SPEAKER_03: 0:00

I'll do a quick intro and then um then we'll just chat away.

SPEAKER_00: 0:06

Okay.

SPEAKER_03: 0:11

Alrighty. Ready to roll? Well it's a real pleasure today to have David Friend, he's the co-founder and CEO of Wasabi join me today. It's an incredible cloud storage company that's disrupting the world. But David, and what we're so lucky to chat to him, he's one of these true serial entrepreneurs. It's actually the seventh company that he's founded over the time. And actually his first company I think we might even get to talk about. And most recently was a company called Carbonite, which was a storage company which he led to IPO, and then that was later acquired by OpenText for$1.4 billion. He's got some really interesting hobbies, which hopefully we'll get to see some of those things behind him. And those hobbies have also turned into some incredible philanthropy projects. So we we might get to delve into a few of these different pieces and what is an incredible story. So today let's start with by asking you a little bit about Wasabi and what is Wasabi? And how on earth did you come to found this incredible company? And we'll we'll start there and see where the conversation leads. Well, Wasabi is cloud storage.

SPEAKER_00: 1:18

So uh whatever you uh whatever you need to store, whether it's uh video from a surveillance camera, genomics data from some hospital machine, medical imaging, computer backups, uh satellite imagery. Uh you know, we have every imaginable kind of data being stored at Wasabi, including uh data lakes for AI and things of that sort.

SPEAKER_03: 1:45

So we I mean we came across Wasabi, um we being megaport because we had some incredibly large customers that were trying to literally split the atom, is what we were we were seeing. They were moving a tremendous amount of data outside of some of the traditional cloud providers and storing it in this company called Wasabi, which I hadn't actually come across at the time. And I remember asking this particular customer, why on earth are you moving out of this huge hyperscaler what was an incredibly large amount of data, uh, probably one of the largest in the on the planet, I suspect. And they were moving this data out and they were put into this company called Wasabi, and I thought, why would you do that? And their answer was it was something like a 90 something percent reduction in cost. And I at the time I didn't think that that was possible because I think, you know, storage is I think all drives come out of a similar factory, and I think you can only compress it by so much. But you guys went and did something very different and really took on the hyperscalers in a space that probably didn't make much sense to be able to be successful, and yet here I was listening to a customer uh aggressively trying to migrate to you. So that's how I first came across you. How how on earth did you build that?

SPEAKER_00: 3:02

How's that possible? You know, when I was the CEO of Carbonate, uh we were, you know, backing up literally millions of PCs every night. And so we got very good at storing data in the cloud. And uh this would have been about um 2015. My genius technical partner, Jeff Flowers, came to me and he said, uh, you know, the Amazon had had introduced the S3 storage some years ago, and Microsoft was already there with Azure Blob Storage. And he said, I have an idea that I'd been working on with our chief architect for a new storage architecture that would be faster and cheaper than what Amazon and Microsoft had designed. And uh partly this was just based on the the evolution of um of both disk drives, which had been getting higher and higher capacity, and uh and the dramatic reduction in compute costs that had gone on during the 10 years or so between uh the time we founded Wasabi and the time Amazon and Microsoft introduced their cloud storage. So I encouraged him to go off and they went and rented a little house nearby where Jeff lives and worked away on it for about a year, and uh they came back and said they had it working at a prototype level. And uh Jeff is uh a real big uh he's a very practical engineer as well as being uh kind of the Mozart of software, in my view. Um and uh I looked at the numbers and I said, holy cow, if this really is accurate, uh we could sell cloud storage that is equivalent or better than what Amazon's selling for$23 for$7 and still make very good profit margins. And uh so uh I recruited a new CEO for Carbonite and left to found Wasabi because in my view, uh storage is on almost an infinite market. Yes. And you know, it's one of those one of those things that you know is never turns around. It's always going up. And everybody stores more data each year than they did the the year before. And uh so we launched the service, uh, our first data center in Virginia in 2017, and it took us uh roughly seven months to sell our first petabyte of storage. Uh today we sell a petabyte of storage every eight or ten hours. So, you know, it it it's been uh it's been a real big ride. And one one of the things um you're in a in a capital, relatively capital-intensive business. So, you know, one of the things that uh is essential for success in a in any business like this is the ability to raise money. And because this is my my sixth startup really, and I've made money for a lot of investors along the way in my prior companies, uh, there were a lot of people standing there with their wallets open saying, I don't care what you're doing, whatever it is I want to invest. And and so we said, look, this is something that a lot of startups can't do. Yeah. And so we we went for a get big quick strategy. We raised a lot of money, we expanded quickly, so now we have data centers all over the world, uh, including in Australia, where you are, as you know. And um we have about 17,000 channel partners, so these would be bars, MSPs, and distributors because we're uh almost entirely a channel-driven uh service. Yeah. And uh and about, you know, somewhere around 110,000, 120,000 customers. That's um, uh ranging from some of the largest companies in the world down to some of the smallest. And uh, you know, you people can just go to our website and put in their credit card and sign up and start storing data in the cloud. And I made it so that you know it it followed Amazon, because Amazon still to this day has the bulk of the of the cloud storage market. But you can unplug Amazon and plug in Wasabi without changing any of your code, and everything works exactly the same. So we thought this would be the example that you gave of this very large customer that's moved to uh Wasabi. It was a you know, obviously a very simple thing to do, but moving a lot of data takes a long, take, can take a long time, even with the kind of speed that you guys offer on your networks. If you're moving a hundred petabytes of data or two hundred petabytes of data, that's a lot, you know. It sure is.

