185: Paul Zeckser, Co-Founder and CEO of LightTable

Paul Zeckser didn’t follow the traditional path into construction. He’s not an architect, engineer, or contractor – instead, he built a career in tech. But as the Co-Founder and CEO of LightTable, he is solving one of construction’s most expensive problems: helping project teams catch drawing and coordination issues before they become RFIs, change orders, and delays.

LightTable was built around an issue Paul believes is bigger than effort or attention to detail. Project teams are managing hundreds of drawing sheets, thousands of pages of specifications, and countless cross-disciplinary decisions. Even strong, experienced teams can miss technicalities or miscoordinate details. LightTable uses AI to review drawings and specs during pre-construction, identify errors and omissions, and rank the issues most likely to create costly problems once construction begins.

Before LightTable, Paul spent years building products by getting close to the end user. At HomeAdvisor, this meant learning directly from contractors and understanding where software helped, where it failed, and where the pain points were for the customer. LightTable carries this same approach into their technology: listen closely to your customers, understand their workflows, and develop accordingly.

Inside this week’s episode:

• How LightTable is solving pre-construction risk, not just speeding up document review
• What a missed coordination detail on a multifamily project reveals about the true cost of drawing errors
• The nerve-wracking sales process that Paul says earns clients' trust
• What's next after LightTable’s $22M Series A as they enter the next stage of AI in construction

Thank you for joining the Design Development community. Subscribe to capture insight from the latest episodes dropping every Wednesday.

Our goal is to help people like you identify opportunities for growth by sharing the journey of industry leaders. There is no one path. Success isn't a straight line. There is something to learn from everyone's story.  Let's go!

Leave your thoughts in the comments and reach out if you want to be a valuable guest on the show!

NEW: More from me on Substack

It's Structural: Get my take on the latest news, trends, and policies shaping commercial real estate. 

Follow us on LinkedIn, Instagram & YouTube

Find out more at https://h-o.engineering/podcasts/

Transcript

Rens Hayes 0:08

Hi everyone, welcome to Design Development, your hub to learn direct from top performers in real estate design and construction. I'm your host, Rens Hayes, co-founder of H+O Structural Engineering, a lifelong learner, and I'm personally obsessed with high-powered organizations and the leaders that got them. Let's go. It is my pleasure to introduce Paul Zeckser of Light Table. Paul, thanks so much for joining us today.

Paul Zeckser 0:30

Hey, Rens. Thanks for having me. I'm excited to be here.

Rens Hayes 0:33

Yes, Light Table is on the cutting edge of technology here in real estate design and construction industry. And I hear you have some exciting news to share with us today. What's going on?

Paul Zeckser 0:45

Yeah, we're really excited to announce that Innovation Endeavors based in Silicon Valley, a venture capital firm that was started by Eric Schmidt from Google, has decided to lead our Series A round of financing. It's a $22 million raise that we're super excited to apply to growing the company, sales, marketing, product development, customer success, going to grow all those teams and continue serving this incredible industry. That is super exciting.

Rens Hayes 1:15

Congratulations to you and then Paul. For those that aren't familiar with Light Table yet, what problem do you solve that allowed you to get here to a successful Series A?

Paul Zeckser 1:24

Yeah, great question. There's so many different ways to slice it. And the way I've been thinking about it recently is understanding and managing risk, specifically in pre-construction. And where we've started is we're looking at drawings and specs anywhere between 50% DD and 100% C D. And we're asking an AI, hey, where's the biggest risk in these drawings? Where do you see errors and omissions across every discipline and every scope? And I want you to rank those for us of which errors and omissions are going to become the biggest problems during construction. Surface that now. I want to know about that now today, and I don't want to find out about it three months from now. And so that's the core of what we're doing is we're leveraging large language models and vision models to perform that work.

Rens Hayes 2:13

And I imagine you have to train your own proprietary model to be able to have this judgment.

Paul Zeckser 2:17

Yeah, we do a lot of work on top of what's currently out there, like large labs, the Googles and the Open AIs of the world have a great foundational product on which we can do a lot of really interesting things, but the bulk of the work is work that we at Light Table. I'll give you a quick example. I I actually recently with one of my teammates, we uploaded a drawing, a demo set for drawings and specs. It was it was a multifamily project. Claude was able to identify when we asked it to say finest errors and emissions in these drawings. It found seven issues. And five of them were pretty good. And two of them were totally wrong. When we ran it through Light Table, we flagged over 700 issues. Two of them, two, 200 of them were critical or high importance. So 20 critical. A critical issue is usually like $100,000 or more. And so it didn't, Claude didn't find any of the criticals. It found a couple of the highs. We found 200 highs. So it's a pretty big delta between what you get if you ask a large language model off the shelf like Cloud or OpenAI to do this versus if you ask Light Table.

Rens Hayes 3:21

What's the deliverable to the customer? How is this information presented?

