Why Construction Tech Fails (and How to Actually Win with AI) | Robert Zimmerman

Rishi Srivastava (00:41)
Today, our guest is Robert Zimmerman. Robert, welcome.

Robert Zimmerman (00:46)
Hi Rishi, good to see you man.

Rishi Srivastava (00:47)
Definitely. Robert’s Construction Tech Executive and host of the CONTEC Exec Podcast.

The first section, Robert, here is on your career arc and credibility. You’ve gone from journeyman carpenter to construction tech executive. What moments in that journey most changed how you think about money, risk, and decisions on a job site?

Robert Zimmerman (01:13)
You know, most technicians, I’ll say technicians when I say that like trades people, they don’t think about, they think about money to them when they work hourly. My position changed on money when I started to own, when I owned my own company. And so that was about 10 years after being a carpenter and working out in the field, I decided to start my own company. And that changed everything. That changes your position on

You know how you’re funding things you you don’t realize you know what a difference it is between you know being paid hourly realizing the decisions that the boss was making or whatever and finally realizing the reason why he was making some of the decisions that were made that may not have made sense to you at the time. So that was definitely when I started my own company.

Rishi Srivastava (01:59)
Yeah, yeah, once you’re your own boss, it’s a very different world.

Robert Zimmerman (02:04)
Well, I always say, and I wish everybody could experience, but what it means to carry a payroll. Right. When you’re looking for collection, collecting money from jobs at the end of the week, maybe something’s not done. Clients are holding a check until you finish something. But on Friday afternoon, you know, everybody’s got their hand out like, hey, I worked this week, pay me. And sometimes I didn’t get paid.

Rishi Srivastava (02:12)
Hmm.

Robert Zimmerman (02:30)
And that changes your perspective really, really quick.

Rishi Srivastava (02:33)
Yeah. You’ve worn almost every hat, owner, PM, operator, consultant. Which role gave you the clearest view into why construction technology so often disappoints financially?

Robert Zimmerman (02:48)
I think consulting has given me a lot of clarity on that because I’m not only get I’m not seeing the obstacles from one side. So when you’re an owner, a lot of times you can’t see the forest for the trees. You’re trying to solve your problems and you’ve got so many other things going on just in the core business, whatever you’re trying to get done. And I don’t care if you’re

framing subcontractor or if you’re an electrician or whatever, you’re really focused on being very, very good at that. so construction has never been one of those types that says, you know, the technology that I’m using should really lead the way. It was always kind of, it’s everything goes, it’s still in the expense column, right? It’s not in an asset column at this point.

And so it’s still in the expense column. So you’re not thinking of that way. So being a consultant, I can see the side of the client. I can see the side of the contractor working. I see, hey, look, you may not see this now, but this software that you’re looking at or these three softwares that you’re looking at, here’s where you can divide up and figure out where it’s best going to serve you. Right. And so

That’s kind of what I do when I help people. I help people decide which one is going to serve the greater good when it comes to bottom line on the balance sheet, right? Or bottom line on the P &L. Which one is going to really serve the most function?

Rishi Srivastava (04:18)
was talking to a construction technology VC recently and he said like, selling software to construction companies is like selling Porta-Potti to construction companies. Now, he said like he’s gonna go build his own CMS and he’s gonna get all this AI software and do stuff like 30 % cheaper and 30 % faster. It’s like the world that we’re living in, selling software to construction companies is not worth his ROI.

Robert Zimmerman (04:47)
Well, you know, here’s the thing. Construction companies, contractors, they don’t have a lot of time and they don’t have a lot of time for nonsense. And unfortunately, in the technology world, there’s a lot of noise. There’s a ton of point solutions and there’s the sales guys that go right along with it. And I guess I just did an interview with a marketing team, a go-to-market team that works directly with Contec and I get

Rishi Srivastava (05:02)
No.

Robert Zimmerman (05:14)
their struggle because it is a hard job. It’s a very, hard job to get your product out there. If you’re a technology founder and you’ve got a good product, the problem is getting past all the noise. It’s getting past the point solutions to get to a contractor who may have 15 minutes to spend with you and say, why do I need your product? You know what mean? Because the last 15 of you guys that have come to my door or I’ve answered their call have wasted my time.

Whether it was they didn’t have all of the products or the product that they were looking for, the features that they needed were on a roadmap, but yeah, we’re going to get to that. Don’t worry. you know, they were, they had like this package of this great bow that said we do something with AI, you know, or, you know, they tried to put that, you know, pretty little bow on that pig and it just doesn’t work. So that’s the thing guys.

anybody that’s out there, if you’re selling software, and I’ve done that, I’ve been on that side of the world, if you’re out there selling it, guess what? You only got about maybe 30 seconds to get somebody’s attention, and then you might get another 15 minutes if it’s worthwhile.

Rishi Srivastava (06:21)
That makes so much sense. And we software vendors are also guilty of this. Like we got like 10 people writing the code and 40 people selling the code. I mean like that, come on. We need like it to be different. Like, yeah. It’s like, why do we need so many salespeople? You know?