SPEAKER_03: 7:53

Well, that was actually what this particular customer needed, and they were in a whole they were in multiple cloud zones and locations all around the world. They needed uh they wanted more. I think 100 gig was what we could what we could get them out and then connecting in. And so, yeah, quite astounding uh the amount of trying. And I I suspect it was probably running for some period of time to get that much data out.

SPEAKER_00: 8:16

Uh probably was, I don't know, but it probably was months, you know, to move that much data.

SPEAKER_03: 8:22

That's the statement you had. We we sort of, you know, data's like gravity, it's so hard to move around. And when you get it, it as you just said before, it just grows by it just continues to expand. And so yeah, it's such a great business model to build out, yeah. It's sort of a net retention machine almost.

SPEAKER_00: 8:41

Yeah, we have very, very little churn. Um you know, people hardly ever leave. And uh we have uh you know a positive net recurring revenue so that even if we didn't sell one new customer, you know, we'd still see that top line grow just because on average people store more data each year than they did the prior year. Going to the channel was a great uh or was a really good decision, I think, on our part, because uh, you know, it allowed us to uh capitalize on the trusted relationships that the channel partners already have with their customers. And you know, a lot of these are big multi-billion dollar resellers who formerly would have sold uh you know a NetApp box or uh or a Dell EMC storage box. And now we've kind of trained them that you know, no matter what it is that you're do selling, whether it's a backup system or a surveillance camera or you know, anything else that generates data, just plug Wasabi in. You know, it just kind of solves the storage problem. And it's simple. And and you know, the channel is we we have some people who really understand the channel, and uh, you know, one thing they taught me is the channel is gonna sell what is easy, makes the most money for the least effort. Yes. And so, you know, we have really focused on making things simple. And one of the things that, you know, you mentioned how do you get from sort of 70 or 80 percent savings, which is what the list prices would would indicate, to 90 percent is egress. And you know, if you get a bill for storage from Amazon for S3 storage, you know, it's got line items and line items and line items. And every time you look at your data, every time you take your data out, every time you touch it to see if it's still there, you get charged. So all these little microcharges, and and we did a research project um a few months ago that showed on average the customers that we have who are also using S3, only half of their S3 bill is actually the storage. The other half is egress and and API charges and things of that sort. So when when the you look at their storage cost of$23 a terabyte, on average, people are really paying something like$45 or$46 per terabyte when you include all the other little microcharges that come along after the fact. And but our bill only has one line item on it, which is how much storage should you use. And once in a while, you know, we get somebody who abuses our free egress policy, but it doesn't happen very often. And it's and it's worth it for the simplicity and for the customer's peace of mind to know before they start how much is this going to cost? Because everybody knows how much data they have. What they don't know is how often they actually touch it. Yes. That's right. So they can their first bill from Amazon could be a nasty surprise.

SPEAKER_03: 11:55

Yes. We had this uh conversation with different customers that was, you know, I can't believe our bill. Well, it turns out we have a chatty app. I thought, what does that even mean? It just means the app is just checking or or communicating back and forward a lot, and so therefore the API calls would go up. And our world is, I mean, we we're very strong in the egress space. I mean, we we one of our products allows you to connect directly with cloud providers and reduces egress. So egress is something like nine cents per gig, I think it is, and then I think we're down, we bring it down, I think 70% reductions as well. So there's there's so many hidden charges that add up, and there's some great companies that add true value to actually solve for that. But that the the piece that you mentioned before, it's actually really hard. You said something, just to be simple, it's incredibly difficult to make products simple, and it's really, really difficult to add the second piece, which is to make it really cost effective. And so making a simple product that's cost effective, and then the last piece is you're incredibly reliable. You sort of have those three pieces. But it's it's everyone would say, yes, I just want to build a simple product. Megaport's the same, we had to keep it super simple, super automated, and had to be lower cost than what our competitors provided. But that's easier said than done when you're automating infrastructure and infrastructure at scale. I assume I don't know how much storage you have, or well, it'd be it'd be enormous, I assume.

SPEAKER_00: 13:23

Yeah, exabytes, many exabytes. Which is what? What does that even mean? Well, an exabyte, an exabyte. Well, you go gigabyte, uh, terabyte, petabyte, exabyte, so a thousand petabytes. A thousand petabytes, there you go. I don't even know how many gigabytes that exists.

SPEAKER_03: 13:42

A lot of zeros. Yes. Which which you which, as you pointed out, is is a capital intensive because to do that you have an incredible number of drives. Yep. Um so how do you do that? How do you figure out what drives to buy? Do you I mean, does it do you just simplify the drive right down and then use the software across the top, which is the magic? Is that sort of the the play or um case sort of like that?

SPEAKER_00: 14:09

The key is to get uh what's called 11 nines of durability, and that's the gold standard for data security across Amazon, Microsoft. Eleven nines of durability. And if you were to think about that, if you gave me a million files to store, statistically I would lose one file every six hundred and fifty thousand years.

unknown: 14:31

Right.

SPEAKER_00: 14:32

So eleven nines.

SPEAKER_03: 14:34

Yeah, I mean we've all heard of five nines, eleven nines.