Paul Zeckser 3:25

Yeah, great question. It's a web interface. So LightTable is a web application. Login. It is both a workflow management tool, but also presents essentially as a report. So you can see all of the clouding and the comments on the drawings and specs. You can easily slice and dice, sort. I want to look at all of the medium priority structural issues. I want to look at all the critical ADA, fire and life safety. You can move around real easily. And then you can create an ability to collaborate with the rest of the team. So architect, engineer, see, everyone gets a seat and you can kind of like Google Doc style, make a comment, tag somebody in the comment, they get a notification, they can jump in and have a conversation with you about how to resolve the issue.

Rens Hayes 4:11

That's that sounds really amazing, Paul. And I'm excited to share some examples and kind of the X's and O's so people understand what it's like to work with Light Table. But I want to get to know more about you and the journey to where we are today. I, for one, and I think we even say that like I get obsessed with like really cool businesses and go-to-market strategy and the leaders that guide them. And coming from the design industry and construction industry, I grew up in in steel construction and I run an engineering business today, structural engineering. So I work in a lot of different assets, facets around design, construction, real estate. The thing that's different about our industry versus, say, a tech founded industry is that these are generally bootstrapped companies. They're professional services, labors, there's a need, there's a cost, and the gap is your profit, right? So they're generally not venture fundable. And it took me a while. I think I think it was Peter Thiel, the concept of escape velocity. I never understood how like valuations and how venture funding and how the technology world worked. I would love to get into that a little bit to share some of those insights. It makes me a better business leader as I learn from other industries like yours. Like when you're in, I think they call it like stealth mode, right? You have this idea and you need to generate a little bit of money, but you don't want the world to know because you don't want somebody else to take your idea. Like, how does when did you first become aware of this problem that you were going to solve with? And then, like, how did you get the ball rolling?

Paul Zeckser 5:34

Yeah, that's a great question. In my view, Light Table, at a minimum, is very much a three-headed monster. At this point, there's many more heads as we've grown. This would not be the company it is today without all three of the co-founders and the key hires we've made since founding. And the reason I say that is because I really think to tackle this space, you have to have deep expertise coming from this space, coming from real estate and construction. And I also think you need deep expertise in large language models and vision models. Um, and I actually, maybe shockingly, do not bring either of those. And so that's why I have two great co-founders. So Ben Waters really was working on this idea first. He's an architect. He was trained at Cornell and worked at SOM and Gensler for a number of years. He and I first got connected and started talking about the opportunity to replace traditional peer review with AI. We had direct experience working on reviewing drawings and being pressured by senior people in the organization, get these out. And he's like, we haven't really QA'd them, and they're like, we don't care. And he had a problem with that and saw the steady decline in quality of drawings. And we then got very fortunate in being connected with my other co-founder, Dan Becker. He's our CCO. Dan has been in AI for 15 years and came out of Google. He started his own AI company, which he sold and then worked for the acquiring company as both the head of machine learning and head of product. And I feel like then there's little old me. And I've been building technology companies for over 20 years and have been an entrepreneur before. And for about 15 of that, I would say I was in loosely like the prop tech space. I built a lot of software for contractors. And so don't feel unfamiliar with the terrain, but I'm glad to have Ben and some of the other people we've hired from industry help beat the guides. And so I think if you don't build a company that has deep expertise coming from the AEC industry and from the AI industry, I think it's going to be hard to make it done.

Rens Hayes 7:39

Yeah, I love that frame of how important the complementary skill sets of the founding partners is to the endeavor. And I think you're absolutely right. When I look at our industry, it is so fragmented in how uh commercial real estate comes to reality. It's like you have an owner, you have institutional capital, you have bank and debt partners, you have attorneys, zoning, community, and then maybe a dozen consultants and then dozens of subcontractor companies that come together to it's like the fragmented delivery makes it really hard for somebody outside the industry to get their head wrapped around how that world works to develop a practical technical solution. So the fact that you have somebody that has some of those insights, but also understanding of the tech opportunity all in the same team is, I would imagine, a pretty compelling proposition to bring to an investor pool.

Paul Zeckser 8:28

Ben and Dan is really the models around which we've been doing our hiring, and we're now up to over 20 people. Most all of us are able to still fit into a room, and we have the AI engineers and the AE talent sitting next to each other. So folks with 25, one case, 35 years experience as architects, engineers, owners rep, and construction sites, general contractor. And so to see that team with that depth of experience on job sites and in the design rooms, sitting next to the you the lunch and learns are so rich because when the AI team leaves, the architect is like, this is all Latin to me, is Greek to me. And then the next week we switch it. And then we're talking about D D and C D and GMP and the AI team is they're taking all these notes and they're learning. And it's really the it's like the confluence of these two different backgrounds, but have this burning, joint burning passion to solve a real problem that I really think that's our are our our secret sauce.

Rens Hayes 9:29

Can you help from somebody that's outside of, say, the venture-funded world? You've now gone through pre-seed, a seed funding round, and now you're successfully at seed. What does it take to get to the round? Like, how are you approaching investors? We have an idea, we have some people and some experience. What kind of are the steps to go through those rounds and like where does your MV kind of show up?