Robert Zimmerman (06:31)
It’s backwards. That is correct. It’s backwards.

Well, again, software doesn’t do you any good if it’s not in the hands of a client. It’s difficult. Again, yes, and this is why I always say, know, VC funding or whatever type of funding that you can get for your software, go get it. Make your product as best as you possibly can. Get as much input on it as possibly can and make as many features as you can. Don’t wait. You know, that’s where you should spend that money.

make that product rock solid and then leave just enough left over, leave just enough left over to hit those conferences and hit those right spots for your target market right at the right time and get out there and hustle. Find the people that are going to hustle. So it’s difficult. It’s a tough balancing act to have if you’re a contech founder. Contractors, we are just difficult people because we just don’t have a lot of time for nonsense.

Rishi Srivastava (07:06)
Mm-hmm.

Robert Zimmerman (07:30)
And that’s just plain simple.

Rishi Srivastava (07:32)
love it, your point of view, The next section here is on technology versus reality in construction. A lot of contractors say, we bought the software, but nothing really changed. Why does that happen so often?

Robert Zimmerman (07:34)
Yeah.

I’m going to blame the contractors on this one, you know, that’s just… Honest truth is, contractors don’t change. Their processes don’t change. Most of them, their processes are broke. I just did an article about this. I call it the coercion economy. When you’re a contractor, you’re working with low bidder, you’re not changing up your processes and trying to find real value, real quality to deliver. I don’t care how much software you buy.

you’re going to be dealing with, you know, rework RFI’s galore. You’re going to be dealing with problems. It’s the same game. I call it the construction whack-a-mole. You know, you’re putting out fires. You’re trying to, what this one pops up next, you hit it. This one pops up next, you hit it. You’re a fireman. You can’t do that. I don’t care how much software you buy, AI is not going to fix it for you. Your problem is your inefficiency in your operation.

Rishi Srivastava (08:28)
second.

Robert Zimmerman (08:44)
So you got to fix your processes first. I’m not saying that if you do this, you know, fix all your processes, you’re not going to have any problems. That’s not what I’m saying. But what I’m saying is, is that maybe the way that you’re bidding, the way that you’re buying products, the way that you’re acquiring jobs, you’re getting maybe in over your head, those types of things, all of that affects the outcome of your process. So you really got to get sharp, right? About one, cashflow.

cashflow is king in the construction business. Two, you got to get sharp about quality of help in your business there. And then if you’re going to involve technology and data, you’ve got to get real strong about your data and how you communicate. Data is communication in today’s world, period. And so your definitions, how you’re defining things in your ERP, how you’re lining up your task codes, your numbering system,

All of that has to be spot on, consistent, and it has to be used across the board by everybody on the team, period. I don’t care if you’re a 20 person company or you’re a 500 person company. This is what you have to do.

Rishi Srivastava (09:52)
From your perspective, what’s the biggest myth contractors believe about digital transformation?

Robert Zimmerman (09:59)
The biggest myth that they believe is that they can hand it to somebody on their team and they should have the bandwidth to make it work. Not everybody is qualified to roll out technology. A lot of times, technology fails. 70 % of technology rollouts fail. And I think it’s simple because they handed it to one person or a couple of people and said, make this work. And that’s not how technology works. Again, technology is a tool.

that affects every aspect of your business. And so that means anybody that’s involved with that technology needs to have some say or some input. And the executives team. So again, if you’re a mid market contractor, specialty contractor, whatever, GC, your executives have got to be on board. You’ve got to get a team of people that know what they’re doing, that know how to roll a software. You may have to hire somebody that knows how to roll out software.

or hire a consultant that can guide your folks on how to roll out software. And then you’re gonna have to get some product champions. And then beyond that, once you get that ball rolling, guess what? Then you’re gonna have to monitor, again, goes back to your data, making sure that it actually is worth the return on investment. You can’t just stick your thumb in the air and figure out which way the wind is blowing on this. It doesn’t work that way.

Rishi Srivastava (11:15)
Actually, on this one, also want to blame a little bit on the software vendors. Sometimes they are just too prescriptive, like how to run their business. You walk in with the software, and you’re suddenly telling the construction company how they need to run their business. That shouldn’t be the case.

Robert Zimmerman (11:28)
Well, there’s a balance. It shouldn’t be, but let’s

be clear about this. Technology runs on sets of rules, right? And those sets of rules need to kind of be abided by in order to make the technology work right. mean, so unfortunately is, is that this goes back to kind of data governance. If you’re not calling things the same deal, you’re not, you’re not saying the same words, you’re not right. You’re not abiding by the rules and then expecting that software.

to and then expecting that software to function, it’s not going to happen. There is a fine line. There is a balance that has to happen between the software vendor and the contractor or the client, meaning the software vendor says, well, this is the way you should do it. Well, you’re going to have to adapt your software probably a little bit to other people’s processes if you can. Right.

you’re going to need to. But that means you need to have somebody, if you’re a software vendor, you need to have somebody that understands construction. If you’re going to be in the construction world, you have to understand where they’re going to need some wiggle room on these processes and these workflows. They’re not always going to do it the same way. And you’re not going to be able to tell them to do it just this way. They’re going to ask, well, what if we do it this way? And then you’re going to have to be able to work with them.

