Navigating venture capital and networking with Alan Cohen

Avi Freedman: Welcome to Network AF. This week I have on Alan Cohen, a networker who has led marketing at Cisco, at multiple innovative networking startups. In fact, he likes to sort of kneecap and put fun at the main stream and drive innovation. We're going to talk about some of his experiences, both in the networking field, marketing business, and life. Please join us, and I think you'll love the episode. Hello and welcome to Network AF, I'm here with my good friend, Alan Cohen, a rebel and recovering networker. Alan, if you could give us a little intro, what are you up to nowadays?

Alan Cohen: Morning Avi, great to see you and be on AF. Because it's definitely kind of an crosstalk.

Avi Freedman: You are AF.

Alan Cohen: AF. Usual stuff, I mean, continue to make investments in kind of companies that are completely changing in the future. I'll be talking about it a little bit soon, but I just made an investment in a networking company of a different type. But I can't go into it a lot, it's... And it's not a network with packets, so... But we'll talk more about that sometime soon. Come back after crosstalk.

Avi Freedman: crosstalk carrier. It's a inaudible.

Alan Cohen: Yeah. It's a bird. It's inaudible but apparently it's not dead, it's just sleeping. And gearing up for the holidays, whatever that means. Or gearing down.

Avi Freedman: Okay. Some travel, whatever.

Alan Cohen: Yeah. Usual.

Avi Freedman: Okay. So, you had a diverse career... We won't say long career, we'll say diverse career. In and out and around of networking from the Death Star of networking to people shaking things up and kicking things around. How did you get into tech in the first place?

Alan Cohen: I was an accidental tourist in tech. I actually started my first career actually in government, and then got an MBA. And when I got out of school, I got a job as many MBAs do in consulting. And actually, my training is actually as an economist. Obviously not engineering, which you figured out in our first meeting. So, the partner that hired me to the consulting firm, about three weeks before I got there accepted an assignment to work on the mining industry, between Singapore and Australia. And he said, " Hey, you're ready to come?" Move there. And I said, "I just got married and I'm trying to hold it together in the first year." So, he actually passed me on to the partner that ran the telecom practice, and I got thrown... In the early 90s, got thrown head long into the growth of cellular, and then broadband networks, back when we called it interactive TV.

Avi Freedman: Oh, I remember that.

Alan Cohen: Yeah. So I came up through that and then went out to US West and worked on its cellular and paging business. And also, the trials that we did in Omaha and with Time Warner, and learned that... How do I say this politely? I won't be polite. It's 4 billion dollars to find out that you don't want your phone company picking your movies. But, we had some of the first ATM networks and saw that and then kind of moved into the internet business actually, at IBM, good friend of Kentik, and built e- commerce. And I think I mentioned to you that I joined IBM on June 5th, 1995, which was the day that IBM acquired Lotus, in a last gas of client server application and paying some rents and some mortgages in Boston. And so, ran the first Windows 95 platform at IBM... I didn't use OS/ 2, with a Netscape browser. And steadily came out to the valley, worked for Cisco... Yes, Avi, I am your father. And proceeded through a great early career. Cisco in the'90s, and then for networking and security startups later. And eventually do what every washed up technologist does is make your way into venture. I mean, you take all your experience and bring it to venture.

Avi Freedman: To mentor the next generations. Although some of them aren't that next generation, some of them are just next generation of ideas. So...

Alan Cohen: To try to hold their beer, give them money and not get in their way. Yeah, you're exactly right.

Avi Freedman: Yes. Or Diet Pepsi as the case may be.

Alan Cohen: That's right. Diet Pepsi. That's right.

Avi Freedman: Or cases of Diet Pepsi, as the case may be.

Alan Cohen: And I think you know, I've had the great pleasure of causing several wars in the networking industry. One in the wireless war between the... At an air company called Aerospace, which was a start up that kind of inaudible got taken out by Cisco. And then later, spent some time at Cisco again, and then joined an enterprising network virtualization startup called Nicira, which was acquired by VMware, and started the shooting war between Cisco and VMware crosstalk kind of fanning the flames.

Avi Freedman: Yeah. Which we'll talk about Nicira, but it's interesting because, the way that SDN was originally pitched, truth in advertising would've been, people hate opening tickets to make VLANs, can we just automate the network people. And that's where we are, versus the flow controllers and magic LANs and all that stuff.

Alan Cohen: Yeah. Well, I mean, Nicira, clearly... And by the way, it's a great antecedent for any discussion with Kentik because it was the first realization that increasingly, organizations were not going to completely own every network they were running over. And that they would need software control layers to be able to provide the capabilities you would expect in an enterprise network, like security, performance, control, visibility, that you weren't going to have when you were using somebody else's network.

Avi Freedman: As you've said, the networks you own are the ones you don't.

Alan Cohen: That's right. Yeah. And so, what I would say is like, a lot of these capabilities have effectively created passports and health kits for traveling into the great unknown of the internet and the internet as an infrastructure... Infrastructure as a service players and all those folks. So, I've kind of got washed, sure in my little rowboat across all of these battles. And quite cognizant of, look, I mean, when I was at Cisco in the" 90s, it was really, two things really drove the business. The first one was email, which drove the enterprise network and then second one was the browser.

Avi Freedman: Right. Yeah. No, you definitely need the applications to drive. I mean, even originally... In fact, net access originally meant email and Usenet... For those that remember Usenet, not live internet access. Because, it was neat for CIS admins and whatever, but it was not what originally was all connected. So, were you at Cisco before startup, before wireless?

Alan Cohen: I was. I was. I worked with Cisco. I came there in the'90s, during the great push to replace service providers with over the top kinds of companies and actually worked on management as well as call managers. I remember, we worked with Groupe Videotron in Canada to build an over- the- top phone network. It was all clunky at the time, the software-

Avi Freedman: Now this is before the Cisco VoIP phone took over the world.

Alan Cohen: No. Yeah. I was there for the early VoIP and the acquisition of Selsius, which was this little white phone that... And the demos would be, you would take the phone and plug it into these different ethernet ports. It's like, " Oh, look, it still works. You can still make calls." And because it got inaudible. What was interesting is that, our architecture was very client- server. It wasn't very intranet- ish, which is why it had to be rebuilt. But clearly, organizations were... As the applications became increasingly networked, we're deploying as fast as we can, and we were hiring as fast as we can. And we had to create things. I was there with Tony Li, came up with BGP. It's like, somehow you had to make the routing thing work. We acquired StrataCom and we had kind of the slow burnout of ATM, as basically IP and ethernet, just kind of rolled everything. It was the Bitcoin of networking protocols. It kind of rolled over everything, well it actually still is, right? That's still kind of rolled over.

