Metromont: The Concrete Behind the Cloud

The Empire of Lights is brought to you by Comcast and Babst Calland.

Transcript:

Jonathan: It's the Empire of Lights, and we are kicking this series off, having a great time here with this new series exploring the whole supply chain, everything behind what is powering and building data centers. And we couldn't do this conversation without our friends at Comcast, and we're starting it off with my guy here, Marshall Sorensen from Metromont.

And today we're gonna have a great conversation because you are so passionate about what Metromont does, and what you guys do is so unique. We don't often think about Precast concrete when we're thinking about- You should ... data centers. We should. I mean, that's what we're gonna find out today.

, You think of a data center, you're thinking of servers, you're thinking of blinking lights. You're thinking of e- all that stuff around it, but you're not realizing that something has to house all that. Yeah, yep. And it has to be something very strong. I can't think of anything much stronger than precast concrete. 'Cause it is done [00:01:00] offsite and then trucked in. Yeah, yeah. Which to me is just really cool because we get to actually think about the nuts and bolts and the growth of this is just being driven everywhere. Everyone right now is thinking about Western Pennsylvania- 'cause that's where we are and we're seeing- Yep ... lots of that, but it's everywhere. It's everywhere. Metromont is, is- Yeah ... they're everywhere with this. Yeah. So Marshall, thanks for hanging out with me today- Thanks for having me ... on our Empire Lights. Yeah. 'Cause this is just gonna be a great conversation today.

Looking forward to it. I'm gonna learn a ton. So- ... first and foremost, tell me about Marshall, 'cause I said you, you are a passionate guy. Yeah. Well, let, let me tell you about Metromont. Yeah. So Me- Metromont is a manufacturer of precast, pre-stress concrete. Yeah. And been around for 100 years, so we've figured a lot of stuff out.

We operate in a lot of markets, all the way from as recently as last December, the Texas market, which we acquired an operation- Okay ... through the Carolinas- .. through Virginia and down into South Florida. So we're in a lot of places. And, and we're, we're students of the market. We, we don't just say, "Okay, we're gonna make this thing," and, "Do you want some or not?"

No we listen to the will of our customers and they tell us what they need, and then we go do it. And we'll go wherever they tell us to go, wherever the competition [00:02:00] doesn't prevent us from going, frankly. Absolutely. Yeah. Okay. So our attitude's a little different in that regard. I, personally, I lead a very small but enthusiastic group of folks in the Northern Virginia operation, primarily in the development of data center pursuits, but not just data centers.

We're primarily a parking structure company. We've done thousands of parking structures. We've done warehouses. We've done all kinds of things. Yeah. My personal passion, as you pointed out, is data centers. Yes. Just because it's so different than where I began in this market a long time ago.

A long time ago. How long ago was that, Marshall? Well, 1988. Oh my goodness. So I've been, yeah, I've been doing this a long time. You know your stuff. Yeah, yeah. That's what, that's why you're here. Yeah. So it's fascinating to me because I got a perspective that a lot of folks don't have. So that's what we do.

And by the way, we, we've done about 200 data center projects, successfully completed 200. That's incredible. And of those, about half are what I refer to as all precast concrete data centers. And, and all precast is kind of a nominal description of everything above the foundations is precast concrete.

[00:03:00] Gotcha. Do we do the janitor's closets, okay, yeah. And yeah, no, we don't do. Yeah. But, but everything in the superstructure of the building. Right. The, the columns, the beams, the double T floor and the roofs the stairwells, the elevator shafts- Yeah ... the pre-ga, it's all precast. The hardcore infrastructure that, that, that, that goes on there.

Yeah. The stuff that really, really matters- Exactly ... it has to go up first so that everything else can follow behind us. Gotcha. So that's, that's what we do. And of those 200 projects about half have been all precast- Okay ... buildings. And that, that trend is becoming more pronounced just because of some market dynamics that I'm gonna talk to here in in a little bit.

Absolutely. But, but the big picture is that we can go really, really fast if you do some things. Speed, and speed to market is- It's, it's- ... very key in this industry. It's it's- why is precast concrete so fast? Well, because and, and I don't wanna speak for the cloud infrastructure folks, but, Sure

or mention specific names, but there are AI companies, as an example , their revenue depends on having IT power. And IT power depends on the, on how much of it they can produce, which leads to the need for us to, to [00:04:00] deploy quickly and then build very fast. So the more- Definitely

stuff, the more data centers that you can have, the more money you can make because you're producing electrons- Exactly, yes ... faster. And, and that, that is, like, an exponential thing. It's not a straight line. It's, it's really transformational. Okay. And I'll give you, give you an example. If you look at a zettabyte- Okay ... - z-E-T-A byte. How big is that? It, it- That's gotta be pretty massive ... it, it's, I had to use my hands, but I- Yeah ... yeah, yeah 'Cause I'm thinking of terabytes. How many terabytes is a zettabyte?