Paul Zeckser 9:52

Yeah, it's a great question. I think in some ways it doesn't take a lot, but what it does take, you just have to be really strong. And I think the first is the team. We've already hit on that point. I think that's one of the reasons we've successfully raised is because the right team from the right backgrounds that also knows how to work well together. Like pre-seeds almost uh more of a bet on the team than it is the idea. I think so. We were pre-product, pre-customer, like all of that at the pre-seed. So it's you gotta say it's the team and what they're selling as the vision. You get to the seed round, and there there needs to be traction with a demoable product, and in our case, five really good design partners. So those are like pre-revenue customers. And we approached them and we said, look, we would love to build this if you're willing to try it and give us feedback on it. No cost to you, just obviously some of your time to give us feedback. And we were getting such positive comments on our first version of our product that that's what we were able to raise our seed round on, which we closed almost a year ago. And then we kind of worked in stealth over the summer to get ready to launch and then came out of stealth last August.

Rens Hayes 11:06

And so if we get a sense of that investment, so I think you said series A was 22 million, is that right? How large are the raises at pre-seed and seed funding rounds to kind of get to a series A?

Paul Zeckser 11:17

Yeah, in our case, 2.2 million pre-seed, six million dollars seed, $22 million A. So our total is about 30.

Rens Hayes 11:25

Very cool. And then the and then that's investing in the team and product development because again, you're pre-revenue and pre-customer in those early days. And then the difference between seed to series A is you're starting to show when you're paid customers. Yeah, you gotta start selling. Well, Silicon Valley, the the TV show said, No, you don't ever want to be exactly.

Paul Zeckser 11:45

Yeah, no, uh it'll never be enough. Maybe I'm a little old school, and I kind of like to sell and show some revenue run a time.

Rens Hayes 11:52

That's where my world is. That's how I think.

Paul Zeckser 11:54

Yeah. So yeah, you we started. We actually almost immediately after we closed our seed, we also closed our first ping customer and started signing customers throughout last summer, and then things really accelerated for us in the fall and the so far the spring this year.

Rens Hayes 12:08

Congratulations. And I know some of your customers, which is how we got connected. Are you able to share some of the customers you're working with or some of the projects you've got?

Paul Zeckser 12:16

Yeah. One of the folks in in your neck of the woods, customer of ours, is Suffolk. We're working with them and have really been enjoying that relationship. What's great about Suffolk is they're they're visionaries and it comes to technology and construction.

Rens Hayes 12:31

Very tech forward, absolutely.

Paul Zeckser 12:32

Very, very tech forward. They have their venture capital arm. They'll do a lot of really cool and innovative things over there. So we feel fortunate because in addition to working side by side with them on lots of projects this year, we're also getting feedback on how to do the product. One of our beliefs is that the closer you are to your end users, the better the product create for them. And so we to the point where Suffolk and others are like, hey, you really don't need to keep showing up. Like we're always willing to be on site and sit with and say, okay, we want to understand you just got the light table comments, and here's how you're applying them in your process. What do you do next? And what do you do with that? And sometimes what they do next is perfect and it doesn't need any work at all. And but a lot of times what they do next are like, yeah, this kind of sucks, but then I have to do this and I have to move this thing from Excel, and then I have to do this thing. And I wish there were some integrations here, but there's not. And so I end up just picking up the phone and doing a bunch of stuff over Teams. And I'm like, can we help you build like an agent to do that thing just talked about? And so that's our philosophy is this is not a successful company product, et cetera, if it's a point solution. Like light table needs to help solve the underlying pain, which is why when I started out, I'm like, this is really about risk, understanding risk, managing risk that's broader than just saying we're doing QA, QC on drawing.

Rens Hayes 13:51

I love the vision that your team has and how you're approaching the work. That follow-up question is what happens next? And then what happens after that? Because those questions lead to finding out that next pain point in the chain that might unlock the actual value creation at the end game, right? Like if you're just totally building on that initial go-to-market strategy. I think I saw in the text that Light Table is able to capture four times more issues than a traditional peer review.

Paul Zeckser 14:18

That's an estimate based on customers sharing with us traditional peer review reports that they've been given that were redacted and appropriately shared without developing confidential information. And then us comparing what Light Table was able to generate on the same set of drawing specs. The thing about it is it's actually that was a I think that was a marketing point we wanted to make early on. What really I've discovered is that AI and Light Table can be too comprehend. And so when you introduce system the system today is okay, hey, let's let's let's let's do a quick peer review. Okay, letter came in, 25 things. All right, I can handle 25 things. I can jump back into the drawings, I can find these 25 things, and I can decide if I think a correction needs or clarification. When you drop first first several rounds of light table, we'd say, Hey, here's 750. You're like, Yeah, thanks.

Rens Hayes 15:11

I'll block off the next month to go through that.

Paul Zeckser 15:14

And exactly as you would imagine, the design teams are like, don't bring that over here. I don't even know. Feels like this got sprung on you, right? And and so what we really focus on speed's always been there. We can do things in hours. It takes a traditional process. The key of it is in the end of those few hours, how do we deliver? Here's the 20 most important things. We can be really comprehensive. We can give you 700, but here's the 20 most important things. And then if you have time, here's the next 30 or 40 and give the user kind of the controls, the throttles over how deep they want to go and in what areas they want to delve into. So we've solved that by just continuing to pair back, but we've also solved it by asking our AI to go through and create executive summaries by scope and discipline. So I might come in and I'm the structural guy, I'm the MEP. Just give me the top three to five to seven things that are impacting coordination, structural, and architectural or whatever. So that's how we're trying to focus it is empowering teams by telling them where to focus in the limited time that they have.