If you’re a software vendor, a founder, whatever, get somebody that knows construction that has been out there doing it, understands the workflow so that you can design your software to be a little bit. You got to be flexible.

Rishi Srivastava (13:00)
Yeah. Yeah. You know, as we are building our product, you know, every new feature we build, we want to bring it as an option. You know, we’re not trying to be too prescriptive. That’s one thing I’m always keeping in mind because you walk into like contractor one, they have certain process and you go to contractor two, they have different process and they like both of them. We want them to be able to work with us.

Robert Zimmerman (13:26)
And that’s not easy to do on the software side, right? That means you’re going to have to expand functionality quite a bit. that, and most folks, if you haven’t been on the software side of the world, they don’t understand that that takes a lot. I always compare it to, if you’ve ever been a plumber, if you ever plumbed with iron pipe. So iron pipe, you’re putting fittings together, couplings together.

Rishi Srivastava (13:41)
It does.

Robert Zimmerman (13:49)
And a lot of times what will happen is you get one fitting tight down here and then you’re putting on another fitting here and it moves and it loosens up this other fitting over here. Like everything that you do within the software may affect another piece of the software depending on how the code’s written. So you’ve got to be very, very careful of that. And that’s what makes it difficult.

Rishi Srivastava (13:59)
Mm-hmm.

Yeah, it’s not easy, definitely. So you often say technology should deliver clarity, not complexity. How do leaders tell the difference before they buy something?

Robert Zimmerman (14:22)
You gotta do your homework. First of all, you have to know your why. Before you buy something, you have to know why you’re buying it. And it’s not because your buddy or your competitor is using it and you need to keep up. You need to know why you’re buying it. You need to know what your processes are. That will eliminate a lot of the complexity because then you will know.

The why basically tells you what the outcome needs to be. So if you’re terrible with payroll and every week you’re rewriting checks or you’re getting it wrong or your prevailing wage payroll is coming up with bad reports and you’re in trouble, you may need some help in that area. So your selection, depending on where that’s gonna be, like if you’re struggling with prevailing wage and things like that,

First of all, you need to get somebody who’s competent in prevailing wage, but you also need a software that handles prevailing wage very well within that municipality. And you need to, you know, that needs to become that’s your clarity point, right? My clarity point says this payroll does all this stuff and it helps with this prevailing wage makes that easy. That’s clarity. But you had to have the why first.

Rishi Srivastava (15:27)
Yeah.

Also, having a free trial of a software is kind of important, I’ve started realizing. Because even though a lot of times it doesn’t work out for the software company, being willing to let the person test your product before fully committing is kind of important in my view.

Robert Zimmerman (15:46)
Yeah, I do. agree. You know, I think if you have a kind of a free version, a trial version of it, you could get it so, hey, we have some limited usage here just to get the basics down. But it’s also up to the vendor at that time really to be very clear. It’s like, okay, what is your why? What’s your kind of end result here? If your trial does not cover that, right, then you’re going to want to say, look,

Rishi Srivastava (16:15)
And…

Robert Zimmerman (16:16)
Our trial’s free and you can look at it, but it’s probably not gonna cover what you’re needing. But how about this? How about I spend, you know, and I don’t think it’s wrong for a salesperson or somebody that understands solution engineering a little bit on the software to go, let me spend a couple days with you on it. Let’s, and that is a, that’s a relationship development moment, really. If you’ve got somebody interested in your software and they’re interested in the detail, spend the time.

You know, this is, that’s a very, very warm lead. Spend the time with them to see where they need to go, what they need to do and how the software may function around that. know, not only will it help you as a solutions engineer and getting to a correct solution for somebody and getting you that knowledge, but it’ll help the client as well. They may not select your product and that’s unfortunate, but you’re going to educate them.

and you may be able to forge a relationship or maybe they need your product later, whatever.

Rishi Srivastava (17:14)
The next section here, Robert, is on AI jobsite readiness and decision making. You created the AI jobsite readiness score TM. What problem were you seeing in the market that made that framework necessary?

Robert Zimmerman (17:30)
Well, a lot of people.

A lot of executives, mid-size larger companies, they have a glossy view of whether their teams are ready for, and it just doesn’t have to be AI, it’s going to be digitization. Meaning they think they’re digital because they’re on Excel. That’s not digitization. And that’s not being ready for AI. But again, so what I like to do is I like folks to take the test.

Rishi Srivastava (17:50)
Thank you.

Robert Zimmerman (17:59)
and it’s just like 20 questions. And then I analyze that because what I get is I don’t just get the CEFOs or the CEOs point of view. I get the formants point of view. I get the office point of view. I get superintendents point of view. I can get some project managers point of view on what their structure looks like. Once I see that, then I know who is going to truly be able to participate in the functionality.