It's, it's a, it's a billion terabytes. Oh, my goodness. You'll find out that, that the world crossed the zettabyte threshold in about 2012, depending on who's doing the calculating. Okay. And that's, like, all the data in the world and things like that.

In all the world. Okay, gotcha. Yeah. Well, it's, look at what it is today. It's much, much higher than that, and the rate of change is nearly exponential. Well, guess what? To synthesize and use all that, the, those zettabytes, you gotta have data centers. Data centers are just the refineries for all this data that's being created.

But- That's interesting you bring that up. I never heard it put in that way, as a refinery. Well, it's exactly- Very, yeah ... it's exactly what we're- That makes perfect sense. That's exactly what it is, yeah ... if you, if you look at all of the unique data, and, and I'm as guilty as anybody. [00:05:00] You send, five little megabyte video clips of your tuna sandwich to your kids.

I've done that, okay? Tuna sandwich? That must be a heck of a tuna sandwich we're eating there. Well, that takes up IT space. Somebody's gotta synthesize that. Yes. So that data is in the cloud somewhere, okay? Just sitting there. Well, it turns out that all that unique data is just sitting there. About 90% of the, of the unique data that's been created since the dawn of man is just sitting somewhere.

It's not being used. So what AI is doing is it's using that stuff. So it's like having 90% of the earth's oil undiscovered. It's just in the ground. You haven't done anything with it. Exactly right. Okay. And so data centers are the refineries of- They bring it up and a- Yep ... we can actually use it, process it, for lack of a better term, and make it actionable.

And, and, and- So we can look at that tuna fish sandwich- Ex- ... and say, "Man, that's pretty amazing." I think it, I think it's pretty amazing. Hey, Dad. Yeah, but Bob Bob- You're making me hungry now here, Marshall, so I'll, I'll give you an example. I, and again, I'm a student of the market. So if you go back in time and you look at, I look at inflection points- Sure

things where great change happens instead of just normal change. Yeah. And the first of those [00:06:00] inflection points, I think, is when the, the subprime meltdown happened, in 2008, 2009. Do you remember that? Absolutely. It, it was painful, yes. And for everybody in- including us And, and so we were looking for other things to do than the traditional things that we'd done up to now, and, and the data center business was one of those.

I'm from Northern Virginia, right? I operate for the last 20 years Northern Virginia. And if in case our, our, our viewers here don't know, I mean, Northern Virginia is probably one of the most heavily dense- Oh, it's, it's- ... data center places in the world ... it's ri- it's ridiculous. It is like one- There's more-

For a lot of reasons I can't- Right ... go into here. But so anyway, we became interested in the data center business back in 2008 because we had to. We had- ... to find other stuff to do. And people came to us and said, "Well, why don't you do this data center?

It's like a parking garage. It just doesn't have a ramp." And slots for servers. , And, we're not the brightest bulbs on the circuit, but we said, "Okay." "Okay." "Well, we can do that." "We can do that." Sure. "Yeah, we can do that. Sure." And so we did, one data center, and then we went back to our day jobs and did other things.

And then all of a sudden it was two data centers, in 14 months or 18 months going, "We're-" Maybe there's a trend going on [00:07:00] here. Maybe this is a thing. Yeah. Right. Okay. So we started we started thinking about it that way as, as, hey, maybe this can become a regular part of what we do. And which is different than a most most other, Yeah

folks did. Then all of a sudden in about 2018 or so, there was a, what I, in my mind, another inflection point were the hyperscalers. The rate of change was, was exponential. Their need for- Interesting. Okay ... for huge amounts of compute power and gigantic buildings at a rate that they needed those things was vastly different than at points before that.

And we noticed that. Had a client come to us and said, "Hey, can you, can you get rid of columns in my white space where the servers are?"

And I said, "Yeah, we can do that. We'll just make different columns, different beam, different stuff. It's no big deal. Why is that important?" 'Cause I'm always interested in why people want- Yeah, of course. Yeah. Hey, there's a need here. Why? Because other people are gonna have that need too, right? . And so the guy, the guy tells me, and this is old news I don't have any misgivings about saying it, but he told me, I'm not gonna say who it was- Sure

but that each of those columns had the economic value, and this goes [00:08:00] back a ways. This is almost 10 years ago. Okay. He said each of those columns has an economic value of about a half million dollars. Back up. That is Yeah, I mean, that's a- Over the life of the building ... that's a, that's a place you cannot put a server.

So I like- Right ... "Wow, that's really cool." Wow. So we connected up those dots, and we invested a truckload of money in developing parts and pieces that got rid of columns in the white spaces, 'cause that's what we were- You get more space. More space You can optimize your space. Absolutely. So we just did it.