SPEAKER_00: 14:36

Yeah. No, and that's you know, it's because people don't want to lose their data. Of course. And so it has to be highly unlikely. Now there's a lot of ways to achieve that level of I mean, that requires a lot of redundancy because a disk drive itself is nowhere near that reliable. Um, you know, a disk drive is almost sure to fail within seven years. Yes. And uh so, you know, one way to do it is you just make multiple copies of your data. So you got all your data on one drive, you got all your data on another drive, you got all your data on a third drive, and then you know, the chances of uh all three of those drives failing at the same time is very low. So that's that's kind of a brute force method of achieving uh 119s of durability, and that's in fact what people did in the early days of CloudScript. RAID, sort of thing. No, it's not different. That's that's a little different. Now, RAID actually uh by itself can do the same thing by using uh what's called striping, and you can take your data and chop it up and spread it across a large array of drives. So in our case, we use an array of 17 drives, and we you can lose any four of those drives before you lose any data. So I can have an array of 17 drives, four of them can fail at the same time, and I still won't lose data. And then you would replace those drives or it would come out of the road. That's right. If one fails, and they fail all the time. So when one fails, you you uh you actually rebuild that drive from the data that's on the ones that haven't failed.

SPEAKER_03: 16:19

Got to automatically sort of pull another drive into it, or would you physically go and replace that?

SPEAKER_00: 16:24

Well, we we actually found at Carbonite that it's better to have spares, hot spares sitting there. Yes. And um and so because we don't we don't go out to our data centers but you know once every three or four months. Um you have it actually plu already plugged in and this is available, this one's loud. It's already available in the pool. Yeah. That's right. So we have a pool of of spare drives that can be automatically turned on and rebuild uh any array. And so this it gets to be very complicated because you know you have to optimize for a lot of different things. But uh, you know, we found that one of the biggest costs in storage is the people that it takes to babysit all this gear. Absolutely. And you know, there are kind of rules of thumb in the on-prem data storage industry about how many people it takes to pet to babysit a terabyte of data. And and we actually operate at about 10% of what that sort of that norm is by by building redundancy into our equipment so that when things fail, which is inevitable, um, you know, the the software itself actually detects those failures and then brings up some redundant system to compensate for it. So it's all it's all automatic. Sorry.

SPEAKER_03: 17:49

Yeah, no worries. It's it's basically built to fail because it is going to fail. So then you build it to expect the failure and then you get it to self-heal itself, so to speak. That's right. We we have a very similar um scenario at Megaport. We we're probably different in that we're in so many, we have to, by our very nature, we have to be in thousands, we're in a thousand locations. Uh, and the only way we could do that is we can't have humans involved. If if you think of a traditional telco, for us, everything would result in a human with a screwdriver and a plug driving to the data center, finding the thing, connecting it in, putting it in the wrong spot, springing the wrong line card, and so forth. So for us, the only way we could actually do it in a small company is by automating everything. And so there's no humans that touch our network. And a bit like you, we build it to fail. We build an expectation that someone's gonna drill a hole in some fibre somewhere, and that's fine. It's totally fine because we've got another physical path in every lo in every instance. So you've done the same thing, but it sounds incredibly complicated across what is probably um a pretty dense data center location, I'll assume. Are you having trouble getting access to the space and and is there I don't think drives are a huge amount of power from memory, but is it uh what's your what is your challenge on that side, or is there one for you?

SPEAKER_00: 19:15

No, I mean the there is a premium for uh for space in data centers. Um we're we're by no means uh a big problem in terms of power because it takes three watts to spin a a hard drive, which isn't a lot. So, you know, the one Nvidia chip running at you know 400 degrees Fahrenheit or something like that, it's gonna it's gonna use a lot more power than than an array of disk drives. But it still takes floor space and it still takes connectivity and uh you know the the competition for data center space is certainly worse than it was. Um but we've we've stri uh uh created some uh more strategic relationships with uh people like IBM who have large amounts of space committed around the world uh in in some cases which they're not using. So uh you know we've we've uh we've found some creative ways to uh to keep ourselves from running out of uh data center space.

SPEAKER_03: 20:22

And let me uh can I ask I'm just asking things that come to the top of my mind that I'm curious about. Disks. When you look at a disk, there's sort of the spinning disk drives that we're I think we're all used to. Then you've also got flash or flash drive. How do you how does that how do you look at that?

SPEAKER_00: 20:40

Do you have both options or is it easier to build with disk or we just launched our first all SSD data center in San Jose. And it's many times faster than the HDD. The latency is, you know, probably one fifth or one-seventh uh of what an HDD is, because you know, if you've got an HDD, you the disk is spinning. To get data off of a disk, you have to physically wait until the data comes under the head. Yes. You know, so there's no way around that. Yeah, it's the actual speed of the drive itself or the how fast it can spin, I suppose. Right. Uh but uh our view is that uh, you know, the for 50 years the cost of disk drives has been coming down, down, down. So when we started in 2017, we were installing state-of-the-art four-terabyte disk drives. Today we're installing 22 terabyte disk drives, and you know, that'll go to 28 and 30 in probably the next year or so. So they and they still cost the same two to three hundred dollars per drive. Yes. But you're getting far more capacity. However, the the hard drive manufacturers are starting to reach some physical limits that are are slowing down that that traditional 15-20% year-over-year cost reduction. Um solid state storage, SSDs, are coming down at a much more rapid level. They're still about 2x the cost of a hard drive on a bit-for-bit basis. But within the next two or three years, that line is gonna cross. Yeah. And then we're gonna case, you know, I don't know what the hard drive manufacturers are gonna do. I think some of them have their head in the sand. But you you know that, you know, semiconductor manufacturing processes just get cheaper and cheaper and cheaper as the volumes go up. And so, you know, the people who are making uh NVMe flash um, you know, are uh they they project, you know, continued rapid decline in price. And so we're getting ready to for a world in which all these hard drives become obsolete, and that'll probably be sometime in the next five years. We'll you know, we'll probably stop buying hard drives and st and everything we do will be based on flash.