You know, nobody wants to tell their boss, yeah, it doesn’t really work like that. Most of the time when somebody is talking to the boss, like, yes, sir, it’s working great. We’re doing great. You know what mean? It could be a totally, you know, hey, are we ready for AI? You could ask somebody that directly and they’ll be like, oh yeah, I think so. I think we’re pretty close. I don’t know anybody that has worked for a company that’s gonna go, yeah, no, sorry, boss, we’re not.

I’ve never met anybody like that. I never have. I’ve never been in a situation like, boss, yeah, you’re you’re smoking crack. We’re not we’re not even close to being ready. So so again, for a CEO or a CFO to tell me, yeah, I think we’re great on our data governance. I’m just going to call BS. I just am. And I’m going to say, well, let’s let’s take the test. Go take the test. I’m going to evaluate the criteria and I’ll tell you where you’re at.

Rishi Srivastava (19:08)
You

Yeah.

Yeah. When contractors say we are not ready for AI, what are they usually actually not ready for?

Robert Zimmerman (19:26)
They’re not ready to get rid of their Excel database, their Excel spreadsheets, their processes.

I literally had a company I worked with. They have all of these great technologies, you know, kind of bolted together. And the way that they’re bolting them together is with Excel spreadsheets, right? It’s so it doesn’t really make much sense to say that, we’re ready for AI or

no, we’re not ready for AI. We or we know we’re not ready for AI because we’re we’re reliant on these these Excel spreadsheets. And look, I don’t like to bash Excel because I love it. Just like everybody else loves it. It’s easy to use, but it’s good for one person, maybe two to share. And that’s it. It’s not good for an entire company to rely on or to say that I can pull data from, you know, Dave’s spreadsheet over here.

Because once you’re tied into that thing and one little thing breaks, only Dave knows how to fix it, right? And so that’s a problem. And that’s siloed, that’s fragmented data. There’s a reason why they created APIs. There’s a reason why people do a lot of heavy work on these integrations is because you don’t want that unsecured piece of the data flow. You don’t want that.

It needs to be locked down so that you can perform an audit trail, so that you can look back and go, well, how can we validate that data? So you have to have that. And to be ready for AI or what AI will become for these companies, you have to have everything like that and that data structure locked down.

Again, AI functions on rules just like anything else, right? And the way that we train it is just as rule-based as, know, rule-based as we’re gonna play soccer, we’re gonna play by the rules, we’re gonna play baseball, we’re gonna play by the rules. It doesn’t matter. You got to set a rules for a reason and it’s to make the game or to make the business organized.

Rishi Srivastava (21:27)
so glad that you talked about like that weakest piece, maybe that Excel file that everyone is – relying upon. You know, making the systems connected is so important, right? And you have suddenly like a really bad integration. Suddenly the things are going to be not as reliable as you would hope for.

Robert Zimmerman (21:47)
Right, well, let’s put it this way. If you’re pulling a CSV file out of one program, one software, one platform, whatever, to generate a report, and then you’re dropping that report into an Excel sheet or whatever, immediately that data is subject to corruption.

because any of those cells inside of that data can be changed by whoever has access to that file. Even the person, maybe there’s one person. There is an accountability level struggle. But then also, when you take that and then you put it into another space and you have to manipulate it, maybe move a column around over here in order to get the CSV to fit, that’s another problem. It’s unstructured data.

Rishi Srivastava (22:34)
What are the non-technical blockers? Culture, data quality, workflows that prevent AI from delivering value.

Robert Zimmerman (22:44)
Non-technical. We kind of talked about this early on. Non-technical blockers is some of our procedures that we have. Like said, I just wrote an article on it. It’s called The Coercion Economy. When you’re using lowest bidder, that’s not a technical thing. That’s the way that we’ve structured business from kind of day one. I think it’s interesting that we will pay lawyers $300, $400 an hour, I don’t know, to

you know, write up these contracts. But then for the actual build, we are trying to find the guy that’s charging, you know, that’s paying these guys 15 bucks an hour that don’t know anything about pouring concrete or they’re, you know, they’re, they’re in way over their head. They don’t have enough to cashflow the job. These types of things. We are willing to pay top dollar for the legality of it all, but

We wouldn’t have to worry about the legality of it all, quite honestly, if we actually got quality work with quality people who knew how to build. And that’s the non-technical piece that we fail at every time, or not every time, I’ll say a lot of times we fail at discussing the project with people that know how to build before we start structuring a contract, before we start putting a schedule together, before we start.

value engineering a job.

Rishi Srivastava (24:02)
You know, one thing I love about AI right now, Robert, it’s like, it’s going after the lawyers’ jobs. It can do lawyers’ jobs better than lawyers. So we needed like, let’s say 10 million lawyers in this country. Soon we will only need a million.

Robert Zimmerman (24:17)
Yeah, it is funny. There’s a level to that. It will get rid of lot of, and I’ll say middlemen, people who have kind of been in the way for a while in the industry. Lawyers have been what we’ll call necessary evils. And some of my best friends are lawyers. I can’t bag on them too much.