Rens Hayes 16:21

And I imagine as you're as you continue to collect data and do more and more reviews, your comp your feedback gets better and probably more accurate on the value of the feedback, which I imagine as you go through this process, the feedback loop from the design and construction team on the information becomes paramount to making to your the improving of your language model.

Paul Zeckser 16:45

Yeah, it is. We love it when our customers choose to share with us RFIs and change orders on projects. It could be projects that we worked on, it could be past projects.

Rens Hayes 16:55

Oh, that's really cool. Like a postpartum, here's things that it did catch and hears, but it didn't. Ah, that's really awesome.

Paul Zeckser 17:01

It's actually it's how we sell. I say to everyone, I'm skeptical of AI. I hope you're skeptical of AI. We should all be. Be skeptical of us, be skeptical of every AI company out there and say to them, like, where's the actual value that's being created here? Oh, cool, it's fast. Oh, cool, look at this 700. Like, where's the actual value that's being created for this project? Best way to test that is give us projects you've already completed. First, we want to see the design documents. We'll put them into light table. We'll show you what our answers were. Now show us the change orders from that project. So this is a job that's already been done, ribbon cutting, tenants have moved in. You still got those change orders somewhere. And now let's go through which change orders would light table have foreseen and could have flagged for you. And which table, which one did we miss? Then the ROI cases, there it is. Show to the CFO, hey, we want to bring on this new tool. It's called Light Table. Okay, can they do what they say they're gonna do? Yeah, actually, they've done five five past projects for us. This is what we would have saved on each one of those.

Rens Hayes 18:00

So good. I absolutely love that strategy. And this is the cool unlock to me in AI and technology development. Is if you if we go back before the light table solution, it's let me show you how much value I can provide you by doing quality control and peer review in your drawings before you get a bunch of like that. Would be nearly impossible to do on your own dime to pure value, right? But here you are, you're able to iterate quickly and get that data to prove and support a sale that almost it shows how that depays for itself. Who is your customer, Paul? Is it the real estate developer, the owner? Is it the construction company? Is it like where's your point of sale?

Paul Zeckser 18:41

Point of sale is real estate developers, owner developers, and general contractors. And then in addition to them also being users, architects and engineers can use it too. Very cool.

Rens Hayes 18:52

And what's your what's the revenue model? Is this license space? Is this fixed fee? Is it per project? Is it performance?

Paul Zeckser 18:59

Yeah, it's essentially so if I'm an owner developer, we'll go through the back test process as we talked about a second ago. We'll say, we want to earn the right to support you on all your projects and tell us what your pipeline looks like. And so we like to do a commit to consume a million square feet over the next 12 months, or it could be 10 million square feet. Obviously, pricing gets better, price per square foot will drop the higher the credit. It can be 50,000 square feet. It's a price per essentially square footage that once a price, so let's say I bought a million square feet and my first project is a hundred thousand. So I draw down a hundred thousand from that million, and now the 100,000 square foot project, I can create as many seats as I want for all the people on that project. They can use light table for the full duration of the project. And what's really cool is we have a versioning feature. So as new sets of drawings are uploaded, we can run them and compare them to the last set and help the team keep track of how many outstanding errors and omissions there are and who's committing to resolving. So you can also run light table multiple times on each project. So all of that's included in the price.

Rens Hayes 20:13

That's really cool. So yeah, if I have a million square feet of total upcoming development, my first building's a hundred thousand. That hundred thousand covers as many iterations on that one building and for all of your supporting teams. So if I buy that as an owner, the contractor gets to use it, my architect gets to use it, my engineers get to use it to improve the outcome of that process. That process. Yep, that's exactly right. That's amazing. And then when I think about that long-term commitment to show that drawdown, you knock that one out of the park after your sale process for even value on the last one. And then they get to realize it already paid. You're ideally paying for yourself in that first job.

Paul Zeckser 20:49

Yeah, that's right. It more than pays for itself in its first job. We're seeing ROI ROI in the neighborhood of 20x our fee when we do the back tests. Sometimes it's a lot higher, sometimes it's a little lower. The thing that gets really cool though is especially for general contractors and owner developers who have fairly robust pipelines and are using it on several projects, is with their permission, it gets smarter on their project. Right. So they basically say, I want you to pay attention to where we have issues, and I want to also give you feedback on the issues you're sending us. And when I see a really good one, I'm gonna tell you, make sure you always look at this. Double down on this. So then the next one we do will look even harder for those kinds of issues. And if they say this stuff is not that important to me, then the next time around we'll we can look even less. We can look less for we can do that on projects starting on day one moving forward, but we can also do that if they share historical project data with us. We can then do that on day one if we've had their last five fixed projects.