Avi Freedman: Yeah. Yeah. I mean, it's definitely funny. I remember the, " Oh my God, we've got an ATM pipe full of frames but no packets coming out the other end because congestion control is not synchronized." And BGP fighting the IGRPs... The Cisco... Yes. I mean crosstalk or whatever. I just saw a funny cartoon crosstalk.

Alan Cohen: Well, and then we realized like, " Oh my God, we've now shipped X millions of devices and we don't have anything that actually manages them all either."

Avi Freedman: Right. Well, that's a separate thing, is vendors and software. But-

Alan Cohen: We were selling too quick to worry about management. That was somebody else's problem, right? Yeah.

Avi Freedman: People used to think I was religious and they're like, " What IGP should I use?" And I'd be like, " Well, look at the cards you need to use, look at the release notes, find the IGP with the fewest bugs, because in the'90s it was like, stick a fitty card in and spit out the hissy port and every other bike gets dropped and you're like, " I don't even know how that could happen," but they weren't real operating systems. What was it like... I mean, that was back when the giants... You said Tony Li, the giants walked crosstalk then.

Alan Cohen: The Joel VaenI mean the people who worked in the lab, like with inaudible at Stanford, right? It was growing so quickly. There was kind of two things going on. At a pure geeky level, there was still an enormous amount of a building in creation, of protocols and capabilities, because the little router printer network they set up at Stanford was not exactly the ideal architecture for something that would manage... Something, scaled Amazon Web Services if you think about the... I always think about things in terms of LANs, right? Like you had your LAN and there's a little LAN in the Cohen household and then there's your office and it was campus architecture. And then there was wide area and data center architectures and then overlay where your experiences on the internet edge. And they were all kind of separate. And so, we were doing, I think, very rapid, incremental innovations on one hand. And then some of my colleagues, not I, were building these... Can I say, BARs, BFGs.

Avi Freedman: Yes. BFRs.

Alan Cohen: BFRs, right? Because, as applications like voice, but more importantly, video started to get layered on in the late'90s, it was-

Avi Freedman: Good freaking routers needed to pass the packets, yep.

Alan Cohen: Yeah. A lot of routers and line card updates were just not enough. So there was this massive explosion and hiring everywhere. And we were doing... I don't know, an acquisition a month, a week, couldn't tell.-

Avi Freedman: Yeah. I hired someone when I was at Akamai who had worked at Cisco in Australia. And he told the story about literally the fax machine was out of paper and they just kept... People kept dialing the fax machine to make orders. They couldn't be bothered-

Alan Cohen: Couldn't change a paper fast enough. crosstalk

Avi Freedman: Yes. They couldn't be bothered to answer the phones unless it was like a 3 million dollar order.

Alan Cohen: I was there for the ELB e- commerce here where we got rid of the fax machines and we just watched the money rolling in on the server. What was interesting also was, I can remember a management meeting probably 1998, 1999, where chambers got up on a meeting... Sharon said, " You're not hiring fast enough. We need the scale." So, I mean obviously, probably Dell saw this, Intel saw this, inaudible saw all this, you saw this at Akamai. There was such an explosion of infrastructure, just the support. Effectively, back office applications and a little communications. If you think of the perfusion of applications that the network has to support today, it's nothing. And we did not have remotely in the amount of developers like HashiCorp went public yesterday. inaudible GitLab went public two months ago and... Which, by the way, I think is fabulous for us Kentik, because what are all those developers going to do? They're going to build stuff that's got to move around on those networks and-

Avi Freedman: Right. No, it's awesome. People think, oh, well the network... It's just APIs, right? There's no people. There's no... It's like, well, automation doesn't mean simplicity. Ultimately, something needs to make it go.

Alan Cohen: Well, it's got to work. I had a license plate holder that said the network works, no excuses. That was my first license plate holder from Cisco, because didn't always work, right? And so I think, it's like every form of communications. When I worked in telecommunications at US West, we got out of the cellular business because we had a very small footprint. And people don't remember this, it was either Boz Allen or McKinsey put out a report in the early'90s, said, it'll be lucky to be a million cell phones in America."

Avi Freedman: Right. Five computers, 1 million cell phones. Yeah.

Alan Cohen: Yeah. I mean, so it's the avalanche of... I guess somebody's got to be the network provider to the metaverse man.

Avi Freedman: Yeah. It's interesting. Next week, interviewing Dave Schaeffer Cogent, an icon at last. It's partnered with Cisco. Did very well.

Alan Cohen: I remember Cogent as a customer from 20 years ago. Yeah.

Avi Freedman: Yes. No, it's amazing what they've done. And speaking of, I remember being... I think the first time I met him was at, George Gilder had a telecom conference which was largely promoting things that I wasn't... There was this concept called Storewith and stuff.

Alan Cohen: Oh, Avi, this is so bad. Before I joined Cisco, I was forced to speak at a conference called Vortex that was put on by Bob Metcalfe and the gang from Network World. This is like OG story time inaudible. Most people have probably like clicked off and they're probably trading Shibu coin right now. But IBM bought a slot to do a speaking slot and nobody would go, because we had obviously lost the networking, and I was the lowest level executive that couldn't say no. And I was supposed to speak after John Chambers and before George Gilder. And I said, that's like going on after the Stones and before the Beatles, you're really going to crosstalk

Avi Freedman: You're a pretty good story teller but even so...

Alan Cohen: Avi, I said, " They're going to go to the bathrooms. They're going to go get a hot dog. So I called up Metcalfe, I said, " Send this a quarter a million dollars back." And he goes, " What do you want to do?" And I go, " Well, put me in with my peers". So actually it was Charles Giancarlo, who's the CEO of Pure, who's the head of business development at Cisco. Odafam Ali who ran 3com's networking business and Tom Rebel who runs Siemens. And we did a panel on internet quality of service and things like that. And I ambushed them, that's how Chambers watched it and that would help me get acquired.

Avi Freedman: I remember at that conference, Gael and I, it was on Knob Hill. At the Ritz? No, the Fairmont. And there was a vintage Mercedes and the license plate said Dainu. And I was like, "Dainu, that's pretty cool." It is enough.

Alan Cohen: It is enough.

Avi Freedman: So there are some people there that were happy. I was telling people at Kentik my Bob Metcalfe story. And I have a great deal of respect for him, but Gilder sort of took my quote out of context because I don't know if you remember, Metcalfe wrote an article saying the internet was going to explode and die and routing inaudible.

