Success Stories: What Good Looks Like

Fraser Gallop and Steve Gross break down real‑world examples of modern data transformation in construction—moving from manual month‑end chaos to automated data warehouses, governance, semantic models, and actionable dashboards. They share stories from the field, highlight how top performers streamline WIP, forecasting, and reporting, and show how better data architecture frees high‑value teams to focus on strategy instead of spreadsheets. A practical, insight‑packed conversation on what “good” truly looks like in construction data.

Learn more about Steve’s work with eCMS Cloud Construction ERP Software at  Computer Guidance Corporation or Fraser’s work transforming construction data into actionable insights at Onware.

Transcript

Fraser Gallop 0:11

Welcome to Hard Hats and Data Chats episode six. This is Success Stories, What Good Looks Like. I'm Fraser Gallup, and I'm here today with Steve Gross. Steve, what's new?

Steve Gross 0:24

Well, we had a milestone birthday in our family uh last weekend. So we had a big party for my mother-in-law, and lots of relatives flew in, and it was a big deal, big family event. So we're still uh recovering from that.

Fraser Gallop 0:41

Oh, so you had you had a bunch of uh house guests and that kind of thing.

Steve Gross 0:45

Oh yeah, absolutely. Yes, yes. Uh a lot of we had a catered and coordinating all that, and uh my wife was was really busy. I was too. So anyway, it was it was worth it though. It was it was a memorable event.

Fraser Gallop 1:00

Oh, that's good. Uh we did uh when my grandmother turned 90, I think it was 90. That we had like a big kit together and hadn't seen cousins for for many years before then. So that's that's always fun to do those kind of events.

Steve Gross 1:16

Absolutely.

Fraser Gallop 1:18

What about you? Uh well, I got back from from my my Paris trip, uh, went and did the tourist thing. I mentioned last time my recommendation is to book ahead and get the tickets ahead of time. We showed up at the Eiffel Tower without tickets relatively early in the morning though that we showed up. And uh they they had on the big digital sign saying top is open, wait time is about one hour in the lines. With okay, that's that's reasonable. We'll we'll line up. So we we lined up and about an hour later we were at the front of the line to buy tickets. By then the top was closed because they were at capacity. So we got to go to the second level, not the top level, uh, which was still fun and and worthwhile, but uh back to the the the whole thing from last time. And if you if I would have bought the tickets a month in advance, we would have been going to the top because we had the tickets in advance. Either way, it was a good trip. We did, we did the tourist thing, had a really good dinner. Um, so that was lots of fun.

Steve Gross 2:17

Sounds like a great time.

Fraser Gallop 2:19

We said today would be success stories, what what good looks like, some real world examples of what data transformation um look like. Before we kind of get into the the after and what companies are doing to to to put uh a modern data architecture in place, let's talk about you know what the current state. I think the the first thing is manual reports or run-on-demand reports where people are are going and running things and typically gathering that into an Excel uh or some kind of intermediary format before they're they're compiling their reports. Um you had mentioned before uh one of the companies where they had month end and uh how did they handle that?

Steve Gross 3:05

Yeah, and uh, you know, it's WIP is an area in construction. It's amazing how manual it still is to this day. It involves lots of compilation. I guess that's the word I can use, is uh just pulling information from different places into one spreadsheet, you could say. And that becomes the basis of what your financial statements are based on. And uh so there's lots of opportunity to streamline that and automate that process in construction, in my opinion.

Fraser Gallop 3:37

Yeah. You're storing that static information in Excel somewhere, right? Whether it's in a SharePoint server or shared office server, capturing at that point in time. And if things change, like if somebody realizes they missed a bunch of postings into the GL, it's start the process again, right? And or redo the process to to kind of have to account for the yeah.

Steve Gross 3:58

Well, it's it's the accounting data and it's a buildup process, you know. I mean, uh all the all the accounting data needs to be there for sure as of the cutoff. That's that's number one. Number two is the field inputs, forecasting is the primary topic there, you know, and project manager inputs into the how their job is performing. That should flow in as well. And then from there you roll up and you might have further adjustments you bake for conservatism purposes, or if you have a margin target that you're wanting to live up, live by, you know, in terms of how you're reporting your work and your financial status. And so all that, it just takes a lot of buildup, a lot of work up, and uh really contributes to the um not only for whip for revenue recognition and whip reporting, but you know, that's a key component to the month in close. That that drives the month in close. So the whole thing is related. So really, you know, when you're doing it manually, that that that's a pretty arduous process. It can be, you know.

Fraser Gallop 5:05

And I think you've probably come across the same thing as us. You know, there's lots of cases where it it is like a CFO or high-level people within the company are the ones that are running these manual reports to to put it together and assemble them. It's not cheap labor, right? It's your expensive labor is the one that's analyzing the stuff.