Rens Hayes 21:51

Quick break from the show. Thanks for tuning in to design development. We're trying to help as many people as possible. So if you can subscribe. And leave a review on today's episode on whatever platform you're listening. It would be a great help. It's the only way we're going to reach more people. Let's get back to the show. You mentioned a critical item being of $100,000 or more. Can you share a few examples on some of your projects where you've identified these critical issues that would have been a six-figure miss for the owner?

Paul Zeckser 22:20

Yeah. Yeah. We just recently did this back test. So I know exactly what the dollar amount was. It was a multifamily project that 300 so units in Southern California. And when we got the drawings, put them into light table, and it flagged a coordination issue in the kitchen between interior design and the electric. Electrician had drawn the placement of the disposal switch in the backsplash, is where the wiring went. But the interior design team put the button for the garbage disposal in the counter. So they weren't close enough to each other.

Rens Hayes 22:57

And this is on multifamily, so it's not a one-time mistake.

Paul Zeckser 23:00

It got repeated a couple hundred times. Um that was missed until all of the countertops were ordered and they were starting to be installed. And they said our electricity is over there, but the buttons over here. So that was a couple hundred thousand dollar change order. And this is again all part of the back test process. And so, yeah, we showed them in light table, like right here, the AI clouded that section of the drawings. And the comment was interior design, electrical drawings not coordinated. And had either of those teams seen that from the outset, they would have not made that mistake. So that gets repeated over and over again. Coordination issues are by far the most common ones that we find. Without a doubt.

Rens Hayes 23:45

And how do you, if I think you you start at say 50% construction drawings and then you'll do it on the versions at 100% CDs and buyout? Do you have to cater your review based on that level of drawings? Because otherwise, like you said, that's how you end up with and as a as running a structural engineering firm, if I review a 50% CD set as if it were CDs, I'm basically redrawing the building in my comments, right? So you have to like aim. How do you manage that in feedback?

Paul Zeckser 24:12

We can both tell the AI what the stage is at, is usually the easiest way. But yeah, if we do not pay attention to, and we'll start as early as 50% DD, if we don't pay attention to what stage we're at, we're gonna give super unhelpful comments. This is supposed to be here, and you're like, no, it's not. We're at 50% D D. It's not, I don't do that until C D. And so the AI, this is something we've been working on for the last several months, and I think it's getting quite good at this, is giving feedback relevant to the stage of the drawings.

Rens Hayes 24:44

Very exciting. Do you see a path where you're actually going to be able to give cost feedback? Let's say your electrical wiring example. A I noticed in this unit, you're running your electrical over to the wall instead of the center counter. The cost premium to do X versus Y is this. And traditionally in our data, we see this as the market standard.

Paul Zeckser 25:06

We we we just did this actually in the data center. We flagged aluminum versus copper wiring. And the AI is like, you don't need to be using copper wiring, you're always yeah.

Rens Hayes 25:14

That type of feedback and intelligence will be so valuable. And when I come back, like your example on the wiring or the alert copper, that could be a typical detail that somebody inherited in our on their drawing set, and they could have thought that it was aluminum versus copper, right? And they just didn't think to go back and scrutinize that detail because they thought they assumed that it was correct because it the one is the last product, or the chaos of a design process when you're under time duration, you have to rush to get to a deadline and you think, or you maybe you had a verbal conversation, like, hey, I know this used to be on the wall, we need to move it to the sink, and then somebody forgets about it because they were doing a million things. It's like they know, but then it's all the contractor should catch that. But then, but maybe not in time because the counters were already made because they were critical path, not where's the butt. Right? It's like all the time. So you can really see how these little things can, it's no one person's fault, but it's the nature of human error through such a fragmented process that this technology is such a risk that this is you asked this earlier, like a light bulb moment on on light table.

Paul Zeckser 26:17

And for me was understanding how many people are involved in the process of designing and then building, and how long a period of time that process is going on. And the idea of trying to coordinate across that is it's impossible. We I think have a tendency to sometimes blame an industry or blame teams. But I as I've looked at this, obviously, as I've acknowledged as an outsider, I don't think it's actually humanly possible to track all of this. Hundreds and hundreds of sheets, thousands of pages of specs, every mining detail, and something can just get missed. And it's the type of wiring or it's the location of an outlet and the backsplash. And that's why I continue to believe AI will be a supplement enhancement to the professionals who work in this industry, and certainly not a replacement of. I think the judgment and the final call on all these things will still reside with the leaders on these teams. It's but how much better to walk in to a review meeting saying, I've had this AI running overnight, and now I've got this report in front of me, and I feel so much more empowered now versus I don't know where the sleeping giants are. I don't have any clue. So I'm just going to carry this really big contingency in and just like hope that it covers it.

Rens Hayes 27:33

I couldn't agree more with your take. And I'll go back to your example of the disposal button and the electrical wire in the wrong spot. Let's say you're totally stuck or you're totally crammed for time and you need to hit this deliverable and light table AI runs after the construction drawings went out. I can't change it. But guess what? We now have this other document to send to the contractor to capture the risk, even though we didn't capture it in our drawings. And I think that's an o'clock. Chasing perfections impossible. But as long as we're creating this communication line to manage risk, we're all we're gonna be far better off.