Alan Cohen: I do remember that, yes.

Avi Freedman: Yeah. I said something that was much more in context, but Gilder took it as Freedman says, he's an elder statesman doing more harm than good. I think I said that statement was not constructive basically. And so then he called me a inaudible ASP brat wet behind my packets. So he can be alliterative and literate as well.

Alan Cohen: Yeah. Like, I mean, look co- inventing ether, that didn't completely suck.

Avi Freedman: I think I need to take a look at, again, someone on the Kentik random channel posted the thesis response about how crappy ethernet was, which will live forever for me.

Alan Cohen: I mean, I can do this ever. My first day at IBM, first week at IBM, I had a flight to Boston to meet the Lotus guys. And I got on online and the executive who ran IBM that work hardware division was online in front of me, was being pushed out of IBM at the time and asked me who I was and he was crosstalk.

Avi Freedman: Did they give you a bag of uppercase letters and back slashes and underscores for using notes?

Alan Cohen: No, but I kind of had it. Oh, no, they were too busy being happy about not having to fight with Microsoft any longer.

Avi Freedman: I guess. Yeah. That was true at a time.

Alan Cohen: I got into a lot of trouble. I actually may have coined the phrase Notice Lotus. I mean the Lotus people were terrific, right? I mean, 123 was the program for such a long time.

Avi Freedman: Yeah. Actually, this is, I believe the better lit fake background. On my actual background I now have VisiCalc as well as 123 on my shelf. So having used pretty much everything on the spreadsheet side that there was. So how did you go from mainstream networking? How did you get into Nicira and what they were up to?

Alan Cohen: Well, that was my second startup. So, I was very clear. I was the head of enterprise marketing and solutions. So all those people delivering those wonderful marketing messages across 24 billion of Cisco revenue were on my team. And I was absolutely paranoid about the cloud because what I think we've seen in the networking industry and the computing industry overall, is whenever there is an easier to use implementation of applications or infrastructure, and if it works pretty well, people are massively inaudible.

Avi Freedman: They're lazy.

Alan Cohen: They're going to move to it, right? I mean, ask Marc Benioff what it was like to build a better CRM system than CBOL and Oracle, right? And so, by the way, you don't have to run it, we'll run it for you.

Avi Freedman: Well, it was ASP, but this was multi-tenant. So, this is better.

Alan Cohen: Yeah. And so I wanted very much to push more into cloud and I was running into the usual resistance. " Ugh, that's for non- essential workloads and it's not enterprise secure." The things that you tell yourself right before the foundations of your house inaudible.

Avi Freedman: This was before the great automation, self- driving, closed loop automation.

Alan Cohen: We're really talking about S3 bitbuckets, right? But you could so see it and was not able to really make a lot of progress and Nicira really was the first networking software abstraction layer that would allow... I mean, effectively it was network virtualization, just what we called it and which is always like why didn't VMware have more play in the cloud. You always ask yourself. I mean, they're a great company, but for the most part they've been in enterprise company that I think they've done pretty good job in the last couple of years kind of making crosstalk.

Avi Freedman: Well and you look at it and what's AWS and IBM, first hybrid partnerships were around VMware because that's what people have.

Alan Cohen: That's what people had. So when I met Martin inaudible and Steve Melania, it was so clear, right? I mean, I went from having, I don't know, 350 people on my team to me and Rod inaudible are pumping out marketing stuff. And I said, there was 48 people, 45 people, 38 people, it was tiny company. We were not that big when we got acquired, but it was so evident. And I think... Kind of, I think the rule of networking go- to- market strategies is to use your technology in your message to allow the market to select you. So you effectively segment your market by people who so care about what you're doing and your goal is try to get to them quickly, in the most cost effective way as possible. Because when you show up one or two things have happened, they've already built a mini version internally of what you're doing, especially if they're a larger company, or they say, " Thank you."

Avi Freedman: I love that.

Avi Freedman: I love

Avi Freedman: that. Sales people hate it when they're like, " Oh, they have their own solution." I'm like, " No, no, no, that's good." They might not buy this month, but they're going to be a customer. Anyone who cares enough to have started trying to solve the problem, whatever the problem is, that's a great sign for sure.

Alan Cohen: So I did a little bit of surveying of a lot of... I mean I was in touch with at that time, a lot of very large enterprise organizations that were evaluating virtualization and cloud architectures. And I'm like, " Well, what do you think about this?" And it's like, "When is it going to be ready? Because we are ready." And interestingly enough, a lot of the core tenant that people thought virtualization was about was, I can get more processing power deployed, right? Most enterprise networks were traditionally at about 15 to 20% capacity.

Avi Freedman: Don't waste all this extra CPU on these machines.

Alan Cohen: It's bad use of capital.

Avi Freedman: It wasn't the stateless resolution. It wasn't production network saying, " Oh my God, we can be stateless and get more efficient that way." It was packing.

Alan Cohen: Yeah. Load it up, right? I mean, " Hey, we go carpool." Don't drive by yourself. But it really, that wasn't the real issue. It was really the ease of automation. They got people and networking didn't have that network automation though we never used the term at Nicira we built a network hypervisor. And it established tunnels instead of spinning up virtual servers. A very similar kind of construct. So it was very easy for VMware to see it and my very famous VMware executive told us when we got acquired and he came in, they came to talk to the team. The acquisition was announced is that, he looked at me in the eye and I remember he said, " Now Alan, VMware is the VMware of networking."

Avi Freedman: Yes. That shows you did well with category definition. Now you don't ever build a company to be bought by a company. You crosstalk to be great, sustainable, independent company that is strategically interesting.

Alan Cohen: And to be fair, we never said, and I never said we were the VMware of networking. We just said, we were the inventors of network virtualization. Which was true. I mean, there are people at IBM that would say they may have invented forms of virtualization too, but we made crosstalk.

Avi Freedman: It's been an interesting journey for me as a technologist who sees things. And when I go to trade shows, I need to de- cloak the marketing website because I can't reason about something, unless I know what it does. What value it provides and benefits and positioning. That's nice, but you could make anything sound like anything, if you're sufficiently talented and compare storage to security, to whatever, to Uber, if it's generic enough. But crosstalk.

Alan Cohen: If you and I were fused together, we'd be like Mr. Jackle and Dr inaudible.