People, people said what they... We listened to the will of our customers. Okay. They told us what they wanted, and we figured out a way to do it, and we did it. Yeah. Yeah. That's how business operates, man. I li- That's how it should operate ... I I like hearing that. Yeah. That's right. Absolutely. So that was what I refer to as the second inflection point. So keep in mind the first two, the third is by far the most impactful, at least in my humble opinion, and that is the the... And I'm gonna call it the adoption rate of- Okay ... of artificial intelligence. Right. Artificial intelligence has been around a long time, okay? It's by, by itself it's not really a new thing, but the rate that it's being adopted is a new thing.

It's on your phone now. It's on your phone now. Anyone can access it. Anyone can, yeah. Before it was just an elite few or [00:09:00] scientists that could grab onto it- Ex- ex- exactly ... and actually use it. Right. Now, now, I, even I could use it but if you're a lawyer, let's say, and, and you had before Judge Chiselowitz, let's just say some litigation. In the old days before AI, generative AI, , , you would send a couple people to the libraries, and they'd crack open books, and they would study those books for weeks- Exactly

to understand what Judge Chiselowitz had, how he had ruled, or she, whatever the case may be, on a particular matter in the past, 'cause that would flavor your approach to the litigation, okay? Well, it would take a lot of time, a lot of people, and a lot of money to do that. Now, you, you mash a button on your phone and you say, "Give me the results of every decision by Judge Chiselowitz in the last two years on this matter before his court," and in t- 20 seconds- You got 'em all right there

you got it. So imagine the downstream effect that that type of thing has on- Exactly ... almost everything you can think of. Right. Now, is AI gonna replace your plumber? No. 'Cause you, you need a real human person to get up there and- Well, for now. ... contort. Yeah, yeah, for now. Yeah. But, but is it gonna replace [00:10:00] call centers?

Oh, yeah. Absolutely. Bye-bye call centers. Is it, is it gonna, is it gonna make almost everything you can think of more efficient? Absolutely. So whoever, and it's been said, whoever controls AI, specifically generative AI or even agentic AI, where you've got a digital teammate- ... is gonna control the world economy, which is a pretty scary and- yeah

and a fascinating thing at the same time. That's pretty, pretty heavy-duty stuff right there. Absolutely. So that's really, that's a pretty bold idea. And so the way that you dominate AI is that you have a lot of data centers, because if you have data centers, you can refine all this data.

You can process all the, all the inquiries- Absolutely ... all the data that's being mined to do that. So it's not normal straight line growth, it is exponential growth, okay? According to us. Yeah. So how do we know that? Because we get c- we get calls from people all the time who ask us to help out with their challenge in this AI space.

So let me talk about challenges right now. Now, in, at, at a high level view, challenges are really, in, in my opinion, , they're in four buckets. Okay. There's [00:11:00] power use, there's water use, there's land use- ... and there's people. Yeah. The first three of those four things you can fix.

You, you just get smart people in a room, and you close the door, and you set them a deadline. And you say, "You, you cannot leave until you figure these things out." E- exactly, yes. Okay? It's just, it's just, it- It's solvable, yeah ... it's solvable. You can , power Homer City Pennsylvania. They, they, they're- Yes, they put their own power on site there and- Put their own power Yes

it's gonna be gas- Ready to go ... gas power. It's y- they're just gonna do it, okay? SMRs are a big thing. You're gonna see commercially viable- Small nuclear reactors, yeah ... small nuclear reactors. They're, they are, they're already out there now for private companies. And there's a three-hour discussion you could have about that.

Yes. So power's gonna get fixed. Water is an easily fixable thing. You don't have to use vast amounts of water to cool these data centers. You can, you can totally- The new technology is amazing. You just reci- You recycle it ... you can circulate, recycle it. But water use can, can be dealt with. It's an easy thing. You don't have to use oceans of water. In fact, most data centers, most data center designs now could be designed that you use less water than a golf course uses.

[00:12:00] Really? Or that you fill a community swimming pool with. That's interesting. People don't know that. Yeah. They, they think that you have to use, ocean. No, you don't have to. There are some data centers that use water less than 5% of the year. Okay. Okay? So that's an easily fixable thing. Land use i- is a big deal.

But you don't you don't have to build a data center right next to a school or a neighborhood. You don't have to do that. You can build it in an area that you'd be, you'd struggle to find unless you got lost driving around. It can be in- Exactly, right. Yeah, all good points. Yes. It's just something that you, you do.

But people is the, is in my opinion, people is the, is the hardest thing 'cause it takes 18 years, more or less, to grow them. Excuse me. And, and- Is that what it is these days? Yeah. Yeah, or 17 or, or 30, whatever. 30, it depends. Yeah. But, but it take, it takes a long time to get them before they can really do much o- of anything.