Steve Gross 5:23

These are high high-paid people and they're smart people, but they're doing some pretty manual stuff that they probably learned when they were working for for a public auditing company or a public firm, you know, and they've way past that. They should be spending their time on other things. Learning how to make you more money and that sort of stuff, you know. So, I mean, it it really pays dividends when you automate this sort of thing because people overall, I'll say let me say this. I think this is the theme. Overall, the the people involved, the players, project team, the finance team, everybody starts playing at a higher level. They're the obvious things are out of the way. They don't have to deal with them. The easy stuff. You're dealing with the more complicated things that take think work and analysis, and it frees them up to focus on that sort of thing. And that's that's really the dividend that's paid by automating this, these processes and really fine-tuning your reporting capabilities and pulling it all together for people.

Fraser Gallop 6:22

Let's talk about how that comes together and what that maybe looks like. When we're talking about modern data architecture, the first thing on the list is this idea of a data warehouse, right? So we start getting data from the source systems like the ERP into a database where the analytics tools and the other systems can have access to that data very easily. If it's a third-party software, um like a payroll system, maybe you're using things like APIs to go and get that data and retrieve it on a regular schedule. It is all schedule-based, automated, so that when the data's moving from one system to another, it's it's not something that you have to think about once it's set up, right? It's it's kind of moving automatically. When I say what good looks like, good is just getting that set up and in place. Great is next level, it would be like streaming data. So you're doing a transaction, like you're doing a journal entry in your ERP, you're seeing that journal entry hit your data warehouse within 15 minutes time. You're not even waiting for like next day refresh schedules. It's just it's happening in almost real time that it's coming through. We've set these up in the past. We've got lots of customers that have set these kind of things up. It's enabling and collecting all that data in in one place so it's there to act on and use it. So instead of going and running those manual reports from three different systems, all that information is sitting in your data warehouse already for you.

Steve Gross 7:49

And that becomes really powerful when you start segmenting that by like period end. So you have this ability to go back and look at certain metrics at the end of each period in the past, which paves the way for trend reporting and as you know, that sort of thing. And that's really where a data warehouse can make a big difference.

Fraser Gallop 8:09

And that's a really good point because maybe source systems don't necessarily take those snapshots, like point-in-time data that you need to have and you need to be able to rely back on. The data warehouse gives you a place to do that kind of stuff, and it's not a static file sitting on a computer somewhere. It's in a centralized, trusted place that's backed up. Right.

Steve Gross 8:32

And and you're pulling in information from disparate systems, possibly, depending upon what you're what you're what you have to work with. If it's not all one system, then a data warehouse becomes even more important because you're combining different sources of data and into one result.

Fraser Gallop 8:50

So that's kind of the base layer. And then on top of that, what good looks like, let's go back to the theme, having some kind of data governance. What's happening to ensure that you have quality data in your data warehouse? There's tools that allow you to do this where if something kind of goes out of range, it's gonna alert the right person to say you may need to take action on this. There's something unexpected. Like a GL transaction was was posted for the year 2036. Who's gonna get alerted when that happens to go and and and fix it? A really great example uh that we had a customer, they had made this exceptions uh report and had a bunch of things that they had identified that uh they wanted their project controls people to be notified on if their jobs were set up in a certain way. A good example of that was just having the correct tax codes set up on a project. You have the job and you put in the address where it's located, and then this uh tool would go out and it would input that address to kind of come back with what are the tax codes for that location and just validating that it was set up and configured properly. If it wasn't, it would tell them and then they could go in and correct that. Do you have any other kind of examples of that where there where there's like a really good kind of data governance in place?

Steve Gross 10:08

Yeah, I think um like AI is really great at that because it can assign um a probability of of accuracy to things, and you can get reports out and exception listings out that say, oh, this is 90% accurate or probable, 90% probable of being correct, or 50%, you want to look at those, you know. Time saver. And uh, you know, good systems validation routines will capture and trap those things and prevent them. But you know what? It's still there are exceptions to everything, right? So if any uh efforts made to clean it up further using tools like that really, I think will help out.

Fraser Gallop 10:47

No. All right. Uh next one was uh data marks. We keep building things up here. We've got our data warehouse, some governance on that data in our warehouse. And the next one would be this idea of data marts. What that's getting at is this idea of you've standardized your naming. So the the way you refer to a job, a subjob, certain calculations and terms across the business. You went through that process, standardized that, and and everyone agrees that this is how you refer to it. That's implemented inside your data models. The term that we hear increasingly with AI today is this idea of a semantic model. So you're not just taking that raw data out of your ERP, you're adding some additional context to it so that the AI knows about your data and how that's used. Um, but even without AI, you're standardizing things so you can use your data across the business. A great example would be a payroll and time cards and this idea of what's the source of truth, what's that kind of golden record that you have for your employee, and making sure you can use that across all your integrations, all of your reporting. Uh you're not having three different sources for that employee's name and address, and you just have that one uh standard place.