Avi Freedman: We'd be, I mean, the Brundlefly of Supreme power in networking. Because I spent eight years desperately trying to sell the cloud vision before it's time. And that was a lesson in hindsight about needing to focus more on product marketing. When I saw Nicira and I'm like, " Oh, it's VX lens because people don't want to open tickets with network people and weight."

Alan Cohen: Well, something also happened. It shifted a lot of the inaudible. And I think it's very relevant for Kentik what was happening in the... You mean, ultimately you have to kind of go into any industry and particularly networking industry with a sense of real politic, like who's controlling things. And by building inaudible one of the catch traces that we didn't market, but trademark was the network above the hardware, because it was a virtual network, right? And basically saying you Cisco and other hardware vendors, you crosstalk.

Avi Freedman: Sounds almost biblical. The things that swarm upon the network and the things beneath. The things that live in the deep, they're the protocols that live in the deep and the packets that swarm upon the face of the network.

Alan Cohen: Oh my God. We were like that old Hebrew national commercial, we answer to a higher power. Or software, not hardware.

Avi Freedman: Matt Oko who's like... So, we usually say, one platform to rule them all and in the DNS bind them. I think he was like, "So that no packet shall be left behind and the lessons there for loss.

Alan Cohen: And so I think... But clearly the thing that was important there is we shifted the control of the control plane into the hands of the server administrator, right? Who was running VMware. Not the network and it turns out that was really interesting and obviously very attractive to VMware, right? Of course. And what's interesting now is as we move well beyond virtualized infrastructure, into multi- cloud, multi- everything, it's actually not as simple. It's actually kind of come back that networking people are kind of darn important again.

Avi Freedman: Well, because there's this thing called Covid which was not there in the before times. And people like SeekerNet, if you don't have connectivity then you can't learn, you can't work, you can't participate in society, it's become real.

Alan Cohen: You can and it has to work well for you to be effective. Hence, you get market leaders like Zoom and major players, like Equinix and security players, like Netskope while working with Kentik because the performance and visibility crosstalk.

Avi Freedman: I was just going to say, those sound familiar.

Alan Cohen: Yeah. We know those guys. I mean, whether I work with Eric at-

Avi Freedman: Cisco.

Alan Cohen: At Cisco and I work with in Equinix in the'90s and all of these players. And so as the world is more dynamic and more distributed, you need a couple of key capabilities that are going to transcend this very noisy... So how I think about it is like, in the old days you would buy a stack, the original days from IBM, you got the whole thing. Client server came, it blew that all up in the air. Hundreds of companies got spun up and then it consolidated down to like 10 to 20 major companies. Now the cloud era comes, it blows up again. You don't get hundreds of companies, you get thousands of infrastructure and application companies. And you get changes in roles with developers and code repositories like GitHub and GitLab. Obviously you have... And so it's much more dynamic and distributed. So the roles of companies like Kentik, which is why I was so excited when we met is like that world doesn't work without things like-

Avi Freedman: Well, the magic is awesome when the magic is magicing. When something needs debugging and you're flying a plane with no console or with a... I mean, I guess it's actually, Boeing does better at building consoles for their planes than most in the networking space do for their own products. But it's improving. I would say for all the vendors, it's been steadily improving but most networks... I mean, Cogent does run on a single... They're committed to a single vendor most-

Alan Cohen: No, no. I'm not worried about one company and inaudible I'm thinking about-

Avi Freedman: No, the hybrid, yeah.

Alan Cohen: I mean, if you look at the little screen on the bottom of your browser and it tells you all the places the packets are running around and Google ads inaudible and there's a lot of people in the mix and it's a little like think about like, if you wanted to be on Twitter all day, your head would explode because the random algorithm, dribble of stuff assaulting you, it's exhausting. And to increasingly, if you think about, if I was an eCommerce provider 10 years ago, or God forbid, 20 years ago, how many vendors I would be involved with. I could count on my hands, and now it's probably 10x.

Avi Freedman: Yeah. There's dozens of SaaS tracking vendors now. I mean, it's even something we do with Kentik people like, " Hey, can you track all the SaaS that I get to?" And yeah. Even we're 140 people and we probably of at least half the number of SaaS, probably more than half, more than 70 SaaS things that we use. So you got SaaS, you've got the internet to connect. We use cloud because we take telemetry from cloud and we use our own infrastructure. And we have some tunnel stuff over. I mean, just to run just about anything. I mean, I think the debate over is it hybrid or multi is dead. I mean, you're always depending on something. I mean, you're going to use an API to Twilio to send SMS. Sorry, sorry, text. SMS is old. So, how do you put that digital supply chain together and run?

Alan Cohen: Oh, by the way, that message is authenticated by another company called TeleSign it's run by Joe Burton, a friend of mine, right? Even using Twilio to make Twilio secure there's somebody else kind of looking over the shoulder. So it's this multiplication effect and you don't put the genie under... You really can't put the genie back in the bottle, so to speaking ever again. So it's on one hand the amount of stuff you can use is fairly exciting, right? And there's kind of an answer to everything. On the other hand, if your business depends on it... Look, what was? There was an outage, it was yesterday or the day before and AWS for a period of time. I know we were cited in some of the coverage on that, but the entire business world kind of took a collective migraine headache for many hours. While that was happening and it's popping up on CNBC and it's changing-

Avi Freedman: Christmas. crosstalk. It's a dangerous. Supply chain is already at the brink.

Alan Cohen: Yeah. I mean, it's like running out of the greatest toy that all the kids are waiting for. And the other thing, and obviously we know this as we learn this heavily during COVID everybody, the entire world's economy communications, the social structures are all connected through global communications networks and applications in a way that, I mean, to your point to the supply chain to the other, we really can't stop viruses from popping up because there's no way to stop global travel.

Avi Freedman: I guess it's not network politically correct to say this, but I still can't believe how well it does work, remembering how best effort worked in the'90s. And knowing what's actually underneath the web of complexity that makes-

Alan Cohen: Those DARPA guys we're pretty smart.

Avi Freedman: The thing that's amazing about the internet is not that it does break sometimes, but that it works in the first place.

Alan Cohen: Well said.

Avi Freedman: But you know what? I mean, now we've got QoS as quantity of service gets easier and easier despite the growth, when all of a sudden you can do 32 by 400 gig. I mean, it's pretty... I turned on my first a 100 gig link two months ago and I'm going to inaudible the servers in inaudible and upgrade a bunch of stuff. And I'm actually going to take this inaudible 723 BXL catalyst offline from the CEO cabinet and-

Alan Cohen: How old is that? The 7,200 it's an old box?