You don't just snap your fingers and get people. And oh, by the way, if you're a young person today and you're not sure whether you wanna go to college, whether you wanna be in the trades- Right ... or whatever, become an electrician. Yes. You will make [00:13:00] truckloads- Of money. ... with a capital T of money. And then you take that money, in my opinion, just my opinion, and you go, you stack it up somewhere and watch it grow for the next 30 or 40 years, and then you go do something else.

But if you're looking for something to do, become an electrician. They're gonna be in demand even more than you could ever know with AI and all of that, yes. A data center really is, and I, I oversimplify things, but a data center is, is really just a gigantic computer with a million wires coming off of it, and somebody's gotta connect all that stuff up.

There's a lot of power coming in. Yeah, a lot, a lot of power, yeah, yeah. But not so much water, exactly. So you got, you got people, water, land and power. Those are the challenges that the industry is wrestling with. Three of those four are fixable. You can fix those. The fourth fix is harder to do, but it'll get fixed as well.

Yeah. It's just gonna take longer- Absolutely ... and be more difficult. So here's the challenge, apart from those, those typical challenges. The challenge is procurement From my perspective. Yeah. Tell me more about that. Okay Everybody loves procurement people said no one ever. Just so you know.

And, and what you have to [00:14:00] do, see, th- there's, there's processes for accounting and for safety and for manufacturing and all these processes, but there's no process for deal-making, which astonishes me. That is where everything starts. And so you've gotta find a way to go from almost nothing to a coherent and durable deal with, with off-ramps along the way that benefits everybody.

And, and a deal is only a good deal if everybody benefits from it, okay? That's just Newton's seventh law. I, I made that up, but you- We can all agree to that. Yeah Yeah. You get the, you get the idea. So here's what h- here's what happens. . There's very few companies that control most of the data, and, and that trend is, is intensifying, by the way. Yes. Very much whether that's 10 years, 20 years, in my opinion- ... there'll be three or four companies that have all the data.

All the data. Okay? We're not there yet. So these, these firms go to, they go really to two places to build data centers. They go to the REIT community. Say you're a, a real estate investment trust person. And they'll go to you between 30 or [00:15:00] 70% of the time, depending on their business model, and then they'll use their prototypical designs that they've developed, and they can deploy on their own.

So they'll use those two mechanisms to build data centers. And it, and it's a situational thing, depending on where in the country they are. So the problem is that, or the challenge is that they don't really know what they want or how much they want. They just know they want blinking lights somewhere-

So what happens is I'll just and I'm not gonna name names, but say, say you're a REIT, well, 10 years ago, there were only three or four or five REITs that you had to beat to get these deals.

Now there's dozens of them. Yeah. Makes sense. So you're, you're locked in mortal combat with these other REITs trying to get this deal, and the deal is a hy- a, a cloud infrastructure company calls on a Friday afternoon. I'm making this up. 3:15, he say, "Hey, I need blinking lights in," whatever, "10 months, nine months, six months."

Six months. Six months. And you wanna say yes. Because these are gigantic deals, and you want these deals, so you [00:16:00] wanna say yes. But you can't say yes unless you've got partners who know what they're doing, that have been meaningfully engaged- And I, yeah ... early on with you. And I feel like this is where a partner with Metro Mont comes in because with- Absolutely

because with precast you can build so much faster, so that time to market- There's, there's really three- Yeah ... there's three, three... Thank you for mentioning. Okay. Yeah. There's three things. There's, there's the deployment speed. I'm gonna talk about how you can condense that. Surely. There's the construction speed, 'cause we can just build stuff a lot faster- Sure

with a lot fewer people on the site. And then there's getting it right. I... And I, and hear, I hear people talk about early, early engagement. Which to me, I, I hear, I hear phrases that are similar to that, like, spot on, which I particularly dislike. Everybody says, "Oh, yeah, early engagement is, is a good thing," but they don't do it.

Okay, maybe they, they, they make a 10-minute phone call on a Friday afternoon to somebody in the business, four months before the, the project, and they call that early engagement. Well, that's not early engagement, okay? That's a phone [00:17:00] call. That's a 10-minute phone call. Okay. So what we do, to get all the way back to where your, your questioning was, what we do and what's very different is we do test fits.

Yes. Talk about the test fits. So this is very unique, I think, in the marketplace. It, it- It's something that I think really sets you apart, and the idea, yeah, that you can actually show how this is gonna go together- Yep. Yep ... and see where it's gonna work, or maybe not work, I guess. Or maybe not work. Yeah.