SPEAKER_03: 23:05

And is flash is flash gives you a similar I thought drives were a bit uh more reliable than flash or something, or the flash burned out a bit quicker, or it's it's never felt for that problem.

SPEAKER_00: 23:17

Flash doesn't last forever. So I mean there's only so many times you can read and write data and erase data from a flash. But uh there again, it's all software.

unknown: 23:27

Yeah.

SPEAKER_00: 23:28

So if you have enough redundancy, um it it doesn't matter. So, you know, things just kick in and things wear out. So you know, it's uh it's definitely the future. And uh and you know, AI workloads particularly need uh higher performance than than HDDs. And and our biggest two biggest customers today are both AI companies. And um, you know, they're they're always looking for better performance. And so as the world transitions from HDDs to flash drives, I think you're gonna see you know the expectations of higher performance just normally. But we want to be in front of that curve. Uh you know, we want to be the first. So we're we're introducing the first uh cloud storage, S S3 compatible cloud storage in San Jose. It's all SSD tier of storage. So much. And it'll be twenty dollars a terabyte, and Amazon's comparable product is$140 a terabyte. So we're you know, we're maintaining a huge price gap between what we're doing and what they're doing.

SPEAKER_03: 24:41

It's such an obvious thing. It's almost uh surprising that you're not I mean, you're you're big and you're growing so fast, but it's so obvious that it's uh surprised the world hasn't caught on. I mean, when I first came across you, I must admit I hadn't heard of Wasabi. So I I'm assuming that one of the hardest things, and by the way, Megaport has this problem. It's like everyone that's at every time we tell mega someone about megaport, they're like, oh my god, I can't believe you exist. Is that the same story that you have? Is it really just an education uh for folks out there?

SPEAKER_00: 25:10

It it depends where you live, right? So if you live in Boston, there's wasabi logos everywhere. If you go to the Red Sox baseball game, you know, the wasabi logo is plastered all over the baseball stadium. If you live in in uh England, you know, if you go to a Liverpool uh well, you can see I got a Liverpool football club jersey on. If you go to a Liverpool football game, uh, you know, you're gonna see the wasabi logo all over the stadium. So as we move into new markets, we have the challenge of building a brand. And um, you know, I think a lot of tech executives uh don't understand the the power of brand. Yeah. But you know, if a customer has the choice between two companies and they've heard of one and they haven't heard of the other, they're going to be predisposed to buy the one that they've heard of. This is especially true in storage, because people take storage very seriously. I mean, it's not like if I've got some kind of uh you know an analytic system for my accounting, you know, if it breaks down for a day, you know, the world doesn't end. But if you can't get to your data, or if you lose your data, it's a real problem. So you know, that that's why we spend money on brand, because uh we want to be a well-known company, and um and it's why we spend money on uh channel marketing, because the channel already has a trusted relationship with the customer. And if the channel partner says you should use Wasabi, we think Wasabi is a good company, uh you know, we're piggybacking off that trusted relationship that the channel partner already has with the customer.

SPEAKER_03: 26:51

When you store data in Wasabi, are you and I and you get the 11-9 component, do you also have multiple data center options? So I could make sure that I've got a backup in one of your data centers, and if you know that thing might be.

SPEAKER_00: 27:23

Crazy things like that. Okay, fine, you know. No problem. Put one here and put one over there a thousand miles away, and you're all set.

SPEAKER_03: 27:33

We're hearing a lot of sovereign requests as well now. So um, you know, if you think of the geopolitics that's going on, everyone's super concerned about having data in multiple locations in case there is some sort of issue with the world. So that's that's playing out for us in a number of ways. I was gonna ask you, you you said before one of your largest customers is an AI provider. Can I ask what we're seeing with sort of these NVIDIA um GPU sort of neo clouds, they're building out massive scale in huge locations and sort of they're usually servicing one or two customers in many cases. Are you also looking to sort of build out a giant would would you or do you build out like a giant storage that connects into that specifically for that customer, or have you managed to keep everything purely a cloud option versus building a bespoke sort of storage instance in their data center or whatever it might be?

SPEAKER_00: 28:31

Um we have done both. So, I mean, we have customers who have come to us and said uh, you know, we have 50 petabytes of data that we want to store, but we don't want our data intermingled with anybody else's data for security reasons, ownership reasons, or whatever. We said, fine, we'll we'll build you a complete standalone bespoke instance of Wasabi in your data center. We'll manage it for you, but it'll only have your data in it and uh you know and it'll be on your premise. Yeah. So we've done that in a number of cases.

SPEAKER_03: 29:07

Uh and it's the same self-white stack in effect.

SPEAKER_00: 29:10

It's exactly the same. And we manage it and we you know, we make sure that uh it's up and working and we monitor the performance and things like that with all of our with all of our tools. We also when we move move into a new market, for example, when we moved into Australia, um, we had an anchor tenant already lined up so that uh you know we didn't open the data center on day one and put the four rent sign out, so to speak, and and you know, nobody was there. So when when we move into a new market, we like to have a customer there who says, Will you please bring Wasabi to Italy? you know. And so for example, we had a a big telco in Italy that said, If you bring Wasabi to Italy, you know, we'll buy half the capacity on day one.