Avi Freedman: No, not the VXR. Not the inaudible VXR. That was'95- ish. Yeah, no, this a just a catalyst Doug's product line for a bit. Doug and inaudible for a bit. Back at crosstalk.

Alan Cohen: So you're going to give Doug and inaudible an upgrade with their new stuff.

Avi Freedman: Yes, we are.

Alan Cohen: Awesome.

Avi Freedman: I am rolling in the new stuff. In fact, I have a picture of Doug pointing to a 7513 or something at the UNIXSurplus. And it's like you can go to the warehouse and find all the things that used to be product manager for. So, VMware bought the virtualization company, bought the virtualization. How long were you at VMware?

Alan Cohen: I didn't go?

Avi Freedman: Did you walk in the building?

Alan Cohen: Oh yeah. Then I crosstalk.

Avi Freedman: It's a beautiful.

Alan Cohen: And it's great. Great cafeteria. It was really terrific. I had just spent a six year run at Cisco and trading one... Like, I mean, I had such a great time at Nicira and literally a year. I'm a little like Benjamin Button or Merlin. I'm going backwards in time where, I mean, I started my career. I actually worked for the US government, the energy department. I have worked for a phone company. I have worked for IBM. I worked at Cisco and everyone said, " Alan, you're such a startup guy, and now you're a VC." And it's like, yeah, if you think working with 300,000 people... I was like, I'm working backwards. It's just me and the dog, you heard him barking a second ago, right? So, for me I increasingly just prefer to be early in things and work with smaller teams and which is... So, after Nicira I actually wound up joining Illumio when they're... I was looking at my paperwork like a certificate number five. And it was a small team, we're in a tiny little dorm room. And to me that's always the fun part where a lot of crosstalk starts.

Avi Freedman: Was micro segmentation, but actually it was programming inaudible for network people.

Alan Cohen: Yeah. inaudible are IP tables. crosstalk. It's network security, right? It's a networking company, right? I mean, I always push Andrew a little like to, you could control the flow of traffic just the same way, right? So it could be done.

Avi Freedman: Focus is important at startups.

Alan Cohen: Well, the same kind of expansive growth that we were talking about a little while ago. So, if you think about it, like one of Alumio's well known public customers, Salesforce that's on everything. Salesforce does. And if you think about everything Salesforce has, talking to everything Salesforce has, you need a really good degree in graph theory to actually build a control plane for that, which is what they're really good at. So I'm always very excited about the ability to master massive problems through computing, because the infiltration, exfiltration, security risks, like Salesforce knows when Kentik salesperson's about to take an account away from company X, right? And maybe we not even used to work at. And so you don't want anybody knowing that information. And given the kind of hypercharged or in turbocharged cybersecurity market we're in, it is yet another clear element. We know this right. At Kentik, we're frequently used in support of people's security mission. In addition to how stuff works and how it performs.

Avi Freedman: Yeah. Especially, it's interesting, even more on the cloud because the telemetry is familiar to security people. Not just that it's net flow, but VPC flow logs are firewall logs on Microsoft and Amazon, not on Google they have permit deny. And so they look like firewall logs. And so people think about that. In the EBPF world. It's a little bit more we're foreign, although I think that EPBF is going to be huge in security before too long.

Alan Cohen: Yeah. No, look, it makes sense as the Linux operating systems get upgraded around the world companies like Silium and folks doing EPBF are going to be really, really important. By the way, Silium started by Dan Wetland, Nicira product manager. And so, I think the thing that you find out in networking industry is we're all basically family. We might fight now but we all basically been working with each other for decades. However, what's different is that the architectures that we have to support are dramatically different because when I could run everything through a box, meaning a switch or a router or a firewall or a load balancer or pick whatever flavor device there you have, I can do something about it, but when there are no more boxes and everything is spread out there, you have to really rethink the architecture of the network.

Avi Freedman: Yes, it's interesting because I think my view is more agnostic. So, as opposed to Silium, my view is EBPF on everything. That's why we've opened sourced it but to run, not needing a particular CNI, as opposed to Aviatrix, our view is people do want to do their own. And so you need a place to see all of everything that's going on with your infrastructure. And if you don't want to... We don't have a bet on a way to configure. That's someone else's business. And to be, we're agnostic crosstalk.

Alan Cohen: Well, predicting a bet is a vendor marketing strategy, right? This is the only way it's going to happen. And it turns out the world's not blue or red, it's purple, right? And you have to design for purple.

Avi Freedman: Yeah. But the good thing is it's really fascinating to see today that you can build a big enough ecosystem with the pace of cloud growth. There are people that want to be, that are happy to be in there, in that one ecosystem if you can build a good enough thing, right? And so Alumio did that and Nicira did that and it looks like inaudible is doing pretty well in their ecosystem but-

Alan Cohen: Yeah. Well, I think, and inaudible I think there's a class of... I mean, honestly, so what you did at Akamai when there was just the internet edge, right? I mean, there were class of market leaders that are emerging, that people are building their applications and their business around. And the great thing about being a strong, early mover is, you get a chance to create a sustainable franchise. I mean, there are a lot of stuff running over Kentik today, right?

Avi Freedman: Yeah. And a lot of directions we can go. We started making networks go, but what we actually do is internet scale data processing and observability is not just network. So we partner around that and there are things that we do directly around that. So, while we're tracking, again, EPBF whether it's... Yes, there's still S& MP, and I know what that stands... You still remember what that stands for?

Alan Cohen: Mm- hmm(affirmative).

Avi Freedman: And net flow and esflow and VPC flow logs and EPBF, and really up to crosstalk.

Alan Cohen: You don't think your manager, you don't think your listeners know what simple network management protocol.

Avi Freedman: No, but you were telling me about someone who had forgotten.

Alan Cohen: Oh, yes. They didn't couldn't... Well, I don't remember what I ate in the cafeteria in 1998.

Avi Freedman: I was just looking at a Twitter thread again, there was a Twitter thread of best names for the IP over Avian Carrier one. I think my favorite was Token Wing.

Alan Cohen: Token inaudible.

Avi Freedman: Yeah. That was a some good one.

Alan Cohen: Spanning free.

Avi Freedman: Yeah. So before we had met, I don't... I go to trade shows, as I mentioned to, we call it vendor torture. Try to figure out what everyone's doing. I have to say, I think I would give Alumio one of the best awards or at least for a few years in a row, best award for best demo.

Alan Cohen: Oh, well, yes. I mean, I'm very proud of that because of the visualization, right?