Exactly. Yeah. Maybe, yeah. So run me through a scenario on, on a test fit. Yeah. And, and simple thing. I'm always astonished that people overlook simple things. Simple things are the most powerful, if you ask me. That's... Those are the things that tend to get overlooked the most, I think- Yeah ... many times, knowing that most designers and owners and developers just don't know what we know. They don't know pr- this whole precast thing is, is not a thing that they see much of or under- unless you're in Northern Virginia or other parts of the country so they don't know. It's not their fault that they don't know Yeah, it's just they're not thinking of precast as something you're gonna build something out of.

Yeah. But it's- It, it's our fault because we haven't, we haven't done test fits for them. Right. We haven't illuminated what, what arrows could be in their- Exactly ... in their quiver. It's not their fault, it's [00:18:00] our fault. So that's why... That's one of the many reasons that we really care about test fits. But the, the, the main purpose of the test fit study, which, which by the way, takes about three, maybe four weeks to do, to do it right.

And it's not easy, which of course is why most people don't do it, because human nature being what it is. Yeah. Well, they want the easiest path in sometimes. That's right. They want that 10-minute phone call. But you could be leaving a lot of money on the table, though, without spending some time on that.

Well, that's, that's exactly right. And I'll give you an example. It's kinda, it's kind of, simple, but there's something that you do better than anybody else, okay? And pretty much everybody that I've had the good fortune of working with in this industry, the data center industry, there's something that they do, that they do better than anybody else.

Sometimes several things. you have to get all those smart people in a room at a table months and months before the green flag drops, because if you don't, you're never gonna get their best things. You may think you're gonna get, but you don't have a prayer of getting their best things.

Very good point. Yes. And and if you don't do that, all you're gonna get is you're gonna get an apparent low bid. [00:19:00] An apparent... I like that. The apparent low bid. That's- Yeah, which just fascinates me, because you, you'll get, you'll get some, some preconstruction person who's, who's got a talent for Excel spreadsheets, who's very, very smart.

Like my ki- all three of my kids are Virginia Tech, but they could make they could make Excel spreadsheets that would cause blood to shoot out of your eyes, that's quite the spreadsheet, man. Good Lord. So what happens often is that this poor preconstruction person who hasn't gotten people in a room and gotten their best things, they've simply...

They've collected bits and pieces of information from a dozen disparate subs. In a shotgun marriage, they put it all in their nifty spreadsheet, and they've created a number, and they say, "Well, look, this, this number's lower than... This apparent is a low bid." It's a great bid. Let's go with it. It's a great bid.

Yeah. Let's go with that. Well, you have no idea, A, if it's real. If it, if it really is true. Okay. A. B, you have no idea if it's as good as it could have been if you had everybody in the same room with their best things- Sure ... or lesser from what. So it's, it's wrong on two fronts, yet that is primarily what [00:20:00] happens.

So the test fit s- smashes through all that. You, you've got to get people to do a test fit study. So the test fit's super important. So our process and this is not, you know- The test fit ... the test fit. Yes. Yeah, it's not rocket sci- Anybody can do this. It's not it, it's really nothing special.

It's just the will to get the right answers. That, that's all it is. Okay. And, and the will to explain it to people so that they allot the time in their decision-making process to do a proper test fit. Most importantly, you're getting all the right brains in the room at the right time, so you're getting them at their best.

So let, let's say you're, you're a data center developer. You're new to Southwest Pennsylvania, and- I like being a developer. It's a good job. Yeah. Yeah. Yeah. Very cool. And you have no idea. All you know is that you, you've got a bunch of people that want your electrons, but you don't know, you don't have any idea, really.

And I'm not making fun of developers. No. I'm just saying you, you just don't know all of your options. Right. Okay? So you call a knucklehead like me and you say, "Hey, I want you to do a test fit of the building I think I wanna build." And I'll say yes. But I'm gonna say to you, "You need, to ensure that you're gonna [00:21:00] give me about three or four weeks to do it right, and you're gonna allow me to have access to your AE team.

And if you don't have an AE team, we'll bring one to you." All right. Okay? We'll, we'll, we'll, we, we are- I like this. We're- Okay. Yeah. Hey, you don't got it, we know one. We're gonna give you information that you can use to make a directional decision, that I'm gonna build a building this way or I'm not gonna build it that way.

That's awesome. And by the way, we can do them without columns. Yeah. We're gonna get col- we're gonna get columns out of your white space- Exactly right ... so you can make more money. They're gonna be like, "What?" So we have that conversation up front. I describe to them the, the elements of the test fit study, which are we're gonna create this basis of design.

Okay? Again, it may not have pinpoint precision, but it'll be close 'cause, this ain't our first rodeo. We know what we're doing. So the difference between 225 PSF, 250 is not so great that it would defeat the framing plan that we've got. Now, if it was 700 versus 250, that, that would be different, but- Little difference there.