Rishi Srivastava (24:37)
Hehehe

Robert Zimmerman (24:39)
There’s a lot of learning that has to be done, a lot of education you have to have, you have to be specialized in. It takes a lot. And it’s no disrespect to them. But yes, there is a lot of area there that can be, a lot of ground that can be covered with AI that may eliminate some of that work.

Rishi Srivastava (24:40)
Yeah.

Robert Zimmerman (24:58)
I’m still, I got to be honest with you, Rishi, I’m still kind of up in the air on the whole thing about, you know, AI replacing people. I’m hoping that we can keep it as a tool rather than a replacement part because again, I feel like people are the most important part of this whole process, whether it’s in construction or in business, it’s really about the relationship and about solving problems together.

Rishi Srivastava (25:06)
Do

Robert Zimmerman (25:26)
Yes, AI is a fantastic tool and can be used for great things, but I do not really want it to replace a bunch of people just because it saves a few dollars on the bottom line. But that’s just me and that’s kind of my personal opinion. But you have to be ready for it.

Rishi Srivastava (25:40)
I see.

Yeah, yeah.

I’m also going through something personal in my life, know, a custody case, and I have a lawyer too. So it kind of lawyers dealing with them is tough and I hope it works out.

Robert Zimmerman (25:56)
⁓ It’s necessary

evil, right? Yeah, it’s a necessary evil. You don’t want to have to go to a lawyer, but guess what? If you’re going to be in business, need the Robert Kissimmee-Ocky thing. You need bankers, lawyers, and accountants. Those are the people you’re going to need in your life. You have to understand that.

Rishi Srivastava (26:16)
Yeah. How should CFOs and finance leaders evaluate AI tools differently than project or IT teams?

Robert Zimmerman (26:25)
Well, I think a lot of times what happens is they don’t, it’s very difficult to measure soft cost, right? And a lot of times, technological tools provide benefits in areas that don’t get measured financially, soft cost areas. If you’re going to be adding tech,

Yes, I believe it does need to add value, but you can’t, again, you can’t just stick your thumb in the wind and go, yeah, I think it looks pretty good. You really do have to find a metric and stick with it to say, A, yes, that metric has improved since we started using one thing. so you have to have a baseline. You have to develop a baseline. And so I think that’s how finance folks on that side of the business can then justify.

you know, technological spins. It’s not easy. I’m not saying it’s going to be easy, but you’re going to have to do your homework and you’re going to have to find out what metric works for you that you want to, you know, say at the end of the day, this was worth it.

Rishi Srivastava (27:32)
A lot of times, you your gut feeling about a tool that you’re about to acquire is very important. Like it’s very hard to measure everything. And if your gut’s telling you, this is not going to give me enough ROI, you’re probably right. No matter how much mathematical analysis you do afterwards.

Robert Zimmerman (27:48)
Yeah, and I always say when you’re measuring something, you don’t try to measure everything. Measure one thing. ⁓ One of my first bosses was saying, you know, if you just measured one thing on a job, it would probably tell you a lot about the whole job, right? Whether that would be productivity or clock in, clock out times, those types of things. Like if you really kept an eye on one thing, it would probably tell you a lot about the

Rishi Srivastava (27:54)
Thanks.

Good.

Robert Zimmerman (28:15)
the culture and the dynamics and what was really going on in that job. I mean, stop and think about it. Say if you’re tracking attendance on a job and you’ve never tracked attendance before and then you start to see trends and you’re like every Friday there’s one guy is missing or a couple of folks are always missing on a Friday. kind of, you know, and maybe the only person that has ever seen that before is the foreman, right? But they kind of let it slide because maybe they need that guy or they need that person.

Once you start measuring it, then you start to get a really good view of, okay, maybe this is how it’s always going. Do we like that? Do we not like it? Are we going to be okay with that in our culture? And think a friend of mine said it best, he was quoting somebody else that says that, you know, as far as culture goes, it’s saying, you know, the things that you let slide, that’s your culture. So it’s interesting. And so I say the things that you…

The things that you measure will be the measure of your culture and the measure of your output.

Rishi Srivastava (29:08)
Next section here is on data, ERP, and financial truth. You’ve worked across BIM, ERP, and field productivity systems. Where does financial truth usually break down between the field and the back office?

Robert Zimmerman (29:24)
I’ve been, you know, my expertise was probably more in that payroll, you know, the hourly time tracking and productivity. I see a lot of breakdown there. As far as…

Rishi Srivastava (29:33)
Mm-hmm.

Robert Zimmerman (29:35)
As far as, you know, what specifically, you know, I can’t, I can’t tell you it’s different for different companies, but

There’s still the manual kind of transfer of information that gets into the digital system from the field, right? It’s Joe had four hours, such as such had three hours, you know, and then they’ve got budgets that maybe they work in. Well, it’s very, very easy to say, well, I’m a little low on that budget. I don’t want to put those hours on there. So I’m going to put it over here where I’ve got some more hours.