Avi Freedman: Yeah. Well, I'm just the script and taking questions and-

Alan Cohen: I know we rehearse the heck out of that and that Matt Glenn, who's the head of product there and I put a lot of time into it. I'm of the school, since I am not an engineer that I have to be able to do the demo and explain it. Which means anybody in the company can, and to do that you... Well, one, you actually have to know what you're talking about and be scripted. And number two, you have to visualize it in a way also known as eye candy, right? At these events to... Because people see the inaudible, they see the illumination map and they see the global view of billions of servers and packet flows. And then you can drill down. I always call that that's the universe view.

Avi Freedman: And to be fair crosstalk.

Alan Cohen: Galaxy view.

Avi Freedman: It's aggregates. It could have been traffic. It's in the state. You don't actually know-

Alan Cohen: There was traffic at one time at crosstalk.

Avi Freedman: At one time.

Alan Cohen: The line. You need traffic, we got the line. And people do that. They look at it. We did that in Wi- Fi by the way, too. That's where we came up with that paradigm at airspace, which is that the traditional management protocols, management systems were like, " Okay, I've got an access point here. Minus 85, minus DBM, is that good or bad?" And I didn't know anything. So we said, " Oh, here's a heat map. And oh, by the way, instead of having to have interpret signal noise ratio, we'll just make it green, yellow, and red for you." crosstalk.

Avi Freedman: It's interesting because we started... The original insight that led to Kentik was really that people needed, they didn't know what questions they needed to ask in advance and needed to keep all the telemetry, and be able to analyze it and make it as fast as if you were doing roll- ups. But halfway through about three years ago, we actually saw, if you can't make it approachable and simple, product leg growth is the way that people think about it in the startup industry now, you vastly limit your adoption. So, I think that's really the interesting trick is, solve the hard problems, make it look easy.

Alan Cohen: Well, I mean, and the best example of that in the last year, obviously, we have somebody at Kentik who's up for an academy award, right? The Doug, right? Yeah. I mean, Doug is basically explaining the internet to the entire world, including the New York Times and the Wall Street journal and CNN on a weekly basis, because he can actually see the macro political implications of traffic. It's like, " Oh, they must have taken crosstalk."

Avi Freedman: No one cares what exact BGP cycle times are.

Alan Cohen: They don't care and they get to say, " Hey, they're giving exams in Cuba." They don't want anybody cheating. So they shut down the internet for an hour. Burma. Yeah. I'm inaudible but I guess it's Burma again, right? Burma is about to have a coup because they just-

Avi Freedman: I know.

Alan Cohen: Just earned the traffic cam. But think about that, enough about ordering toys over the internet, just think about how people, governments and crosstalk.

Avi Freedman: I come down instead of-

Alan Cohen: Populations.

Avi Freedman: I'm definitely for the free market in many ways, but I come down on the side of human right also. Generally I'd prefer that private economy run it with the right incentives than the government, all be one government program. But yeah, I mean, we saw that with COVID too. You can't participate. You really are on the outside unless you have enough connectivity, unless it works well and definitely FCC is thinking about that. And the internet researchers we work with are thinking about that. So, well,

Alan Cohen: I mean, what's interesting is that the internet is effectively... People always say data is oil. Data is water. To me, it's interesting, I actually think it's water, it's oil. But to me, networking is electricity, right? So if you think about it, what do you need to live? You need food, water, shelter, right? Those are the basic things. Try living without electricity, right?

Avi Freedman: Well, I'm like number five short circuit. I need input.

Alan Cohen: Yeah, input, right? So, you say, man's living without electricity. Really hard to do that. And increasingly the internet's starting has really become that kind of fifth absolute essential ingredient of life. I mean, I suppose you could live without the internet, but it'd be pretty difficult or certainly a lot less convenient. And so for me, that's why I'm so excited about Kentik. I know we're not here to do a Kentik, but it is because it is touching everything and you were adjusting an explosion. I just looked at a whole bunch of IOT stuff in the last couple of months and everything's going to be connected to the network.

Avi Freedman: Oh yeah. Yeah. No, it's definitely not slowing down, but that gets to, I guess, maybe next chapter, as you said, and you're going backwards, Benjamin Burton style. When I met Matt inaudible of data collective.

Alan Cohen: My partner, by the way, for folks. I work at a VC firm called DCVC, where I work with Matt Aco, who is the founder of my firm.

Avi Freedman: Matt's superpower is explaining a founder's vision to them in the Maximalist sense. So connecting with someone crosstalk.

Alan Cohen: Most amount of words.

Avi Freedman: No. The most expansive world dominating maximum universe of their ideas. And usually founders struggle to help people explain why this is interesting, especially if it's technology. And like I said to Matt when we first met, " Should we go through the pitch deck?" He's like, " No, I understand it." I'm like, " You are so full of shit." And then in a few words he explained to me and yes definitely things don't always go as you expect. So, I guess I was not surprised when he introduced me to you and you have some of that skill too, but I am curious at a point in your life you do many things, what drew you to the VC side other than?

Alan Cohen: This is not very networking oriented, or even... What drew me... Well DCVC really was a large part of it. Having spent several decades in enterprise technology and having just work with so many amazing people, so many amazing experiences, I was kind of looking for the think different side of things. And what I recognized is that, the technology world is about 4 trillion of the economy worldwide. That's about add out of to a hundred, depending on how you count. So in our world, like I say, it's this huge giant industry, but in the scheme of things it's kind of small. And most of the world has not had its digital transformation, internet moment. Agriculture has had very little of it. Healthcare, which is certainly very interesting is and certainly we've seen this with the mRNA and vaccines is just beginning to have, it's kind of computing moment, right? I always say that God's a programmer and the human body is 10 trillion in lines of code. Supply chain, food, energy are all industries that have to transform one, because one, we're going to burn the planet down. If we do two, we're going to start revolutions because the equanimity and distribution of goods and services has to be like you can't add a couple hundred million people to the middle class every year, unless they're going to be able to have an iPhone or it's equivalent. And the ability to actually put some fingerprints on the future of other parts of the economy and the other parts of the world and do some things that would have some other legacy that would make the world safer, cleaner, and maybe more equitable for my kids was a big driver. DCVC were seed investors in inaudible. The original name of DCVC was data collective, that's what the DCCVC is. Having made a lot of bets in a lot of big data infrastructure companies, folks like Databricks and Elastic and so on and so forth. But in the last five years or plus they've really shifted it to deep tech.

Avi Freedman: Space.