SPEAKER_03: 30:13

We're actually very similar. We're trying to double, maybe triple that, let's say. And the question is, where do you land? And the it's the easiest answer of all. You just listen to your customers. And so if they say, hey, we love what you're doing, but you're not in here or X, Y, or Z, or this country. That's right. It's a much better proposition to follow that. So w we're in, I think, 26 countries. We're trying to add a couple more at the moment. How about yourselves? Where are you how many countries do you end up in?

SPEAKER_00: 30:49

We're in 15, 15 different places right now. And um you know, we have uh two or three more slated to open next year. So it's uh we're you know, we we get pretty good coverage. Um you know, we don't have any data centers in South America yet, and we don't have a data center in India. Those are two markets where we we have a lot of customers, but you know, we'd have a lot more if we had data centers on site because a lot of companies say I don't want my data to leave the country that I'm in, you know.

SPEAKER_01: 31:24

So that's uh increasingly uh an issue.

SPEAKER_03: 31:28

Yeah. So we just recently we launched in Brazil, it was tricky, and there was a lot of uh a lot of taxes to get the infrastructure in there. You almost pay double whatever the cost is of your infrastructure to land there. So it's actually it's a difficult place to land. We've actually been really successful in Brazil though, I must say. So it was it proved to be worthwhile. And India is for us we need telco licenses and carrier licenses are tricky, and so yeah, that's that's always that's that's actually the next big uh country that we're attempting to enter into. But we've run low on the countries that are easy, you know. The rest of them have become they're kind of like in the very difficult basket from here on in. But we do have a lot of customers that really, really want us to get there. Yeah, well, that's an awesome that's it's the the company is phenomenal. I think that the deeper you go, I can't believe the 119s. Uh I mean the software, one thing I know about building infrastructure infrastructure at scale that's automated, it's the software is the key. And so when you link software and hardware together, you get a really great outcome. So I always looked at VC sort of base companies, and everyone that was building a software company, everyone used to love software. Software was so cool, and you know, anyone with hardware was very, very, you know, not cool. What's interesting is that world has just dramatically changed, and I think NVIDIA has really, really um led that charge. So all of a sudden you've got an incredibly powerful software stack, which is hard to replicate. But let's say that that was easier to replicate than uh then the hardware that you've rolled out as well. And now what you've got are two elements, a software stack that links to that hardware. So it makes you very hard to catch because if someone wants to try and build out a wasabi or do something similar, you've got to first build wicked software, and then you've got to go and get all the capital gone, you know, pay to build it and just deploy it everywhere, get the get the contracts and so forth. So I mean, it doesn't seem like there's anyone else out there doing what you're doing.

SPEAKER_00: 33:28

It's it would be very expensive to enter this business today. And as you know, uh anytime you scale something out, you know, you get problems, you see. It worked it worked when I was this big, but when I get this big, you know, all of a sudden things things are falling over. And uh, you know, so we've spent hundreds of millions of dollars on RD. And you know, if it and it takes years too, right? Because uh, you know, w on day one you think everything works, and then you know, a year later, you know, you're bigger and things start to fall apart and you realize, oh my God, this database is not gonna scale, or you know, oh my god, you know, that the the channel, the data channel is is not gonna be big enough to handle these new disk drives that are coming along. I mean, there's always something architecturally that you you know that comes along that sort of bites you at some point, and then you have to do this mad scramble to redesign some com completely redesign some part of your system to uh to accommodate you know the changes that are coming along in the tech world. So yeah, it it's uh and furthermore, you know, there then there's the the trust issue, which is you know, if a if you were to start Wasabi and you know, do you have brand permission? You know, somebody's gonna say, well, send me some customer references. Well, we don't have any yet, you know. Exactly. Wasabi has a hundred thousand of them, you know. Which one are you gonna go to? Which one are you gonna choose? That's right.

SPEAKER_03: 35:01

That's astounding that number. We um we thought we were pretty good at 2,900.

SPEAKER_00: 35:08

Well, yeah, some of ours are very small. I mean, we have our customers range over many orders of magnitude in size. From uh somebody who might be paying us, you know,$500 a year up to people who are paying us, you know, millions of dollars every month. Yeah, it's incredible.

SPEAKER_03: 35:25

It's actually incredible that you've scaled from so small to so large inside the same product set, and you just keep the pricing pretty much the same in effect to solve for it. Yeah. It's perfect. If you build the product to differentiate price at the start and you hold it there, it actually should scale, actually.

SPEAKER_00: 35:44

It does, yeah. Absolutely. Absolutely. Yeah. No, it's uh it's a good business, and uh, you know, it it took five years to kind of get over the hump, so to speak, where you know, we start to get to the scale where things work because you know, our business is a lot like real estate. You you put up an apartment building, you hang on day one, there's no tenants, the thing's losing money hand over fast. You put the forent sign out, tenants move in, and when the building gets to sixty or seventy percent occupancy, it starts to make money. And we're the same, you know. We build a data center somewhere, we got no uh tenants, so to speak, and we put the forent sign out and do some marketing and hire a few salespeople and get some channel partners, and that people start you know renting our disk drives, so to speak, sending us their bits.

SPEAKER_03: 36:37

Could you what's the thing that limits you? As in if you could is it just capital or i uh have you got enough capital that you it's just you keep a consistent supply and it's taken up appropriately? If you supplied more, do you do you end up consuming more? Is it uh I guess what what's the cons or the the the bottleneck for you?

