructure layer. How do we make a GenAI-oriented software development lifecycle workable to crank out really meaningful and interesting enterprise apps that can radically change the legal department, the marketing department, the sales department, the support department? I know there's all these stories of, oh my God, we've doing everything with GenAI, we've got 3 support engineers. I, I literally do not believe it, having played with 3 different frameworks now and the one that we're building. It's really hard still to make these tuned and interesting and learning oriented and non hallucinogenic. And so it really works at the quality level that we need. So I guess my first breathing of it's okay is no one's losing their job anytime soon, but it will happen over time. Like it's, it's going to climb the curve pretty quickly, but I'd be investing. Obviously, I think most companies have missed the core model wave that's being dominated by a few. Super successful, very rich. It's hard to do it, do it well, but all the magic's going to come in building the— in helping to build the applications. And how do you have specialized models? Vector DBs are obviously going to play a huge role in, in where we're going. Um, but that plumbing is important. And then the promise that— where everyone jumped to right away, but I think they're finding, like I'm finding, you can get there, it's just hard. The promise is going to be in apps. But we launched Sykarian in 1996. 1995, '96 when the browser came and again, none of the core infrastructure was there and we got an app out and the app was a multi-tiered app and it was all browser-based, but it was 20 times harder than if we had built it in 1999. And it was 100 times harder than if we built it in 2008. And it was 10,000 times harder than if we built it in 2015. We were buying data centers, leasing data center space. Begging Cisco for core routing equipment and Sun for servers. We had to build our own middleware layers and web app layers. We had to build our own coding frameworks, workflow engines, like all the stuff that's free, totally available, has been since 2015, '14, '13 on the clouds. We had to build from scratch. And that's what I feel we are now with these GenAI apps. And so the next Salesforce is out there, but the money is gonna be made right now. I obviously, ChatGPT and the core frameworks are where the money is immediately. And the, yeah, NVIDIA and the chips and then the frameworks. But I can see that every cycle is the same. Go back to mainframes, mini computers, client-server revolution, internet apps. Like you can go through the, I still believe in the OSI 7-layer stack. You can just see layer versus layer what needs to be built and where are we yet on maturity and then where do you fill in. And so I'm excited about the networking layer in this world. Like that's way down there. Layer 3, Layer 4, that needs, it's gonna need a lot of work as well. But it's awesome. I am definitely excited about the next 5 to 10 years.

I think the thing that I've been reading about this pretty extensively, and one of the most interesting pieces that I recently read was I tracked these CIO reports from the different banks and the research organizations quite closely. And Morgan Stanley pushed their most recent second quarter 2024 CIO report. And in that, they asked the same question of, within your IT orgs, which are the areas that you're most interested in adopting GenAI in? And in the past, last year, it was very much around, we want to use GenAI to help our customers. We want to improve customer success. And that has completely turned. And I think part of it is exactly why— what you mentioned, Doug, which is that folks have realized it doesn't work as well yet. So they're using— instead of focusing on the outward customer-focused experience, they're really focused on the inward. How can we improve employee productivity so that each of our different units can work harder?

Yeah, which, which again was a, a whole series of applications. Like, you go back every generation, like, what, where, where was all the big impact made? Well, it started with core financials, core HR, core sales, and then where did it migrate to? Oh, it all systems engagement. It's all about customer. But until you get that core stuff built, it's hard to get the systems engagement. To work. Yeah, so that is not surprising. It's good. I had not read that report. I'll have to go look at it now, but it would follow my intuition.

You've been very good at catching trends and understanding how markets can expand. What is the secret behind how you spot these opportunities? Is it something you read or people you speak with that are giving you all these lines of insight?

I am super curious. I listen to probably 10 different podcasts and I try and read a lot. But if I really trace everything back from Accenture, honestly, it's a network of people over time that most of the ideas that I, I don't think I've looked for a job since Accenture. A bunch of guys from Accenture that I knew, men and women went over to Oracle and it was like, hey, why don't you come over here? And then that has happened everywhere I've gone and almost every investment that I've made and I've gotten in way too many companies now. We have to talk about that separately. I'm in over 20 private companies. Like what the heck am I doing? I'm not a VC. But they've almost all been either friends at VC companies saying, hey, you know, we know what you like and this sounds pretty interesting. Do you wanna take a look? Or friends and ex-coworkers that's like, hey, I went out and I started this new thing, you know, can you help? Do you wanna put some money in? And so I, I think going back to GenAI, like there's, humans are social creatures and the premium on in real life I think is gonna continue to go up month over month and, and year over year. And there's no substitute for the type of trust that you and I have because we've known each other for, I think, 20+ years now. And we got a ton of face-to-face time together and this modality is not bad and it helps, but that's most of my learning is my text threads, my friends calling me up and people just like grabbing me by the ear and saying, hey, pay attention. There's something interesting happening over here.