And so what happens is that starts to get skewed a little bit. So maybe your hours don’t really represent what’s really happening on that said cost code. And maybe it’s one of your biggest moneymakers, right? Or maybe it’s one of your biggest money losers, but you wouldn’t know because you’re not necessarily getting exact or truly accurate because somebody doesn’t want to look bad during the week, right?

And so they or they just don’t want to have to deal with that phone call. Why are you over four or five hours on said cost code or why are you over 40 hours on said cost code? So they don’t want to deal with this. So they just shuffle the numbers. I call it the number shuffle. So, again, that is that’s where that can happen. And so sometimes from a financial, you know, ERP standpoint, if you’re sitting there trying to crunch some numbers and figure out.

Rishi Srivastava (30:50)
Mmm.

Robert Zimmerman (31:02)
what jobs or what is your limit? Like what jobs are going to make us money and how small can we go? You know, should we take a $50,000 job? That probably doesn’t make us any money. We should probably only take the $100,000 jobs and up, right? So again, if you’re crunching numbers like that and you’ve got some skewed data in there that you’re crunching from, you may not be getting exactly what you think you’re getting.

Again, it always comes back to these humans. These humans are, we’re a problem.

Rishi Srivastava (31:28)
lot

Yeah,

the data needs to come from the humans into these systems.

Robert Zimmerman (31:37)
Yes.

Yeah, and it has to be accurate. And so you got to do your best to try to make sure that it’s accurate. So, yeah, that’s I think that’s that’s the ultimate breakdown.

Rishi Srivastava (31:46)
Yeah, also like the incentives of the field people sometimes are not necessarily aligned with the incentives of the back office, right? So that could create issues where fields not trying to look bad on certain cost code, you know, and trying to fuzz numbers.

Robert Zimmerman (32:01)
Well, you have to understand too, what, what incentive is to somebody may not be incentive for that person. I’ve worked in the field a long time and I know a lot of these guys and you know, had some of the same mentalities early on was I just, I wanted to fill out my time card and go home. I want, know, I had beer to drink and fish and to, you know, fish to catch, you know, those kinds of things. Those are the things that, so my incentive was get this thing done as fast as I could and, head out. Cause

I could really care less. They don’t understand the impact, right? Again, it doesn’t really, it looks good from their house. That’s the kind of attitude. So that’s what you deal with.

Rishi Srivastava (32:33)
You

Yeah.

So well said, Robert. Why is data governance such an unsexy topic, but such a big determinant of margin and cash flow?

Robert Zimmerman (32:51)
Because it’s called data governance and anytime you put govern in something everybody thinks government and everybody runs in the opposite direction. They just do it. It is a you want to fall asleep every time you hear it. You don’t want to talk about it. It is a boring job to try to figure out how you’re going to name files, how you’re going to

Rishi Srivastava (32:58)
You

Robert Zimmerman (33:14)
you know, use the same definition and the same words for things that are going to become your data. That is a boring job. So you have to find somebody who

You have to find somebody, have to find the right people that dig, you know, organizing that really are, you know, exact in their detail to really say, you know, this has to be this way. But then there’s a balance to that. You also have some have to find somebody that says this is good. This is done right. There’s no such thing as perfect. It has to be done right. So you have to have to find that little bit of a balance in that person.

I think a lot of times what people get into is when they get people like that that are very exact and very detailed, they get way into the weeds and they can never finish. Data governance, the only way that you’re going to get to good data governance is to say that 80 % is better than zero, right? You’re always gonna be tweaking a few here and there. It’s never gonna be perfect. Get the people that are okay, get it.

Rishi Srivastava (34:02)
Good.

Robert Zimmerman (34:20)
Hey, is it 80 %? Do you think we’re 80 % there? Good, we’re done. Stop. You know, we’ll move on down the road. Because right now, if you don’t have any data governance and your stuff is all over the place and you’re totally unstructured in your data, you got to start somewhere because right now you’re at zero.

Rishi Srivastava (34:37)
What do you think if we hire a bunch of DMV people for data governance?

Robert Zimmerman (34:41)
You can’t hire C students for this kind of thing. So they have to be smart people. Nothing against DMV people. No, you can’t. You can’t. Again, that’s not it. Somebody has to be interested, you know, data analyst, somebody that says, OK, I know how to get this right. This is how we structure it. So that way I can read it afterwards. So you have to be.

People, can’t be, it’s not a data entry job, right? It is a analytics type of job, I think.

Rishi Srivastava (35:10)
Hmm, makes sense. If a contractor only fixed one thing about their data before investing in AI, what should it be?

Robert Zimmerman (35:20)
Hmm.

I think standardizing your cost codes and standardized cost codes, kind of that’s base. That’s where you make your money. You make your money on your standardized cost codes. So that way there is no question.

Rishi Srivastava (35:24)
Mmm.