A factor of three. Yeah, no. Yeah ... , but we know a lot. Yeah. And we can operate within, the gray area a little bit. So we'll create a basis design. We'll say, "Okay, here are the jurisdictional requirements. This is the wind." [00:22:00] I may ask you for a geotechnical report so I can know about seismicity.

Of course. Yeah Make sure they're at the ground cheeks, that- Exactly. So we'll collect that basic information and you'll tell me about loads and the, how long the building is and how tall, and all that stuff. And then I'll go to our estimators who are really, really good, and I'll say, "Okay, this is the criteria.

Go draw the building." And they'll, they'll create a model of the building, a 3D BIM. Amazing. Yeah. A- and it'll be of all the elevations, building sections- Okay ... connection concepts. You can actually, you can see this as it- Yeah ... as it would look and probably as it would behave- Yeah, yeah ... and function as well too.

So here's the building. Yeah. There's no- Wow ... there's no mystery. There's no six months later, "Oh, I thought you were gonna..." No, no. Here it is. This is the building. It didn't change. Here's what we're gonna do. So we create the model. In our system, the model propels the estimating process.

The estimating process is what tells you how many pieces there are- Okay ... what type, how much they weigh, everything you need to know which are important for other reasons I'll get to in a little bit. So we create the model. We do an estimate so that we can have a, a reliable indicative price. Is it the final price?

No, [00:23:00] because the building's gonna change. So- It's getting you in the ballpark, though. I mean- Yeah, but it's gonna be really close. You could- It's, yeah ... you can make a directional decision based on the, that information. Exactly. Right. People do that all the time. And then we create a mechanism where the price can go up or it can go down based on changes in the design.

So we do the model. We do the estimate. We create a schedule. We create what we call a block plan Which is really important. Okay. A block plan- I feel like that's important. That sounds very foundational to me. Well, it's, it's, again, I like, I like simple things. What is a block plan? It's simply a depiction of the way that we're gonna build this building with our cranes.

Here's where our cranes are gonna start. Here's the direction they're gonna go. These are how many pieces they're gonna erect within certain block of time. Here's when your follow-on trades can get in and work safely- Yeah ... whether it's cast in place, what- whatever it is. Which by the way is a unique advantage in our system because you don't have to wait for the whole building to be up and stabilized for people to get in.

We're stable as we go. So our- the follow-on trades can, can have access to building very quickly. And they c- they can be laying foundations while you're actually pre, while you're actually casting- There, there's a lot of over- ... [00:24:00] parts ... overlapping activities that could be done. Yeah. Once again, saving that time.

Compressing that overall schedule. Yeah. Because again, , you being the developer, you want, you've been told that you have to have blinking lights and whatever in, nine months, okay? And you said yes to that. I did? Yes. You did. I did, yeah. You did say yes. I remember that now, yeah. Yeah. And, and oh by the way- It was a busy day

every, every week on average that a big building, say 250, 300 megawatt building, is delayed, has an economic value of about $300 million. Wow. So there's gigantic stakes. That is- Yeah ... that is massive. So you don't wanna, you don't wanna be late. Okay. So you're gonna, you wanna involve people. When you said yes, okay?

So you better have folks that know what they're doing can, can get to that target you've been given. So we create this model. We do an estimate. We do a block plan so you can understand the flow of the work and when your trades can get on. We study site logistics. I'm an old school guy, so site logistics is a very big thing to me it's a ballet. You've got some choreography. You've gotta do all that safely, which is the most important thing. Of course. Yeah. Safety's always the top [00:25:00] concern. Absolutely. 100%. And so you've gotta know how you're gonna do the work, where you're gonna start, where you're gonna end.

You've gotta know where you're gonna put stage trailers, where the cranes are gonna go, how much bearing capacity you need under the crane. You gotta figure all that out. I, I have people tell me all the time, "Oh, no, we'll figure that out later." No. What? No, you figure it out now. No, no. We gotta figure it out now.

Yeah, before anything- We're, we're... Exactly. ... happens, we know exactly what's going on. We, we, we are not gonna have a surprise six months from now. We're gonna have a, we're gonna have a surprise now- Exactly ... if there is a surprise. And site logistics is a big, big deal. That's part of our test fit study.

And then we take all that data and we... A- and again, simple things are the most powerful. We've got like a 20-line item side-by-side analysis. We put all the basic information about our thing in there. Yeah. And then we say, "Okay, Mr. Reed," or whoever, "you fill in this other column of whatever you might do differently than us, and you monetize the things that are different.

We'll help you do that if you want." But, but at the end of the day, let the market decide. Not me. Yeah. Or, or you- [00:26:00] Market has to decide ... or any- Absolutely. Of course, yes ... anybody. But, but let, let the market decide. One of those numbers is gonna be better than the other, and then you can... Then you have a way to say to your shareholders, your board, whoever, you say, "Well, I'm gonna do this because there, it's right there.