Thank you, Doug. Listening to 10 podcasts. Is clearly one way to keep up with the trends. Maybe we should include in our show list a few of those podcasts. In your career as CEO, scaling multiple companies to billions in revenues and even creating a decacorn, can you talk about one or two key traits that you've consistently observed in successful companies? What attributes seem to be crucial for achieving that level of growth and success?

Again, probably super trite, but customer obsession, that curiosity, growth mindset, and customer obsession combination is just so important. What got me excited about being in tech started with my realization in the interview cycle, because I thought I wanted to be an investment banker, that technology liberates information and knowledge. And I've had this strong attachment of humans do better when When we're less ordered on all aspects, um, where can we wander to and what data do we have access to and what education we have access to? And it's like, oh my gosh, this is whether it's way back into business or whether it's, you know, something like Khan Academy or this is unbelievably good for humanity if I can be part of it. But then through that, I was all B2B. It really, if you don't, if you're not really hungry for solving someone's problem, like that deep empathy of. What is it like to be a network engineer? And I've never been one, so I, you know, spent a year now of, of meeting with a ton of network engineers. Like, what is your day in the life like? Oh wow, like, that's really painful. How do we make that better? But that customer obsession and genuine curiosity and empathy, and like, we are changing our whole tagline. A couple, one, the network is the cloud, which I'm— we got to keep explaining and we got to get better with videos and, and other But I love that concept of you don't want to be stranded in a data center or series of data centers. You want to use the entire landscape to your advantage. Like that's the whole point of these investments that people have made. And if you don't have networking flexibility, you can't get there. But the themes underneath that are we need to make networking cool again. It's so important to the world and those valiant network engineers that people think almost forgot existed, like they're networking heroes without them. All this cool stuff, this conversation we're having, it's not possible. The network goes down. We can't do this gosh darn podcast. So how do we give those guys, those men and women love, like they're, they've got a really hard job. They're the first people blamed. They spend all their time trying to justify it's not their issue. And then they're trying to think around corners. Like you guys think this is super easy. You don't have to worry about us. Hey, connection is hard. There's laws of physics. Networking's very complex. Like, thank God we are thinking about it so that you can actually get the way of life that you want. So that's, That is, it's kind of manifesting what we're doing here at Aviatrix. I just get so passionate about like, how do you, how do you find people that need help and then be open to think through how to really help them? And I think if you do those two things, you're going to wind up building a medium, large, or extra large company, likely not small though, unless you picked a tiny, tiny audience. But most of the audiences that we'll pick are big enough that they can feel that, that size. Last question.

Where is Doug in 5 years?

Uh, if we do things right and we get lucky and the timing is right, yeah, then hopefully I am talking to my successor about when it's time for that person to succeed me as the CEO of a public company, Aviatrix. Um, and, and hopefully at that point in time, I'll actually be ready to retire and see the last few years of my kids' high school, my youngest kid's high school, and travel a bit more and That'd be ideal, but, but in all cases, no matter what pivots and twists occur, I can't imagine not being super deeply involved still in tech. I just, it was interesting trying to take a break, just how clear it was that I'm way too excited about this industry and what it does for people and how important it is to not be deeply involved in now as an operator. And, and then we'll see what happens. With the outcome of Aviatrix and whether I eventually get to join you on your side of the table, Sid.

Well, you are too kind, Doug. And I think it's just been so amazing to see your journey and career trajectory. And I'm so happy that you took on another CEO role, that you didn't go and retire. I know we had this conversation a year and a half ago, but I'm so glad that you didn't choose to kind of move away. And I'm just so excited to see how you build out Aviatrix and Hopefully we'll have you on in a few years as you're taking the company public. I'll look forward to that as well. Thank you so much, Doug, for joining us. Thank you for joining us Inside the Network.

If you like this episode, please leave us a review and share it with others.

If you really, really liked it and you have some feedback for us, wrap it on a bottle of Yamazaki and send it to me first. No, don't do that.

Mahendra gets too many gifts already. Please reach out by email or LinkedIn.