Steve Gross 11:56

Right.

Fraser Gallop 11:57

Um, but that's kind of the the thought behind the data mart is to try to keep that consistent um across everything that you're doing.

Steve Gross 12:05

Makes sense. Yeah. And it's like your template is what it how it should be organized at naming conventions of the columns and so on.

Fraser Gallop 12:14

Exactly. Uh on top of that, uh, that's where we would build our dashboards and our reporting. Very often, what does good look like is you're going beyond just outputting data uh as you see it in the ERP. If you just want to be able to like go look at transactions, your ERP system does that just fine by itself. Uh if you want to go beyond those transactions, like layer on that additional data that's in your data warehouse and having that as part of your drill down, that's where you're really getting to be good and great. One of the uh desires that so many like PMs and executives have that uh we kind of hear all the time is this idea of being able to click through for details. We've done this safety reporting, for example. Like if there's a safety incident that shows up on the dashboard, being able to click on that safety incident and pop up your safety system, go strict straight to that record so that you can see the details of it and and and read about it. Do you have some other examples of that kind of click-through idea where people are kind of setting that up so they can go from their reports down to details?

Steve Gross 13:18

Yeah, that's it's really useful, especially if you're a visual person. I mean, often seeing a picture of the source document jogs your memory as to what that was about. And also while you're at it, if you have uh approval notes on there, you know, if it's been through an electronic workflow and someone's approved, if you have, well, who approved this? And if there were, they had any questions. Good approval systems will also include the ability to archive any email correspondence about that particular transaction. So to be able to get to that and drill into that and see what what you know, the QA on the item before it was approved, the extenuating the context uh of what the item is is available that way and really makes it the whole, if you can drill into that from a from a graphic visual dashboard, wow, that's kind of soup to nuts. It it catches your attention and then you can do your own analysis right by clicking and getting to getting to an answer and putting two and two together. And, you know, and it helps you determine what your next step might be uh quicker than having to go open up another system or look at another report.

Fraser Gallop 14:21

So on. That's actually that just made me think. I was actually just uh on a call yesterday uh where we were talking about equipment on a job site. And PMs, they'll know if this piece of equipment was not on their job site. And they'd be like, Well, why are you charging me for this piece of equipment? Right. And so if if you do have that kind of audit trail available to them in their dashboards, they can say, Oh, yeah, it's for X, Y, and Z reasons. Yeah, I know that call. It wasn't on my job. Yeah, exactly. So yeah, giving the giving that context is really important. Um, another piece of what good looks like is making sure that your dashboards and the stuff that you're implementing outside of your ERP is respecting things like job level security. Right. Because in the ERP system, everything is very controlled. You've got really robust security, so people should only see the things that you know they're they're allowed to see. People can can go off and implement dashboards that go around that security and not think about that, uh, where you're maybe exposing things like salary information for for people that don't want their salary information exposed. So part of the implementation of that data architecture is being mindful of the security that exists in the other systems and making sure you don't lose that as part of enabling that access.

Steve Gross 15:41

Right. I agree a hundred percent. Yeah, the more uh the more you involve people in the management of a project, suits and outs like cash flow, receivables, the more they know about the business, they will have interest on how other jobs are doing, you know. That can be that can be good or that can be bad, depending upon your your organization. Sure. So these things become important. This ability to secure things at at that level is really critical for anybody that's really diving into that level in detail.

Fraser Gallop 16:15

And then um the last piece that I want to include in this idea of the modern data architecture is just this idea of the the pulse metrics. And I think I probably mentioned this before. Connecting the right people with a daily tracking of job hours and production and getting that information available on their phones, right? So that they or you know, it's a night, nightly email that kind of comes, but it can often be way too late if you're looking at it a week or a month after the fact, if you're going way over on the hours and the budgets on a job, enabling those kind of things that give you that pulse of the business and the pulse of what's happening on site, uh, even though you're not physically present there and you can see that data coming through very quickly. That's four or five bullet points there of the modern data architecture. What does that look like and the different pieces that people are putting in place? Once you get that in place, the next piece to talk about is what does that enable? Because before we had our manual reports running things on demand. Once we put all of this into in place, what comes next? Well, the first one, of course, is project managers have visibility, right? They they can see what's happening a lot faster on their jobs. An example that keeps coming up is this idea of uh giving them the receivables information. What's happening with the invoices? Are they paid or not? Uh, what's another example?