Robert Zimmerman (35:35)
Beyond that, standardize your cost codes and then standardize the how you collect it and enter it. And let everybody know that, you’re not going to get in trouble if you do it accurately and you’re over on these hours over here. Again, you’ve got to be able to get the raw data that is correct to find out where your weak points are. standardizing cost codes is, I think, number one.

Rishi Srivastava (36:00)
makes so much sense.

The next section is on ROI, risk, and executive accountability. How should executives think about ROI when it comes to construction technology, especially tools tied to finance and operations?

Robert Zimmerman (36:16)
I think they should let the delegation happen to the people that are going to be using the tool ⁓ and let that input drive the tool selection.

Rishi Srivastava (36:22)
Hmm.

Robert Zimmerman (36:26)
I think I said it before, a lot of times what people do is they put the wrong person in charge of a person or persons in charge, maybe because, well, they’re part of the IT team, so that should be software. And I don’t necessarily agree with that. IT teams, yeah, they make sure that the software is secure and they make sure that there’s a lot of, that it’s installed correctly. sometimes they get into a…

understanding how the software works, but the workflows inside of those, those are best understood by the people that are going to be using the product. So you’ve got to have a collaboration there. If you’re going to give it to the IT team, then make it in conjunction with somebody in operations, somebody that’s in the field. Get a team together of product champions that can then really make the product work well.

Rishi Srivastava (37:18)
A lot of times IT team doesn’t know the internal processes of ops and finance teams. They’re there installing stuff and helping out in some IT stuff.

Robert Zimmerman (37:30)
Well, I mean, I’ve worked with several IT professionals and great guys and they know their systems and they know their stuff. But most of them didn’t know construction. They didn’t know, you know, guess what? It doesn’t look like it looks in the movies, you know. Construction is a difficult, very, very stressful job when you’re out there in the field.

If you’re a foreman and you’re trying to use a tool that’s digital, or even if you’re a project manager, you’re getting 15 phone calls and you’re trying to put out X amount of fires and then you’re trying to schedule products out here. So if your tools are in a mess and they’re not functioning, that’s the…

never seen more broken out iPads in my life, you know, when guys just when it doesn’t work, right? They are just throwing them across the room because they’re worthless, you know. These people have tempers. We do. And we were under deep stress trying to get a job done. And then when your tools don’t work, and I don’t care if it’s a drill, you know, whatever that it’s very frustrating.

Rishi Srivastava (38:23)
Thank

There’s a reason we call these things a tool. They need to work when we need them.

Robert Zimmerman (38:37)
That is correct. And there is such a thing as the right tool for the job, right? There is such a thing. Well, if I ever see you use a chisel for a screwdriver, Rishi, I’m gonna get on you.

Rishi Srivastava (38:48)
You

Yeah, thank you for saying that. Actually, I’m about to go and do some wood cutting and I’m going to use some saw. So my friend was like, watch the safety video before you come.

Robert Zimmerman (38:51)
Yeah, you’re what you’re

Yeah, dude. Yeah, definitely watch the safety video. Try to keep all your digits intact. That’s always a good plan when you’re working with those things.

Rishi Srivastava (39:07)
Yeah.

In your experience, what’s riskier, moving too fast with new technology or waiting too long to modernize?

Robert Zimmerman (39:17)
I think waiting too long to modernize is probably a big problem because technology moves so fast in itself. If you’re waiting too long, you were never really even in on the ground floor of a product to know kind of how it functioned from the beginning. So I think there are areas where it’s like, you’re just going to to bite the bullet, jump in and go for it.

Rishi Srivastava (39:23)
Yeah.

Robert Zimmerman (39:43)
when it comes to tech. If you’re running a company on QuickBooks and you’re planning on trying to run more than 10, 20 million out of that QuickBooks thing, and I’m not saying that you can’t, I’m just saying you’re gonna start to run into problems. Nothing against QuickBooks, again, it’s great tool for small business and it’s a great place to start, but you’re gonna have to then get into an ERP and then don’t spend a lot of time. There’s plenty of good ones out there.

I always say with the ERPs, work with the one that wants to work with you. I think that’s the best way, because I think that’s where you’re to get the best support, which is what you need if you haven’t been in an ERP before.

Rishi Srivastava (40:14)
dinner.

In personal life as well, a lot of times people are chasing other people who don’t care about them and that this other person right here, she’s right there for you always and why are overlooking her? And trying to go after all the other people who don’t care about you and don’t respond to your messages.

Robert Zimmerman (40:32)
Yeah. Yeah. Yeah. And

yeah, I get it. And again, it’s a relationship business. And, you know, a lot of times the support teams

You may not be working with them upfront. You might be working with a great salesperson who’s there for them. What you do have to find in the background is you have to find out how well you’re to work with support, meaning how well we’re going to work with a team that we’re in basically have a relationship with after we buy the product. That’s what I want to know. Like you can send your salesperson to me all day long and I’m like, okay, I hear you. It’s a tough job. I get it. I’m going to give you a listen.