Paul Zeckser 28:04

I I we first started selling to developers, and the first time a GC saw it, they said, This will be my pre-bid RFI list, and I can even continue using it after GMP because there might be things I missed, but I want to understand that. If I'm gonna bring a team on site, I don't want to be like, oh, sorry guys, like we got the wrong countertops, gonna have to send you off.

Rens Hayes 28:23

So learning about risk sooner is always better than learning said. I you mentioned the fragmented delivery or how many people touch these. I actually even look at the whole process. The fact that we design a building which is not construction level drawings. That's why there's this whole shop drawing review process, is like we created this whole nother sit series of coordination documents because we know that this isn't going to be perfect and ready for the contract. So there's design drawings and construction drawings. And when I look ahead in the future, especially in my world, which is structure, I almost look at it as thankful on how simple it can be. You have a wood skew, a concrete skew, and a steel skew. It might be some version of wood or steel. You might have rebar or structural steel or light gauge metal, but those are the three skews. I could today provide construction level drawings to all of those trades. The problem is my fee to do so would be probably four, five, six times my engineering fee, which when we go back to how does a commercial real estate development happen, that's during general partner money. They haven't got an approved deal, they haven't raised an institutional investor, and they don't have bank debt. So they don't have the resources to invest in that level of certainty in the design. They need to get to a fundable deal to then release it to them. I think technology is going to solve that problem. I want to get to construction level drawings and eliminate all of that. And I see technology like light table, like in the core path on getting there. And when I look at a simplified structure, I I that feels borderline impossible on the architecture side. I have three different types of SKUs. They have, I don't know, hundreds of SKUs, like everything under there's so much room for like coordination that this type of software is so important. I want to go back to Paul and Paul's career journey on how he founded Light Table. You studied international affairs at University of Colorado Boulder. Is that right? How did you think your career was going to be?

Paul Zeckser 30:13

I'm pretty sure at that time I thought it'd be pretty cool to be a foreign services off foreign service officer, kind of go live in some cool new place every two years. What did you even know? Who inspired you, or how did you even see that was a potential career path? When I was a freshman in college, I interned on Capitol Hill for a member of Congress. And maybe even backing up a few years before that, when I was in high school, our high school took a trip to DC. I thought it was that's awesome. And um, so I got I wrote in and applied and got an internship my freshman year. And it was surprising to me how this was a long time ago, but in my view, how high functioning it was, despite that not being the reputation that it had even back to had a lot of admiration for the person I worked with and his staff. He had been a Marine in in Vietnam and an attorney, and I just thought he was a really incredible guy. And so at first going to college, I thought that might be something I wanted to do.

Rens Hayes 31:14

I think it's cool to hear a high school field trip inspired this exploration for you.

Paul Zeckser 31:18

It did, it really did. Kudos to the teachers out there who are like grinding, especially this time of year. They're like, could the end of school please come? Yeah, with summer is around summer break for grinding out there and are doing awesome things for kids. And so anyway, that's where it all started. But I had started a couple businesses myself and had always been an entrepreneur. I was in high school, my first job was landscaping and lawn mowing. And I joke by necessity sprinkler repair. I learned sprinkler repair because my partner gets going to be sprinklers when we were cutting the grass, and my dad's like, You got to go back and fix their sprinkler, and I'm not coming to help you. Um, yeah, we just I've always been like, Let's go start something, let's go do something. And were you surrounded by entrepreneurs, like a business, small business owners? What kind of my grandfather was an entrepreneur. Both my parents were teachers, but my grandfather was an entrepreneur, and I thought that was a pretty cool thing every time he'd tell me stories about his ups and his downs and the businesses that made it and the ones that failed. Very cool.

Rens Hayes 32:17

If I share some of my journey, my my father ran a small business, a steel construction business, which I got to grow up in and see for hard work. My now father-in-law was ran his own real estate agent for single family homes. He ran his own company, and my other friend ran his dad ran a painting. Like, I didn't know anybody in corporate that I saw. So, like I literally grew up thinking like, well, the only way that I'm gonna be able to do this is if I start my own business. And so, like, I like was laser focused on that from day one. And did I see you were a sales director and account executive at compete?

Paul Zeckser 32:47

Yeah, I was. I got a degree in a graduate degree in business from Maryland and moved to DC, excuse me, moved from DC to Boston looking for a job. My fiance had just gotten into gradual in Boston, and so I wanted to stay close to her. So I started knocking on the doors of any place because no, this is like 2004, kind of early days of online applications. So I would still walk in and drop off a resume, and no one was really taking an interest. And so I grabbed a beer with a buddy who told me he was at a tech startup, and I was like, What's a tech startup? And he's I'll see if I can get you an informational interview. So I go in and talk to them and they sent me around and talked to a couple people, and the sales VP said, Maybe you can make slides for us. Maybe our sales team goes out on the road, they do these pitches. A lot of them don't know how to make slides. I presume you can do that. You just went to business school for two years, and I'm like, Yeah, I can make slides for you. And so that's what I did is I was prepping the sales team for their pitches. I do research on the customer and I'd put together the story. And after doing that for six months, I was on one of the pitches. And like when you're trying to present someone else's slides, sometimes it doesn't go well. And so this guy kind of started going a little bit off the talk track and struggling in one of the slides. And I just kind of jumped in and finished the presentation for him. And on the drive to the airport, the VP of sales was like, that was a great save. That ended really well. Would you like a job a sales guy? And my first reaction was, is that really what I want to do? Like, do you go to do you go to business school to be a salesperson? And we were a small company, 40-some people, and the CEO offered to grab breakfast with me and gave me the advice of everyone needs to spend time in. I did sales. If you want to be anything, you want to be a CEO someday, you want to be anything, get some time in the seat selling. And I loved it. It was great experience. I would not trade it for anything. And yeah, that was my stint as a sales director at Compete.