Steve Gross 17:33

Well, right along with receipt, if you're tasking a project manager to be responsible for collections on the job, you're also most likely their marching orders are to maintain a certain margin, certain profitability on the project. So being able to show that, that's part and parcel of the WIP discussion. You really can't show accurate margin unless you have your ducks and row in terms of how you're recognizing revenue. Their forecasting drives that, and in turn, it'll show what margin is. So it's just part of that whole picture. And so that would be another job performance margin, uh, productivity levels is another thing. If you're capturing quantities against man hours, wow. The minute you do that, it's like the third dimension. You can it just you're you're going from managing cost to managing crew productivity. And like if you're a mechanical or electrical, crucial, crucial metric. Another thing would be um man hour scheduling or um man-hour requirements. If you're if you have the mechanism to forecast what your labor needs are for a new project you've won by period, that drives your hiring decisions and and uh whether you hang on to a certain career or a certain person or or or whether you are out hiring new people, you know, just any labor-based um anybody that does their own work, that's critical. And and it the more you have a handle on that, the more foresight you have into uh what what backlog looks like and your ability to maintain levels of backlog.

Fraser Gallop 19:04

So that's yeah, that's where some of these project management systems that are outside of the ERP may be where they're capturing that change order information, uh like for proposed change orders before they're even finalized, you know, get those estimates, adding you know, hundreds of hours to the job, that's gonna impact your planning and your management. But you'd see that alongside everything else that's happening in the ERP and the other systems.

Steve Gross 19:29

Yeah, and another example, like uh in uh heavy highway contracting, you may mine your own material. You may have your own asphalt plants and things like that. So as you win projects all the way from the CRM side, uh, the minute you win a project, one of the metrics that is very useful to be captured right there is uh quantities of material needed. And that can drive your production runs and your planning as far as your materials requirements if you're producing your own material, or if you need to buy it from the outside, you can be lining up because it's a little more complicated when you're doing work in a different region. You don't know, you haven't done business with local materials suppliers before. You need to go out and establish those relationships and so on. So it's just uh it all depends on where your exposure is and what your major costs are on the job and just having uh having more foresight into what the what those costs are going to look like and who you need to buy from and so on, and how many people you need, all those all that sort of stuff is just really crucial to keeping a handle on things.

Fraser Gallop 20:30

Aaron Ross Powell We've done joint case studies in the past where we've looked at monthly reviews and forecasting and enabling systems to get that down from 40 to just four hours. It could take 40 hours for PMs to go and gather all that information for all their jobs and review it and do the forecasts. When you've got systems that are pulling that together for them and just give them a very easy entry screen to okay it and send it for approvals, it's a lot faster. We talked at the beginning too getting that monthly reporting in place is just like a such a time consumption, uh stealing a bunch of time away from executive and these guys that should be looking at other stuff. You can go higher and you can think about like the board and executive reporting. That's maybe not a monthly thing, that might be quarterly meetings. It can take a good 20 to 40 hours to put those PowerPoints together that they're being shown. It's the same thing every time. You can automate that. And we've seen it time and again where where people are turning that down uh to a couple hours to to put those PowerPoint presentations together now, uh, where it took way, way more time in the past.

Steve Gross 21:35

Absolutely. Agree a hundred percent.

Fraser Gallop 21:40

That was the meat of what we want to kind of cover is you put all these different pieces together, and that's really what COVID looks like. You're automating things and you're putting things together in new ways. You're kind of changing from the way that you're entering data on the job site uh to the you know, the board level reports that roll all that up at the end of the quarter, you're really re reducing the amount of manual work that's happening, assembling that's happening. It's becoming all data driven.

Steve Gross 22:10

Um any other kind of final thoughts on that and and what good looks like I think things just run quicker and you're freeing up your expensive people to do what you hired them to do. More of more of that versus compilation work, which just time consuming. And if you have a way to eliminate, why not eliminate that? Why not do it? You know, and it's just very compelling. This is all great stuff, and it really the um the top performers are doing these things to some degree already. There's a high correlation between margin and success in a wedding projects and whether or not you're doing this sort of thing internally.

Fraser Gallop 22:50

So all right. So um that's a great place to leave it, I think. Um the next episode uh is gonna be kind of our finale for this uh first season. Uh, we're gonna talk about the future of construction data. Please uh subscribe and and uh tune in next time. It's gonna be uh probably three weeks, give or take from now, I think. Steve and I are are are planning to be in person together for this finale episode. So I'm I'm really looking forward uh to that. We'll see you in Phoenix next time we meet. Sounds good, Fraser. All right, thank you. Watch you then, bye.