SPEAKER_00: 37:00

Well, you know, there are are uh the the bottleneck is uh you know, geographic expansion because moving into a new market is very expensive. And the the payback on that first installation is is long. By the time you're established in a market and you're already at, let's say, a hundred petabytes and you're going to 110 petabytes or 120, you know, the payback on that for us is very fast months. So, you know, because we're we're just adding a small amount of incremental capacity. Right, and we're also amortizing all the networking gear and all the other stuff. Because on day one, only half of the money that we spend on a new data center is actually storage. The other half is all the other infrastructure, of course. Database servers and network devices and you know various security things. Uh and that you can amortize that over a much bigger installation later on.

SPEAKER_03: 38:05

Well, I think um megaports connecting to you in most countries, we see a huge um a huge push from the US, from Australia, some of some of the Europe locations. We saw this funny thing. We we where every time we land in a new country, we have this issue where no one knows who we are. So we start again. You know, we we turned up in in Spain, same thing. They're like, who are you? We don't know who you are. We turn up to Brazil, they're like, who are you? We don't know who you are. You gotta you gotta start. In fact, in Italy we start. You gotta educate. You gotta start. And it takes about a year for people to realize that that you do that you uh do you do you need to get going? Are we and what are we running out of time?

SPEAKER_00: 38:43

Uh let me just send a note to somebody.

SPEAKER_03: 38:53

All right. Just so they don't well it's probably a good spot to just quickly pivot because I was I wanted to c ask a few questions. Background-wise, I did some research on you. You're an engineer by trade. I think you uh was it electrical engineering?

SPEAKER_00: 39:09

Is that your Yeah, my my undergraduate degree actually is in music composition.

SPEAKER_03: 39:15

Of course it is.

SPEAKER_00: 39:15

When you get out of when you get I got out of college in the middle of the Vietnam War, nobody was thinking about careers. But when you get out of college with a degree in music composition, and you say, now what the hell am I gonna do? I'm not gonna survive, you know. So I went back to grad school in engineering, electrical engineering.

SPEAKER_03: 39:36

Okay. And then and then you ended up so you yeah, you obviously how on earth do you end up creating companies? That's a big change to there, and you've built lots of them. So that's I'd love you to just sort of how on earth did you do that, and then how did you do it seven times?

SPEAKER_00: 39:53

Uh well I dropped out of grad school uh to combine my interest in music and engineering and started a company called ARP ARP that was uh a big deal in uh it was the first one of the first two synthesizer companies. Okay. And uh if you know any musicians, they'll know ARP. Yeah. Um and uh so I spent ten years chasing rock bands for endorsements. And you know, I'm I don't know if I can share my screen here, but uh you should be able to, I think. Let me see. I'm not quite sure how to do it. But uh, you know, I've got pictures of me with Pete Townsend, Jimmy Page, Stevie Wonder, you know, on and all these rock stars. And I have this box of photos from when I had long hair and was doing all this. And when my kids were teenagers and thought I was just a stuffy old dad, they discovered this box and they go, geez, my dad was actually cool. So yes, you you were literally following these guys around hanging out with the and you know, uh uh my synthesizer was the voice of R2D2 and in Star Wars. It was used to communicate with this with the spaceship in close encounters of the third kind. No. So, you know, I mean it was it was a very big deal in the back in the 70s and 80s. Um so where did the it it was a lot of fun?

SPEAKER_03: 41:22

Where did it end? Where did that end up? Is that good a question?

SPEAKER_00: 41:25

We ended up selling the company to CBS musical instruments. CBS was on a on a they they wanted to turn themselves into a music conglomerate. So they bought Steinway, they bought Fender guitars, they bought Rogers Drums, they bought my company ARP, which was at the time the leading synthesizer manufacturer in the world. And they screwed them all up. Yeah. And uh they're now back all under private management. And my company got absorbed by Fender. So in the end.

SPEAKER_03: 41:56

That's fascinating. So they and and then you what did you pivot from?