I love the fact that you're able to go through these processes and really make these things, dare I say, more concrete. Yeah. Yeah, I couldn't resist- seriously, I mean, you're talking like just a lot of money being invested, and little screw-ups can cost big dollars. Oh, yeah. Yeah. And this is, I think, avoiding so many of those. Yeah, yeah. If we're gonna have an issue, let's have it now. Let's don't have it six months from now. Ex- exactly, before the cranes are up and all the good stuff, so yeah.

So the test fit process again, it's not rocket science. Anybody can do it. You gotta have the will to do it, the ability to do it, but you, you gotta do it. And, and- It feels like that's like the absolute most foundational stuff that you can take when you're pursuing this ... and, and that's, that's part of our, our DNA.

People call me all the time and they say, "Hey, I'm thinking about developing this deal." And I say, "Well, you need to give me three or four weeks so we can study it." And, and you can know, not guess or think, but you can know the direction that [00:27:00] makes the most sense. Yeah. What are, what are some of the common mistakes that actually do pop up when you're doing this?

Well, a lot of times people rely on fake information. They think that the lead time for X, I'm not gonna- Yeah ... throw any particular- Of course ... material- Yeah ... under the table, but is, is this much, when really it's this much. Gotcha. Because they, they- You're assuming it takes six months to get a load of something, but actually it could take you a year or something like that.

It, it... Or, or years. Or years. Who, who knows? Oh my goodness. Wow. Okay. And it also... And it's complicated. I'm not being- Yeah ... critical. It's, it- No ... they have a, they have a hard, they've got a hard job. Of course. Like the equipment, I mean, the lead time for the stuff that goes in our battleships It's a long time, whether it's low voltage switch gear, it's UPS, it's what- I've heard there's two and three-year backlogs on a lot of- Yeah

equipment out there. So in some- I mean, how crazy is that? And the prices can fluctuate. That's right. You might have agreed to a price, but that price could still change on you. Right, the price will be what it is when it delivers. Exactly. I feel for them because they, they, yeah, we can go really fast and we can deploy really quickly, but it may not matter in some cases.

If you can't get the equipment to put in the building, [00:28:00] then why is it important for us to go as fast as we need to? You need to, yeah. At that point, it's like, we don't need to be on a rush if you're not gonna have anything to go in here for another 16 months. Right. So- So give us a little more time, maybe you save some money.

So what's happening is that there in my opinion, is that people developers are calibrating all of those things so that one doesn't precede the other by months or years. It all gets there at about the same time, and then they just delay their project to the point where all that stuff can get there at the same time.

Yeah. So deployment speed is just as important as it's always been for us. Construction speed is absolutely i- important. Y- but you may delay the project for a while until all that other stuff happens. Comes together, yeah. Yeah, whether it's community pushback- Yeah ... or it's the lack of power, interconnection, whatever it is, you can delay that building later.

But the underlying benefits, the features and benefits of our system are still there. They just don't happen when they may have two years ago. It may be f- four years down range or something. My takeaway from this is, is like literally how foundational, of course, the structure is.

I mean, everybody's thinking [00:29:00] about the switch gear and all the stuff, the, the servers- ... and all, all these, all these issues. But when it comes down to it, like you gotta have this structure. And by working with a company like Metromont on the structure, it's like you're also walking through really the basic planning of the whole thing- Yeah

when it comes down to it. I find that just fascinating- in so many ways. Yeah. I, I, I do, I do too, but one of the other things that is fascinating is the attention to driving activities off of the job site. And that's always been, in- inherently important, but it's, it's more important than ever now because of the risk of having people on the job site.

W- we're gonna have a, we could have a gigantic building, a 300 megawatt building that's, 800 feet long and two stories, three stories. Well, we're gonna have two cranes and two crews on site. We're a single intensity subs, so we're gonna have like 25 human people On that site erecting this gigan- the size of an aircraft carrier.

Yeah. It is, literally, yeah. Yeah. We're not gonna have hundreds of people that are competing for access to the site that you have to find, keep healthy, and [00:30:00] keep safe. We're gonna have 25- Yeah ... on a- on average. And also, you're, you're making... All these parts are being made offsite, and they're being brought in.

So you have a controlled environment- ... to where you're making them, so that's really good for safety. Yeah. Oh, a- and absolutely. Like, I mean, you don't have wind, like, pushing something over. Yeah. Yeah. And then you bring these parts in. You can assemble them knowing when they're gonna come in 'cause you know when they're coming.

I just feel like it makes all the planning so much easier- ... and more efficient. And once again, it, it comes down to the speed to market thing, man. I know that is, like, the number one thing is speed. Speed is so important. Yeah. But also certainty of outcome. I mean, let's, let's face it. Right.