Rens Hayes 34:48

I agree with that advice. I do think everybody should try to you really learn how to communicate to a customer and you get to handle the objection and then go sit with that objection and be like, how do we actually solve this problem? Or are they right? And it helps you iterate. It almost like when you describe what you're doing at Light Table, the same process you learn by going through a sales process you're doing with Light Table by asking those follow-up questions, like, what do you what happens next? What happens after that? What's the next? That's really cool. What led to the transition to home advisor? Seems like you had a long stint there and some pretty good success.

Paul Zeckser 35:23

Yeah, we decided to come to Denver where we're both originally from, and opportunity to work on something new. And home advisor was the first time I got to work in product, which I found I have a natural skill set for. I'm just a very curious person on how folks interact with the tools around them. Their smartphone, their dishwasher. I'd like, I can't look at a user interface and not start breaking down like what is good or bad about it. And so I came into home advisor in a role where my job was to look at data, like how our customers, how the contractors and the homeowners were using the platform and propose hypotheses about how to make it better. So one of the problems we would have is like contractor churn. Okay, like what's a way to fix it? And every time I had thought I had an idea, they would say, you need to talk to the engineering team. Like they're the ones who run the website. If you want to run a test on the site, and they'd say you got to write some requirements and how do you do that? And I just like self-taught into becoming a product manager and self-taught and got a lot of help too. And so I did product for over a decade there. And it was some of my most favorite work because I would get to go on ride-alongs with a plumber or roofer or painter for the day. And yeah, you'd see how they interact with your tools, but you also just get to spend time with them. And I found that the most high value insights came after you just spend a couple hours with the person and then they'd make some offhanded comment about, yeah, I always love getting your guys' text messages because when I click on the lead and I'm like, wait, you get the faster on text than you with app notification. And that and like you get these incredible insights about how your product's being used, but you got to get in there and hang with them. So I love that.

Rens Hayes 37:12

And if I if the data I got or the information I got about home home advisor, you're at this business when it was from about 200 million and eventually went to a six billion dollar public, right? There's this like growth period along the way. And here you are telling a story about how you actually sat in a truck with a plumber and an end user. So this isn't like a small 10-person business. You're doing this while a business is at massive scale and you still find the importance of being with the end user and building a relationship. And that relationship created the trust for them to actually share something like that.

Paul Zeckser 37:43

Like what a cool insight that I think that's how you build great products. You spend time with people and get to know them and their world and don't just look at the data and make decisions from afar.

Rens Hayes 37:53

I actually, when I think about our business, I spend almost no time whatsoever thinking about my industry or my competitors. And I spend almost all of my time thinking about my customer and my industry partners. And that's the lens that I communicate through, right?

Paul Zeckser 38:11

I I think we're well enough past this to be for me to be sharing any kind of non-public information. But when Home Advisor acquired Angie in 2017, we kind of went into it thinking we're the biggest home services marketplace in the country, and they're the second biggest. I bet we have a ton of customer overlap. And so the deal closes and we start sharing lists, and we're planning out the integrations of how Home Advisor and Angie's list are gonna combine forces. And there was less than 10% overlap of their customers and ours. And that was felt like that was a lottery ticket. And that was after two decades of insane TV spend and hundreds of salespeople call 10 less than 10% of 90%. And I think it makes your point, which is don't worry, you don't need to worry about your curves. There's more greenfield than any of the companies added up will ever really get to. Worry about creating an incredible product that your users are like, I'll die if this goes away.

Rens Hayes 39:08

Absolutely love when I and I don't know how accurate this data, I was using AI like ChatGPT or Claude to do research. Private commercial real estate development is 1.2 in construction value a year in just the US alone. Single family homes are another 600 billion, and then there's the entire public markets. It's I'm pretty sure there's room for quite a few structural engineering companies to serve that market. I don't need you to fail for us to be successful, right? We can all succeed together and serve our customers.

Paul Zeckser 39:35

No, in fact, it's probably better that you're really successful because you're gonna push it. The more success, the more we all get pushed. The more we're pushed to innovate, the more we're pushed to be the best versions of ourselves. I'd rather have that. But we all want the seven game series. We don't want the four-game sweep. Let's battle it out. Let's have some fun.

Rens Hayes 39:53

Amen to that. You mentioned a multi-time founder and you started with Landscape. What other businesses have you launched along the way?

