From Gut Feel to Data Driven | The Keys to Job Shop Excellence

Glenn Holmgren, Western Hard Chrome
Drew Daavettila, Steelhead Technologies

[Glenn]
I'll tell you a little bit more about myself and Western Hard Chrome. I am a mechanical engineer by training, but, uh, hopefully, you won't hold that against me. Never truly applied my engineering training in anything, sort of, in terms of classical engineering work. Starting out of university, I began in the oil and gas industry working for Schlumberger and, and working on the field, getting my hands dirty. Then moved to the manufacturing company where I did research and development, quality, and those sorts of things. And then the last little bit that I had with the, the Stream-Flo group of companies was implementing SAP. Um, and anybody that's worked in that knows that that can be quite a, quite a challenging endeavor, so we implemented that for a group of companies. And then after that, I felt like I needed a little bit of a change, um, in getting out and trying to run a business, uh, on my own, uh, so I ended up... That's how I ended up at Western Hard Chrome as the general manager. So

[Glenn]
Western Hard Chrome is primarily a repair facility, so it's different than some of the ones that you've talked, uh, that some of the folks have talked here today. Um, we're essentially doing industrial coating, so hard chroming, thermal sprays, electroless nickel, um, and then we grind and polish those coatings to repair products for the oil and gas, mining, uh, and, uh, utilities industries in Alberta and the provinces around it. 99% of our work is sort of small, custom, uh, job shop-type work, right? When I came in to Western Hard Chrome, I, you know, I'd met with the owner and, and really liked what I was hearing and the support that I was, uh, felt that I would get. When I got there, certainly did find that a family-owned business, uh, they've been operating in, uh, in Edmonton, uh, effectively unchanged since the 1980s. We had an experienced wait- uh, workforce. As many of you talked about here today, you got long-term employees. We had a great reputation, um, that had spanned decades, um, but were really struggling from, uh, trying to recover from the COVID hangover, uh, and to meet our customers' needs in sort of an increasingly compliance and document-driven world. I'm sure most of you can agree with that. We s- couldn't seem to get back to our customers in a timely fashion, so we talked about poor customer service response there. It was hard for us to be able to give a straight answer to anybody. And a lot of our customers, when we talked to them, felt like we were ignoring them, we were hiding from them in a lot of ways. And then looking at it, clunky data from the production floor. All of our production data... So you can see here, uh, that's a copy of our old paper work order that we had. Keeping track of data on the shop floor was extremely difficult. Um, guys were... Everybody had the piece of paper that got passed around through the shop and everybody wrote their numbers on it, and then when they made an error, they crossed it out, and then you couldn't tell who was writing it down or who wasn't. Um, that was also the sheet of paper that had the carbon copy attached to it so you could do your pricing on it. So if you look at some of these examples you'll see numbers over top of numbers. It was also where we tried to keep track of our non-conformances when they were happening. So, super challenging to try and understand what was happening at the shop floor level. And then trying to communicate that to customers, of course, was also a challenge, right? What that ultimately meant was that as an organization, our reco- our results were declining and our profits and revenue were dropping off.

[Drew]
Just to paint the picture, this is early 2023, and then we adopted Steelhead, what timeframe roughly was it? January of '24?