SPEAKER_00: 42:01

How did you go from that to storage? Well, when I was running ARP, uh, you know, the key really was artist endorsements. I I wanted to sell a lot of synthesizers, and that means selling synthesizers to high school rock bands, not just not just the Who and Led Zeppelin and that kind of thing, because you know they only need a few, you know, they don't need a lot. Yeah, you've got to scale that thing. Yeah. But if you want the high school rock bands to buy your synthesizer, the the the big groups had to come out with hits with using your synthesizer. Then they'd you know, if if they had a hit with my synthesizer, they'd all go out and buy mine. If they had a hit with my competitors, they'd go buy that one. So I I started getting obsessed about artist endorsements. And uh uh the the key was to find a group that was on the upswing and get them to sign an endorsement before they got so big and so famous that you couldn't get to them before they wanted millions of dollars, you know. That's right. Wait for them to become famous and then be like, here it is. Right. So I taught myself to program and I created a a little app on an Apple II computer that I called Trendspotter, and I fed in data from uh record sales, radio airplay data, and things of this sort, which allowed me to do analysis and find these groups that were as they're starting to trend up. That were trending up. And you know, I learned to figure out you know which markets were the most influential. You wouldn't be surprised, California was kind of the most influential market. And uh, you know, and it was really effective. One day I had a venture capitalist in my office to try to raise money for the synthesizer business, and he wasn't interested in that. He was convinced the Japanese were going to take over that market, which he he was right about that. But he wanted to know what what I was doing with a computer on my desk because this was, you know, probably 1981, 1982, something like that. And uh so I showed him how I could whip up these charts and visualize all this data, record record sales, airplay, and he said, I'll give you a million dollars to start to seed a company to commercialize that software. And I said, Nah, you know, I'm uh but six months later, to long story short, you know, I had sold ARP to uh CBS and launched a company called Computer Pictures using this guy's million bucks. And it was a big success. So we sold the company two and a half years later to a bigger software company. And uh then, you know, my partner Jeff Flowers and I, who've been together this whole time, we said, let's do another version of that. Well, Jeff was with you way back then. Well, I had yeah, when I was just starting Computer Pictures, uh, I had read an article about Jeff in the Boston Globe because he and and another guy had kind of saved uh one of the big mini-computer manufacturers in Boston called Prime. It was a big deal back then, and they had a uh relational database product that was uh all kinds of problems, and customers were suing them, and the stock price was tanking, and the engineering department said it was going to take three years to rewrite this relational database product. And Jeff went to the CEO and said, uh, I'd like to try writing the rewriting the product myself. And so he took three other guys and they rented a little white house out in South Datic, and in six months they completely rewrote this product called Midas. What year is this? This is in the IES. This would have been 1982, I guess, 1981, 1982. And it saved the company, literally. And the Boston Globe wrote up uh a full front-page article in the business section about how these four young programmers had done what the engine or the whole engineering department with hundreds of engineers said would have taken three years. So it was one of those soul of the new machine kind of stories. And I said, wow, that's a great story. I'm reading this, you know, reading, drinking my cup of tea and reading this article. I got to call this guy and have lunch, just say, you know, see what happened. So we had, I called him up and we had lunch, and uh it turns out that instead I thought he was gonna be like a hero at Prime, and you know, they were gonna roll out the red carb and hang in millions of dollars. He made the engineering department look so bad that they blackballed him and they would not let him come back into the engineering department. So they had stuck him in marketing. And believe me, he does not want to, you don't do not want Jeff in the marketing department. And so he was pissed. And I said, Well, you know, I'm I'm doing a startup, you know, why don't you come join me in this startup? So he went back and quit the next day and came to So he came to ARP. That was the He came to the No, this was Computer Pictures after I'd sold ARP, my first software company. And he took one look at the code I had written and said, You are banned from writing software. So no longer code, you're out. He rewrote this product and uh he brought the three other guys with him, and and two of them are uh are still with me today. That's unreal. You know, we're all in our 70s and we're all working at Wasabi. That's true. And uh so it's been a great it's been a great team of uh really brilliant people. And so that's you know, that's how we we got to where we are. So we've been doing databases and you know complicated stuff for a long time.

SPEAKER_03: 47:54

That's phenomenal. That's such a good story. It's it's amazing you've managed to keep that group together. You obviously like solving complex problems because it's certainly you know, I'm sure you've got enough money to keep yourself satisfied, so it's like it's got to be driven by something else. So it's uh it's it's fun, you know.

SPEAKER_00: 48:11

I mean, I feel like uh at at this point in my career I actually uh I've done it so many times so I don't have the anxiety that I had when I was young. Yeah, and I feel like I know what I'm doing, and I'm I'm kind of like an orchestra conductor, you know. I don't have to write the software, I don't have to write the marketing campaigns. Yes. I I just provide a an environment in which everybody else can flourish.

SPEAKER_03: 48:36

And you manage to hire some great great talent along the way.

SPEAKER_00: 48:39

Yeah, and they're just that's the key, really.

SPEAKER_03: 48:42

Yeah, well that's exciting about you. You're someone I mean you you as you said before the session, you you're raising more capital, you're talking to different folks, you're and you're very calm. You're very calm in this environment. Most founders are manic. It's just you've done that plenty of times before and it's proven, so it's not a worry. Um incredible. So the other thing I was going to ask you about is uh you've got some interesting hobbies which are sort of behind you. I don't know if the audience is gonna be able to see it, but you were into some incredible gems and minerals. Uh and so uh and I know you've also donated a lot of those, or at least um we were having this chat before, and you were saying that you you know you've stored a whole bunch of stuff in your house and your wife is starting to go crazy that you've actually taken the house over with these incredible gemstones. And now, most males, when they do that, uh the answer is whatever it is that they've been storing in their house just usually has to get thrown away. But in your case, you were storing such interesting items that they've ended up in a museum. So Yeah.

SPEAKER_00: 49:44

Well, there's still millions of dollars worth of minerals in this room right behind me. When I was eight eight years old, I think I was in uh maybe fourth grade, something like that. They were building this big superhighway down through New Rochelle, New York, where I grew up. And they would uh do blasting every day to build a road, you know, a level roadbed. And I used to crawl under the fence on my way home from fourth grade and go through the construction site to pick up leftover blasting wire because I liked building electromagnets and things like that. And one day I looked up on this cliff that they had just blasted and I saw this sparkling stuff. So I kind of climbed up and looked, and I thought I had found diamonds and I was going to be rich. You know, it was actually just little quartz crystals. But uh a guy who lived a few houses down from me was a retired professor of mineralogy at NYU. And uh so I I told him about it, and he crawled under the fence. He must have been 80 at the time, but he crawled under the fence and brought his rock hammer and he showed me how to cut these quartz crystals out. And that was the beginning of my rock collection, and it has over the years kind of taken over my life. And I just here here's one that uh just arrived in the mail this morning. Uh wow. This is uh fluorite from the Rogerly mine in England, and this is a classic specimen uh of mineral from a very famous mine, very famous location. You can't quite see the beautiful green color in the camera, but Yeah, that's the inside of the rock.

SPEAKER_01: 51:29

Well, this would have been in a in a in a cave of sorts.