Paul Zeckser 39:59

Yeah. So to be clear, this is the first time I've been like in the founder CEO seat for a venture scale company. About five years ago, I started my own product strategy and engineering agency, like SHOP, the perfect post-home advisor, COVID era job where I could say, got a playbook here for building and scaling, really large technology companies. I'd love to come in and help you do that. And so I had a couple of employees and a network of contractors, and I did that while I also helped manage kindergarten or doing education during COVID. So it was that flexibility thing. But every company I've been in has been in early stage growth stage. The one I joined in Boston, I joined at 35, 40 people and left and was 150. And then you already mentioned kind of the growth curve of home advisor. So it's where I get the most enjoyment, is when there's a lot of opportunity out in front of you.

Rens Hayes 40:55

From my seat, they seem like really fun journeys to be on and to be contributing to. When you look forward, what's say like a three-year vision for Light tape to be?

Paul Zeckser 41:04

Yeah, we want to be working on the most complex, which often are the most high profile projects out there, construction projects. We think complexity is where we create the most value. And I would love to see us become a tool that is used throughout pre-construction, mitigating, managing risk in all its various forms: QA, QC, uh bidding, estimating, value engineering, procurement. I'm okay not being the number one tool in every one of those areas, but what I would like to be is the system of record or system of work that people can connect in with. And the AI that they're working with is getting smarter on every project. And so gone are the days where you're like, oh God, we keep making this having this problem. I want the AI to be like, I gotcha. The next project, we're not making this mistake. I'm gonna look for it and I'm gonna tell you about it as soon as it comes about. And so it's really a companion, a buddy along the way for the top owner, developers, architects, engineers, general contractors.

Rens Hayes 42:11

I just think of how valuable that is from a company in operational standpoint. I want those same things for my at in an engineering company. And if I go back to some of the environments I've worked in in the past, it's like you learn by working under somebody that's experienced. And then you find out that all the mistakes that you made were mistakes that all these other people made around you. And it's like we didn't have any mechanism, the company to learn to reduce the learning curve on all the other future hires. And now you're like automating that through technology for this process and in an area that's super costly to like how valuable. Entrepreneur, you've sat in chief product officer, you're a founder, you've raised capital. What's the best piece of advice you've received in your career journey that's had like the biggest impact?

Paul Zeckser 42:56

I think it has been there's too many people to point to that that I could say I give credit to this to, but it's about the that you're on. It's about who you surround yourself with. That's ultimately, I think, yes, of course, thinking about my end user gets me up each morning, but showing up for my team and no one there showing up for me. Hiring really great people who are professionally ambitious, but personally humble and don't take themselves too seriously. We like to say take our opportunities seriously, not ourselves. And I think any great company is gonna have a great team. It's not gonna be like the cult of personality founder thing. Like it's gotta be beyond that. It's gotta be the whole group. I've seen that modeled really well in several places. I've also seen it not modeled well. And I am now 25 years in believe like that. If you're gonna say there's one really key thing, it's you got to build a great team.

Rens Hayes 43:53

Great piece of advice. And I think part of how we started to build a relationship is like you guys seem on a great path, but I think also like morally grounded with great values. I think I I got the chance to meet your partner, Ben Waters, too, who I thought super highly of and like on this exciting journey. And at the end of the day, I really think you have to align success for your team and your customer and your shareholders. And when you can balance those, if you ever have friction where one of those is falling short, you're gonna run into a wall. You guys have an acute awareness of that. The importance of having great people on your I find that there is levels to that lesson. That is something you say, and everybody's like, but it's like there's hiring great people at the entry, then there's hiring great people that are on the operational level that can like secute the job and take instruction and work in a framework. And then there's the people that like build the plane after they jump out or that can drive business and like champion sectors of your there's growth levels to like how big of an impact that great person is. And I, for one, am still on that journey as I continue to learn those levels, but it's like such a cool thing to find. Yeah, me too. Well said, I like that. What's your favorite book or podcast you share with the audience?

Paul Zeckser 45:00

I'm reading a book right now, it's super long. I've been at it for a while. It's called Centennial, and essentially, so I live in Colorado, it's it is a historical fiction account of the entire history of Colorado. Get this from the age of dinosaurs up to modern time, and it's all anchored around a town that is about an hour northeast of Denver. And so you just see the passage of time. And I particularly like it because it's so different than what I do during the day, but also a reminder that we're just here for a short stint and bring our best and do our best and uh life well lived. Just one of those is enough if you do a good job at it. And so I add some nice end-of-day grounding for me.

Rens Hayes 45:49

Paul, thank you so much for sharing your journey and being so open about kind of the path that Light Table is on. I, for one, see a ton of value in the problem that you're solving and have gotten to see your product firsthand. So if there's anybody listening that's in development, construction, or design, like absolutely to check out Light Table to consider it for one of your next developers. Hey everyone, thanks for tuning in to Design Development. Real quick before you leave, our goal is to help as many people as possible. We're a growing community and you're a big part of it. So just click that send button, send this episode to a friend to let them get the same insights that you got today. We appreciate you. See you next time.