[Glenn]
Uh, it was... I think we were fully up and running in March. We started our work in January and we were fully up and running in March of 2024. Um, it took a while for us to kind of convince ourselves, even though I had come from an industry where we had ERPs in place, that we needed to have a some sort of a system there. Drew and I talked a lot about single source of truth. That was a big thing that we had when, uh, when I was working in the other company, and it was really... It spoke to me. We needed a place to pull all that information together so that we were all talking about things the same way. And so, like I say, from my perspective as the, you know, somebody new coming into the organization, and I'd spent this time, I was s- struggling in that first six or eight months that I was there to try and understand what was going on, right? As an engineer, I love data. I wanted that data, but you just couldn't get it in a timely fashion. So, I was struggling to understand the business, how to make things better. And while ex- I had the experience running SAP and I knew that that wasn't the right choice for us, I knew that we needed to have a- an integrated system that was better for our, our business, right? So, we wanted that single source of truth. We wanted to be able to pull that information together. We needed that real-time visibility. And, and, you know, as you could see, the paper copies would come into the office and then somebody would get around to entering the data either into accounting or into a spreadsheet so that you could try and look at what had happened in the previous month or two. We needed to up our, our customer service game. As we talked about, you know, if we're not satisfying the customers you start losing and then it becomes a snowball where people are leaving and you're not, you're not getting the outcomes you need for your customers, right? We wanted data. We wanted to be able to look at what was happening in our business and be able to, to make some decisions based on that to actually make improvements. So we really wanted help, and that's where Steelhead came in. So we did... Some of you talked about it before, the, you know, sort of looking around on Google, trying to find somebody that could help us out, um, and we were fortunate enough to come, uh, into contact with Steelhead and, and get going on what we were doing, right?

[Drew]
What we're gonna follow along with, uh, through the next few slides here is just, uh, that... What, what the Steelhead tool brings as far as data, how Western Hard Chrome has utilized it, and, and the picture that we really wanna paint for the audience, what's, what's often misconceived by, from our experience at Steelhead, is that the system will, will do this for you.... and that, that's not necessarily true. We, we can have the best tool, the best system in the world, but what it requires is, is a captain to steer the ship. And we think that, that Glen at Western Hard Chrome has navigated the course quite well. So that's what we're gonna illustrate here from Q2 of '24. So we implemented Steelhead in March, and we're gonna fast-forward, compare those results from Q2 '24 to Q2 '25.

[Glenn]
As a key part of any, uh, business fundamental to everybody really is profit growth and, and cost control, right? Um, it's important. These are important elements to making the business successful. So there were several key elements that I wanted to understand but weren't immediately apparent to me when I was looking at what we were doing in the paper world, right? And so we... as a small business, couple of key things we're looking at, we had lots of long-term customers, right? We've done business in Edmonton for... well, we're into our 64th year now. We've had lots of customers that we'd worked with for decades, um, which was great. Um, and, but we're essentially taking every bit of work that they had, right? Anything that they would ask us to do. And so my am- initial mindset as, as coming into this was, "I need the revenue. I, I wanna bring that in regardless of what's going on with it," right? Um, we couldn't s- easily see whether we are making money on these jobs or not. We had a good sort of, like we say, gut feel about what was going on, but you didn't have the data to support the decision that you were gonna make, right? We had little visibility of the costs that were being generated, right? So, uh, many of you kind of talked about that already, is how do you know where your money is going? What are you spending it on? Especially on ours, most, so, so much of ours is driven off of labor. If you couldn't tell who was doing what and how long it was taking, you didn't know exactly whether you're making, you know, making good choices about who was doing the work in the shop. And then on the data gathering side, if you got crappy input, you got crappy output, right? So the data gathering, the entry, the analysis, everything that we had was so prone to error, data entry errors, different things like that, the guys not recording time. Somebody had mentioned that you don't record it in, uh, you know, when it's actually happening. You don't have a timer running. You've got, "Oh, uh, I forgot to write that down, and that job was a week ago now. What did I actually do on that job?" So you had a lot of lag in things and a lot of error, right? So it was making it challenging to make any kind of, uh, actual decisions about what it was. So wanted to look at a way, uh, well, h- you know, how could we do a better job of deciding what we wanted to turn away and what we wanted to keep in terms of jobs? So those are some of the key things that went into that, right? So, you know, looking at the jobs that we take, looking at what the margins are, those are sort of the things that we, we wanted to achieve, right?