Alan Cohen: So space. Two days ago we had an IPO and a company putting up a rocket in the same day, Planet Labs went public on Wednesday and the same day our other company Rocket Lab had set a bunch of satellite and electron. Which by the way uses additive manufacturing 3D printing to print its rocket engines, right? Print me a rocking engine, okay?

Avi Freedman: Right. Click.

Alan Cohen: And maybe it takes more than a minute, but the print. And so that was important. And I think just for everybody, for their brain, the ability to learn a whole bunch of stuff, particularly a whole bunch of technical stuff and how other industries work was like, couldn't believe somebody would give me a job to do that. So it was highly, highly appealing to have a different chapter in the technology industry. As I'm using myself, my feet in the third version of the internet, and now some of these other, these other industries and just amazing companies and people working on really hard problems. I think it was true in the last 50 years of tech and it'll be true in the next 50 years of tech and new tech, the really hard problems make for the most interesting companies and outcomes.

Avi Freedman: Yeah. And what is interesting as someone who's led companies is how hard it can be on the inside to explain and how critical to the success of companies it is to frame and explain. Which I was thinking about in hindsight, again, it's a super power of inaudible and it's something you're pretty good at as well. And that is from marketing perspective, right? How do you get deal flow? Being able to connect with technologists entrepreneurs and help them understand, play back their vision in a way that can be played to others.

Alan Cohen: I like to tell people I'm only a VC on Zoom Monday to Friday, 9:00 to 5: 00 because I still see myself as an operating firm by the only VC can come up with the term operator. I mean, we're people who build and run stuff, right I mean, the operator's term is some kind of tweaking term that crosstalk

Avi Freedman: Sounds more crosstalk. Sounds more crosstalk.

Alan Cohen: Operator. And my initial job at DCVC, I was an operating partner. And people don't operate businesses, they run businesses, they build products, they take care of customers. And so for me, it's also, I'm extremely fortunate, I guess, just to work on great stuff of pretty much my own choosing. And I get to spend as much time as I want to, as long as it's useful to entrepreneurs like yourself and as opposed to running around and I don't have to chase deals and things like that. You do have to do that to be successful this crosstalk. You need a deal inaudible

Avi Freedman: We have to chase or market and make aware there's different strategies for that network in some way, ultimately.

Alan Cohen: And there's also been a big change that increasingly people are looking for investors who don't just have... But interesting is capital is not as scarce as it was 10 years ago, they are looking for some folks with relevant experience or at least a mindset and not like some inaudible founder or friendly, but saying like, "Well, what can you do for me other than give me money and critique my PowerPoint?"

Avi Freedman: Right. Yeah. In 2014, when I started Kentik it was sufficient that I viewed that you would not add stress in stressful times. That you would be supportive, but not prescriptive. And we've chosen well, so generally the answer is, " Well, we can give you advice, but we're not running the company."

Alan Cohen: Yeah. Best we can be is a really good sounding board to help shape your view. By the way that's a great networking terms, Joe Pinto, who was the guy who ran customer support for the long run in 25 years, he was in the first couple hundred and 200 employees of the company, where we said, the Cardinal rule of networking is when you add something to the network, do no harm.

Avi Freedman: Yeah. That is the Cardinal rule. I don't know that that is followed. I mean, I have given a lot of shit privately to people pointing out that if what you ask vendors for is a new protocol every month, you will not get stability. If what you ask vendors for is stability, you might get stability. And you have to be careful about the pace of innovation on the core of the network. And I think that the world has come a long way from the'90s when people viewed it as, " Cisco gave me an iOS image with my initials in it, how cool am I?" To, " Oh my God, that would be the worst thing ever." If you had a vendor that needed to do that. And inaudible have led the way to single train. And I think Juniper is there too now, and Cisco's not, and rest is still head on regression testing and things that frankly, many SaaS companies, could do. So I'm thrilled, about that. So, normally at this point, I ask people about how to get into networking, but you are only an honorary and spiritual networker, but you are now off on the dark side, on the VC side. So any interesting ways if people are interested in having this intellectual eclecticism?

Alan Cohen: Well, look, I mean, I think there's multiple ways to do it. In fact there's more ways to get into networking now than there ever has been. All kidding aside, where, I mean, yes, you can join a networking company. You can join somebody who joins networks or you can be a developer.

Avi Freedman: inaudible write your own layers inaudible.

Alan Cohen: You write and you could be building an application and inevitably you're going to have to learn how the network works, right? I used to have this triangle. If folks on, maybe it's not easy to visualize. At the top of it, I called it the cybersecurity triangle, or the networking triangle, depending on who I was pitching at the day. And at the top, it was no. Like you can't do that, you can't touch my network. And then when you're going from the networking people to the application people, it was slow. Like don't break what I'm doing. But then when you got to the absolute person running the application itself, it's like, go. Like, let's get our stuff out there. And so you're always somewhere on that triangle of slow down and go, to do that. And so, we were just talking a little bit earlier about Silium or you talk about cloud stack or building something in Amazon or Azure, you have to attach network or make sure things work and the network and the speed of which applications get pushed and revved means you have to be much more cognizant. So I think there's just a lot more ways to get into it. And five years ago, I think a lot of people thought networking was just inaudible. It was like electricity. If the electrons floated the socket, they floated the socket. That's all I care about. And I think maybe it was wrong. And so I think it's extremely relevant again and actually kind of exciting. It has some really good energy.

Avi Freedman: Yeah. It's very interesting to see. There's also some challenges because there are still some concepts of networking, which you do need to understand to debug because there are enough bugs that you get into the physics homework of, what are the first principles? I think this is a vendor bug. But for networkers, it's not sufficient to understand distance vector algorithms and you need to know up to service mesh and understand into applications. So, I think it's fun, but can be hard for people to get into and sort of, as you were saying, I think that is a great way for people to get in is to actually just sit down and write things that live on the biggest decentralized infrastructure that we have, which is not Bitcoin, but is the internet,

Alan Cohen: Is that the blockchain

Avi Freedman: Blockchains are good and semi decentralized, although we'll see what way different things go in different.

Alan Cohen: Well, I mean, I think maybe should bring in some of the web 3. 0 data infrastructure people, right? Because you're going to kind of want to know if your employees are actually using their machines to do their job.

Avi Freedman: Well, I think we'll get their crosstalk.

Alan Cohen: Or they're mining.