Y- do you want a surgeon to operate on your 10-year-old daughter that, that's done three operations- ... or do you want one that's done 133 operations? So certainty of outcome's a big thing. We have a lot of repeat clients. They don't even go through the test fit. They just say, "Make it happen."

Yeah, I mean, well, they understand at that point how you work- Yeah ... and they're like, "We like this." I had a very good client of ours, they go- called up and said, "Hey, we got a project." I'm not gonna mention the state, but it was really far away.

And they said, "We got a project, and we need you to do it." And, and I said, "We really, we don't have a [00:31:00] presence there." And they said, "Well, you're gonna do it anyway." Help me with this then." Yes, sir. So- Yeah, yeah, yeah. And we did, and it was- That's awesome ... yeah, 'cause they- I love that.

That, that- Because- ... that's saying something right there. Yeah, I mean, we're like, we're like that broken-in pair of shoes in the closet. You reach for them 'cause you know they're gonna fit. They're not gonna give you blisters. And, and- It'll take you where you wanna go. Yeah, exactly. Very much so

so anyway, all, all of that is in my opinion, flows from a test fit study. Let the market decide. Important stuff. Yeah. I mean, I feel like that is literally foundational. Yep. And that is amazing, and that's why I'm so glad we have conversations like this 'cause it, it shows that as these things are coming together, this is not willy-nilly stuff- Mm, mm

uh, and that there's lots of problems to be solved, but they can be. Yep, that's right. Especially when you're dealing with all these things up front through doing something- Up front ... like a test fit. That's the key word. Yeah. Is, is just absolutely amazing. I can't look at these things the same ever again 'cause I'm like, man, that's crazy how this whole thing came together.

And it started with a test fit for crying out loud. So oh my goodness, Marshall, fantastic stuff.

So Marshall, you gotta tell me, like what have we [00:32:00] not covered in this, in this scenario? I've learned so much. Mm. Anything that we've, we've left off the table that people need to know about Precast? I don't think that there's anything that we haven't covered, but I think it could, we could congeal it a little bit.

Yeah. At the, at the rate of change- Okay ... that the data center market is, is undergoing, and it's just phenomenal. It's mind-bending- Yeah ... frankly, to me, and I, and I live in it every day. Yeah, no, it's fast. You, you can't do the things that you did last Tuesday, today. You've gotta find ways to go faster.

You've gotta, you've gotta become meaningfully engaged even earlier than before- ... so that you can positively affect the course that a project takes. And the other thing is that users, end users particularly, they're, they're not, they're getting over this onesie, twosie, not that that's a bad thing- Okay

Procurement things. Yeah. And they're saying, they're coming and saying, "I want you to build six buildings." Yeah. Or, "I want you to commit," whatever, "15% of your plant- ... capacity- ... for, for some indeterminate period of time." Because the, the way things are headed, the bigs are gonna get a lot bigger, and then the not bigs are just gonna become [00:33:00] irrelevant.

So people need to adjust their thinking to thinking big and, and doing big chunks of stuff and, and organizing their affairs in a way that they can do that, and that's different. You didn't see that, three years, five years ago- ... certainly 10 years ago. Oh, yeah. You would, you would pursue one project as if it was the only project on Earth.

Yeah. And now it's like, "Well, tell me about your campus." "Tell, tell me about how we're gonna do all" ... you got 1A, 1B, 1C, 1- Yeah. Exactly ... it's not just one building. It's, it's- Yeah. It's, it's- I wanna commit- It's a campus. Yeah. Yeah ... why would I, why would I do your thing on one building and then have to pr- and shake that Etch A Sketch clean on the- And then start over again

start over again. How do I make this repeatable for across- Ex- exactly ... across all the things? So that's a, that's a big different thing. Very cool stuff. Interesting conversation all the way around, and it's just, this is just one of the aspects within the whole data center industry. I mean, we've learned obviously the backlogs-

that, the the, the different industries that all flow into this. It, it's this complex weaving- ... in order to make these things come up, and the idea is we're not [00:34:00] building any less data centers. We talked about the zettabytes of data out there. It needs to be processed, and this is how it's going.

Yeah. And this has been a fascinating exploration of just the foundations. So oh my goodness. You could have a lot of fun doing what you're doing. I, I do. And I can tell you really enjoy it. Yeah. And that's just awesome. Yeah. Yeah. Great stuff, Marshall. Well, thanks for, thanks for having me. That's what it's all about.

That's why we have the Empire Lights here with Comcast, because this is how we find these things out and get to the truth as to what's really making things grow- ... and how things are coming together. Pleasure talking with you today. Likewise. So glad to be- Yep ... part of the show today, Marshall. You are the best.

Appreciate it. And in case you forgot, this is Jonathan Kersting with the Pittsburgh Technology Council. We're having conversations like this each and every day, and I look forward to seeing you on the next one.