[Drew]
One thing that I wanna point out for the audience that I noticed while working with Western Hard Chrome is Glen and crew w- would actually review the margin on every single job that went through their shop. And if it was not ideal, they would investigate why, like, like nearly immediately, not a month later, not a week later. Maybe that day, the next day. Within a couple of days of completing that job, they would investigate the bad margin jobs. And, and one thing that I wanna ask you, Glen, for the audience here is how... You mentioned, uh, having these, these customers that you maybe have done work for for decades, and how difficult was it to increase the price or maybe turn away some of that revenue? As, as a small shop, you, you would hate to turn away revenue. But if it's truly the bad apple-

[Glenn]
Mm

[Drew]
... how difficult is that to do when, when it's been such a long-term customer?

[Glenn]
It's very difficult. It's, uh, it's a very difficult conversation to have internally because you, you've supported these people for a long time. And it's still a lesson that we're learning today. Our numbers, you can see, we have seen improvement. So we're talking about, say our typical margin at the time was say about 25% and we're up to the 35% now. But there's still definitely space there to grow, right? But it is... it's certainly a difficult discussion to have and a decision to make. But with the data that you're getting from the system, it makes it more, way more obvious to us that what are losers and what are winners, right? So and that, and that's really the key for this as well, right? And there's certainly, there's fluctuations that happen here too. So but overall, we've seen that. We've seen, uh, we've seen an increase in our margins there, and that is tied, you know, how we're working in some of the, you know, some of the slides that are coming up, but also just being more choosey in the work that we pick, right, and going after the ones that are actually in our wheelhouse, right?

[Drew]
So then naturally we're, we're going to dive in not, not only in margin here but in, in the following consecutive, uh, couple of, of slides to come how we, we achieved this increase in net margin.

[Glenn]
Mm-hmm.

[Drew]
What other statistics other than margin have we been improving that we should also feel an increase in revenue or net margin in the future?

[Glenn]
And like I say, looking at our costs in real-time, being able to make transitions in people on the shop floor, or who... like technicians that are working on jobs or processes that we're doing. So as we said, hard chrome shop, primarily hard chrome shop, but with thermal spray capability. Lot more, uh, control over our product output with the thermal sprays. So some of this is driven by saying to customers, "Hey, we know where your part's always challenging for us. We have lots of re-works on it. Can we switch to a different process that helps us and it helps them win too," right? So the next thing up, talking about quality and efficiency on the shop floor. So-... couple of the challenges that we were talking about there, what was actually driving our quality issues that we were having, right? We knew that we were having non-conformances, but a lot of that stuff was sort of... It wasn't documented. It was dealt with more on the fly. You had no long-term data to actually understand what was causing the problem and, and how to make changes. And then, you know, looking at our labor, right? As you saw when we were pa- we're doing it in paper, how do you actually judge who's doing the better jobs, right? Who's, who's actually performing higher or lower than other folks, and how do you deal with that, right? So when I first started there and when we were having the discussions with Steelhead, we were, we'd been struggling with trying to meet the evolving needs of our customers. So we touched on that a little bit before, but we were really stuck in a rut. We had great, long-tenured employees that had done things a certain way, but there weren't do- there wasn't the documentation that went along with that. And a lot of it, what we're finding even today, a lot of that success that we were having as a business heavily relied on the customer relationships, um, that we'd built up over decades. And we got a lot of customers that would say, "Oh, it's good enough, we'll take it." We never got that feedback as an organization. So again, you're not filling that plan, do, check, act kind of cycle that you'd wanna see in an improvement cycle for a business, right? So now today, what we have is as people retire, you get the new younger compliance-driven engineer or project manager that you're working with, uh, for the business that isn't willing to make a concession, right? And that's really forced our hand to, to look internally and see how we can make improvements, right? So, you know, so a couple of things that we've been doing here. We knew we needed to up our game in capturing and rectifying issues, certainly before they became c- a customer's problem. We also needed a way to assess our people's performance, right? So with Steelhead, we now have the ability to... Everybody on the shop floor can stop a job, which has been huge. We talk about engagement with the shop floor, that empowerment for people to be able to say, "Let's stop. Let's fix what we're... You know, we're seeing a problem, let's make it better," um, has been really big, right? And it allows us that time we need to engage our quality team and, and help to make things better. Um,