SPEAKER_00: 51:34

And these crystals you can see on the back is the r is just rock, right, which they would have cut out with a diamond saw. Uh but these crystals uh it's a calcium fluoride, which is uh, you know, uh a mineral kind of like salt. And in fact the crystals are cubic just like salt crystals are. And uh, you know, over time uh uh the uh calcium fluoride that was in the dirt uh essentially above these these hidden caves uh would have dissolved with rainwater to come down into uh this these these small caves and coated the inside of the of the rocks with uh with these crystals. So this would have taken millions of years to grow to this size in all likelihood. But you know, anybody who's into collecting minerals, if they saw this specimen, they would say, Oh, that's Roger Lee mine. And and there are a few other mines in the region, and the crystal color would be a little different, the size of the crystals would be a little different. So you get so that you can identify these things and and uh see them right away. But the conosphere of crystals. Yeah. And if you Google David Friend Yale, which is where I went to college, I saw that uh you'll see the David Friend Mineral Hall. Yeah. And uh it's full of gems and minerals uh that could no longer fit in my house. Some of the specimens down there are two or three thousand pounds in weight. So there there was no way I was ever going to get them into my house in the first place. So they they sat in a warehouse until we built this museum. That's amazing. Is that that's the uh is that the Peybody Museum, is it? Peabody Museum of Natural History, yeah, which is the oldest natural history museum in the U.S. And uh has a fabulous collection now. It's uh uh and you know the the these specimens have gotten very expensive over the years. I I think the uh mineral hall at Yale is insured for a hundred million dollars. Oh my god. So, you know, they were too pleased when they got the bill from the inch from Lloyd's to insure it.

SPEAKER_03: 53:49

Oh yeah, yeah. Well I'm I'm gonna let you use have them, but you've got to insure them.

SPEAKER_00: 53:54

That's right. That's incredible. So yeah, it's uh it's a marvel what comes out of the ground. And you know, when people come to my house here and they look at some of these things, they can't believe they're natural. Yeah, that they weren't carved or cut. But I I still like going digging around, you know. I'll go into these mines and be up to my knees in mud and water and find some fascinating. Yeah, go in with with some people that know what they're doing. And uh, you know, I'll I'll be going off in February to do a little exploring.

SPEAKER_03: 54:28

So this is what keeps your mind uh focused outside of work, in effect.

SPEAKER_00: 54:33

This is the it's a hobby that's uh kind of taken over. But I still keep active in music. Uh you know, tomorrow I'm I'm spending the day at Berkeley College of Music here in Boston where I'm on the board of trustees. We have we have music students living with us here in the in the house. Uh students that couldn't afford to go to Berkeley if they didn't have a free place to live. That's amazing. We have a Yeah, so we have uh these students who've lived with us over the last thirty years, they're all over the world now. And you know, when I go to when I go to Spain I can visit them, when I go to England or Turkmenistan or China or That's incredible. Whatever. Yeah, so that's a lot of fun. Yeah, you manage your kids.

SPEAKER_03: 55:23

Yeah, I bet that yeah. And they're probably all going on and doing something phenomenal in their music career if they've gone to Yale and and studied music. You don't go there.

SPEAKER_00: 55:32

Yeah, Berkeley is uh Berkeley is a special school because it's not a conservatory. You go there to learn to be a rock and roll star or a jazz musician or a studio musician or learn how to do engineering and production or music law. So it's all about the music industry, and unlike me, when they get out of college they actually can find employment.

SPEAKER_03: 55:58

That's incredible. I don't know actually how you do all of that, to be honest. It's incr it's amazing.

SPEAKER_00: 56:03

Uh I don't commute I don't waste time commuting and I don't watch TV. Yeah if you don't do those two things, it's amazing how much time frees up. Yeah, I bet.

SPEAKER_03: 56:16

Well, still you are also building a pretty phenomenal company. You're buying kit, rolling it out, look at leading a team, raising capital, doing all those pieces. I'm sure at some point you want to re-IPO again, and then you've got you've got all the gems, so I yeah, still don't know how you do it. So it is pretty incredible. Yeah, it's amazing what you've built. Look, uh thank you, thank you, Michael.

SPEAKER_00: 56:40

Yeah. Well, I I hope we can do uh you know do more business together, and uh, you know, great admiration for Megaport and what you guys have built. Uh it's a cool partnership, that's for sure.

SPEAKER_03: 56:52

And there's so many of our customers that love what you're doing, and and we we came from we're sort of cut from the same cloth in the fact that we have taken a full automation path for infrastructure. So again, it's that software and infrastructure and stitch them together. And and as long as yeah, as long as there's um people trying to get between different locations for us, then then we will always be the connector. And in your case, you're I and as you said, storage never stops growing. It really is it. I remember I remember thinking about Snowflake um and some of the database companies. I just remember thinking, you know, MongoDBs and these sorts of basically as soon as they had one customer, it was like this instantaneous net retention. You know, Snowflake for years had something like 180% growth of their existing base. That's in almost no churn, but just it just grew 80% per annum off existing customers. No one could even help the growth, yeah. So it's such a good business, but it's it's very difficult, I think, to to solve for. So yeah, congrats on what you've built. We're we're excited about continuing to take you to our our customers and allowing people to connect to you. And yeah, I don't know how you've done what you've done, but you you're you're an inspiration. Absolutely uh amazing.

SPEAKER_00: 58:05

So it's really good to chat. Network and storage go together. You know, if you have a storage and and no network, there's no way to get the data to us. So we depend on companies like Megaport.

SPEAKER_03: 58:17

Correct. Uh you're a legend. I tell you what, I'll I'll stop the recording and then it should do its upload.