But before I put my money down on this thing, I want to talk to the help desk. I want to talk to the people who are going to be supporting my product. Because I want to know if I can get along with those folks. Because if I can’t get along with them and I can’t get you on the phone, or if we can’t exchange emails without it taking three days for you to get back to with me, that’s a no-go for me.

Rishi Srivastava (41:32)
It makes sense to me. And one thing I’ve personally noticed in construction technology, maybe it’s just this industry. Like there’s just too many salespeople versus like support or product people.

Robert Zimmerman (41:43)
Well, don’t know that, you know, support departments, I feel for them because they’re often the complaint department, right? You don’t really get the good news in the support department, right? You don’t always get the attaboy there. So that’s a tough job too. So again, you got to have thick skin, just like, you know, sales, you got to have a bit of thick skin. In the support department, you got to have thick skin, but you also have to have the attitude of…

I don’t care how bad this person treats me, I’m gonna fix this problem.

Rishi Srivastava (42:16)
Yeah, yeah. But I think, again, as an industry, as a construction tech, implies far too many salespeople. That’s my point here. We need to kind of cut some of these salespeople and help more support people.

Robert Zimmerman (42:30)
Yeah, I think there is a lot of marketing. I call it the noise. I think there’s a lot of noise out there. But you’re never going to get, I don’t think we’re ever going to get there because ⁓ people want to be a part of the gravy train and that’s a big part of the gravy train.

Rishi Srivastava (42:39)
with it.

The last section here is looking forward. And this is the last question as well, Robert. Over the next five years, what will separate contractors who win with AI and technology from those who fall behind? And what should leaders be doing right now?

Robert Zimmerman (43:02)
I’ve often said it, I think there are three pillars to this thing. There’s not one thing that they should be doing. The first one, of course, we’ve talked about it plenty, the data governance portion. You’re going to have to be able to communicate with whatever AI you’re using, whatever digital technology you’re using, and it’s going to have to be consistent so that you can understand whether you’re getting an ROI out of it.

Two, you’re going to have to train your people to also understand what that language means, what that communication level means, what data governance means to the company and to them as they move forward in their careers. Because AI is not going anywhere. Digital systems are going to become more more prevalent in their lives if they’re already. And then I guess thirdly is

understanding that the data that you’re putting in like productivity data, that type of raw data, not monetary data, but that type of raw data, then we’ll be able to validate the money collection that you’re doing at the end of the day. You have to have kind of almost trying to triangulate whether you’re being successful or not, because obviously you can make money on a job, but does that mean that you made money because you had a lot of float in it?

or is it because your margins were too high or you had a buddy deal and they were just gonna give you the job anyway. So there’s a lot of factors there. So you’ve kind of got to triangulate in on why you’re making money. And then that goes back to that data governance, training the people to understand why it’s important to get it right and then measuring the right things.

Rishi Srivastava (44:39)
And looking at the amount of investment these companies are getting, like OpenAI just raised $120 billion. It’s like we’re not talking about a million, $120 billion. I mean, the kind of investment that’s going in here. Go ahead.

Robert Zimmerman (44:50)
Yeah, I’m still wondering. Yeah,

I’m still wondering if that is, I mean, obviously that’s not a drive on demand for the product at this point. There’s something that’s driving it. I don’t know what it is, whether it’s some sort of military, you know, national security thing. There’s something driving, driving this bus in this direction. I don’t know if it’s a wise direction, but there’s something driving the bus. Obviously this

When people are investing that kind of money, they think they’re going to get a lot of money back out of it. So like I said, it’s not going anywhere. know, we’re not going to invest trillions of dollars into something and go, it was a good, it was a good try. This is just not going to happen.

Rishi Srivastava (45:32)
Exactly. Like when you start putting that kind of money, you know that this technology is not going to go anywhere. You know, so we better try to learn how to live with it.

Robert Zimmerman (45:38)
That’s right.

And so that’s why I say with construction companies, train your people, understand why data is important. You understand why data is important. Everybody else should too. We need to understand that data is the only way we’re going to be able to communicate with this, what I call the age of AI. And to communicate correctly, if we’re giving it bad data, we might as well be raising an unruly teenager, you know, because

giving it bad data and expecting it to do something fantastic is just a terrible, terrible idea.

Rishi Srivastava (46:14)
Garbage in, garbage out.

Robert Zimmerman (46:17)
Right. And like I said, you raise your kids and you try to teach them right from wrong. And so you don’t, you know, you don’t put up with too much, right? You, you have to give them the right information so that they understand how they use it because they’re humans is going to be different. But AI pretty much going to take you at your word. So that’s why you, you know, you see AI, people say that it hallucinates. Well,

Yeah, because there is a lot of garbage out there that it’s been feeding on for a while and you wonder why it sometimes makes things up or whatever. Well, guess what? It sees you do it. Why wouldn’t it do the same thing?

Rishi Srivastava (46:58)
I love it, man. I enjoyed chatting with you, Robert. Thank you so much for coming onto the show.

Robert Zimmerman (47:04)
Always good talking with you,