Avi Freedman: At some point. I think there's still a lot of the, " I'm going to build a computer on the internet," things that are built by people, there's no applications. I mean, fundamentally, but now that's not to say I'm negative on it. I'm just saying, multi- billion dollar companies where there's traffic or nothing being stored or nothing being computed versus billion dollar CDNs, which actually deliver life. I'll just say there's a way to go. I'm bullish on what will be done there, but that'll be a different topic and we can get your update from the field of looking out there. So, awesome question. I'm sure that early Allen had opinions and was stubborn, but if you could go back, Bill and Ted's style. And play George Carlin and give some advice to early Alan. Any things you would... Any experiences you would have short-cutted or any crosstalk.

Alan Cohen: Well, that's bought Yahoo at the IPO inaudible Google. No, I mean crosstalk.

Avi Freedman: crosstalk bought doge.

Alan Cohen: Accepted a whole bunch of roles. I think there's kind of two things that I always think about. And I'm in very much, and this is not like hyper- relevant this conversation, I'm in a huge Shakespeare kick right now. I've been thinking a lot about it, because it's an incredible treaties on human nature. And maybe I get to a more reflective period.

Avi Freedman: That's not boring.

Alan Cohen: It's never boring. Characters are so much better than what's on Netflix. And there was a tragedy of Macbeth with... Not Macbeth. Of a fellow, and fellow was a soldier in one of the city states in Italy and he was extremely jealous of his beautiful spouse, younger spouse. And he winds up getting killed at the end. I mean, there's always blood on the floor at the end of a Shakespeare tragedy. And his final lines is that, " I wish I had loved wisely and not too well." And for a lot of people who are building careers in tech, what you really, and I think you said this a little bit, is there an application for it. My first startup was a colossal fireball. It was a company called Tahoe Networks. The product was absolutely insanely good. It would load balance 2000 sessions on a single router. It was built for the internet edge to allow traffic to come onto the network for mobile internet. And we got one thing wrong. We got the call set up right wrong. So, it turned out in Japan and Korea where the first mobile inaudible and the first mobile internet really showed up at 10 o'clock at night, the rates came down. So everybody, all the kids would like in their bed have their phones and they'd get online, right? Because it was too expensive during the day. And what we had a problem that we didn't set up fast enough and technically, but the real issue is that there was no mobile internet. There was no iPhone. So we were all dressed up, but no place to go. And it ultimately, the company got sold in Nokia, but it's not the two. And I mean, I didn't mean there was anything wrong with the people causal inaudible but we were so ready at the wrong time. And the hardest thing to work through is to kind of get your timing. And I always like, you always have to ask yourself a question if we didn't exist, what would happen? And if there's a good answer, you're working on the wrong thing.

Avi Freedman: Well, I think that if I look back, and this became clear even in the'90s with the internet bubble. Having a great idea and executing it well still requires luck. Like most good outcomes still require luck, but you can absolutely put yourself in a position to take maximum advantage of it. And you can definitely see great returns. And the other thing is, as is true with all business, right? Sometimes, statistically if you pick good ideas and good people and execute well, you will do well but you can crosstalk

Alan Cohen: But you have the time, right? I mean crosstalk.

Avi Freedman: Draw a conclusion from one thing because there's macro and sometimes you're allowed to get lucky. It's allowed to get... It happens to get unlucky.

Alan Cohen: I mean the entire front end of my career was things that didn't happen. I went to a phone company and as it got out of cellular... People did great on cellular, we got out. We put our money into interactive TV, that wasn't so good. I went to IBM, I built an e- commerce business, it was right but I should have been at Netskope not at IBM because it was not material enough what we were building for company of that size. And I mean, they built a lot. Actually, my team built into called net. commerce, which is a well known eCommerce. So, I really got... My service never worked, but my application lived for a really long time. And then I came to Cisco for the service provider binge in the'90s and all the over the top service routers the most part and all that fiber that flew around the world just splashed out. And then my first startup was a company building the mobile era, which was really smart, just two, three years before Steve jobs made that really useful for Americans. And so, I wouldn't trade it. I mean, I can't trade the experiences unless time machines really back to the future really works. But what I missed was the market demand and the matching of the technology or the solution right.

Avi Freedman: And it's something that you mentioned when VCs asked me to talk to a company or when I talked to Kentik about some of our coming forays on more general data plan. I say, " Look, when we started the company, 90% of the hundred meetings were like, ask two questions. How much does it cost was usually actually first. And then, and when will you have it?" And we've had some things that we've looked at where people didn't ask those questions and we didn't have as much pull, or at least at the time, maybe we were a little early and then those are really good questions for someone to be asked. And if that sounds interesting, I might play with it. They're being polite. That's not crosstalk.

Alan Cohen: No. And that's the lesson. Large companies will always play with your technology. They have large teams to do it. And you inevitably say, " Oh, well look, XYZ company is using it. Everybody wants it." It's like, they have a huge team and all of them have a lot of spare time on their hand. And they're paid to look at new stuff, but it's hard to get the timing right and the model right. Like Webvan burned a billion dollars, but would've been pretty good to be into Instacart or DoorDash, right?

Avi Freedman: Seems frankly small right now to own the... They only burned a billion dollars. I mean, what's wrong with them?

Alan Cohen: Yeah. No, that's true too. Never can tell us if you see like your heart inaudible lost a billion.

Avi Freedman: Well, you know that's not my style.

Alan Cohen: Yeah. And we answer to a higher authority. They're called limited partners.

Avi Freedman: I know, but it's funny because people are like, people in the family are like, " Oh, so you're profitable yet." Almost like, " Oh my God, I'd get fired." No, don't say that. Not in the next two years.

Alan Cohen: Oh yeah, no, no. I see business plans all the time. We're going to be profitable in 18 months. I go, "Well, did you run out of market that fast?"

Avi Freedman: We'll do another show and break that down for people because it is a little counterintuitive, but.

Alan Cohen: Yeah. Well, I mean, that's... Basically you're investing in growth, right? I mean Amazon built one of the largest companies in the world by having an investor base that wanted growth over profitability.

Avi Freedman: Right. Dominance.

Alan Cohen: Dominance. It really worked.

Avi Freedman: At Akamai, again, another show, but Akamai actually had to stop saying dominate crush, compete, whatever.

Alan Cohen: We were trained in that at some of my large companies experiences. You're not there, there were this little group crosstalk Washington called the Antitrust Division and you really don't want to invite them in.

Avi Freedman: Cool. Well Alan, thank you. I can see we have some future shows we can do and maybe we'll have the crypto overview in a bit.

Alan Cohen: Awesome.

Avi Freedman: It's been great working with you and thank you for sharing your wisdom and being on Network AF.

Alan Cohen: Thank you. Great to be here.