[Glenn]
now we write up issues that we see within minutes of them happening. We can create non-conformances. We can define a d- a disposition for a product and, and get a plan rolling, uh, within minutes of it being i- identified. And the big thing for me from a quality, sort of the quality background, is being able to look at that trend of data, right? What, what's been causing our problem and how do we look at where we can put some continuous, uh, improvement horsepower into things and make jobs better, make our product better, make our output better for our customers, right? The other thing we are able to do now too, which we weren't able to do before, is look at the data, um, for real-time and historical analysis of labor, right? So now, uh, on a weekly basis, I can look at every member of our team and see how they're doing in terms of productive versus non-productive time, right? And that was never ha- that had never happened before. So the information for the labor hours comes from the timers, the timer dashboard. So everybody, um, that we have in our system, every job that's going through, they're punching in and punching out. As they accept jobs and pass them onto the next stage, there's a timer running. So there's a rudimentary sort of analysis that's done in Steelhead with regard to sort of, uh, everybody's basically working on an eight-hour day. So we get that kind of gross percentage of labor efficiency. Um, and then I take it a step further and I pull information from our, uh, payroll system and take it that way to get a, a more accurate prese- presentation. But it's actually very close for the most part, because not a lot of people are going beyond that eight-hour day anyway for our shop.

[Drew]
Glenn, Glenn gave me one, one very good, uh, tidbit. For those in the audience that, that don't track the three key things, you can't fake all three, and that's clocked in time, billable hours, and throughput. You can fool two of the three, but you can't fool all three.

[Glenn]
Mm-hmm.

[Drew]
That's a line from Glenn next to me.

[Glenn]
Yeah.

[Drew]
And that, that holds true a- and it's a very, uh, interesting thing to look at when we're looking at production-

[Glenn]
Yeah

[Drew]
... is that you can fool your throughput, but you can't fool all three of those things.

[Glenn]
Uh, so now when we have that data, our leadership team can have those conversations one-on-one with our shop floor folks and say, "Hey, your, your numbers are lagging behind the, the other folks in your department. What's going on?" Is it something where, um, more training's required or they're not the right fit for a particular machine that we think we've put them on, right? There's that opportunity for us to engage with the employee and try and make them, you know, fit a little bit better, do training if necessary. And then if, if need be, that's also an avenue and some backup to go into progressive discipline if it turns out that somebody's knowingly doing something incorrect, right? Uh, fortunately, we don't come across those problems too often, but it is something that's there. Um, and now you have that data in your hands to be able to make those changes and, and have that to back you up, right? So again, touching on the outcomes here. So our time on hold, uh, as you can see, is a huge decrease. Um, our... The improvement time has been huge, right? So in the paper system, we weren't really tracking that at all. But even now from quarter 2-24 to quarter 2-25, the time it takes us now with everybody understanding the system better and knowing how to function has really decreased for us. And that's great for getting stuff back on the shop floor if we do make a mistake and need to get it turned around.... um, our rework work orders, so every job that we have non... that we find non-conforming that has to go back through gets a rework work order created for it. That's been a 23% decrease. So we're also seeing less errors, so I mean, that's tied to just, again, people understanding the processes that are in the system a little bit better now that it's more defined, and being able to execute on things a bit better. And overall, our rework cost has dropped by about 8%. It's not exactly where we want it be just yet when we're talking about absolute values, but we are moving in the right direction, which is huge for us and huge for our customers. Customer service is a big, uh, big key to things as well, so it's sort of the final pillar in the things we'll talk about here today. Our customers really were telling us we were horrible. [laughs] We couldn't do a great job of keeping them up-to-date on a job and getting the customers what they needed to know when they needed it. Um, we were, we were really struggling with that, right? And a lot of that, I think, was tied to just our team not being confident in what was going on in the job at any one time, right? They couldn't tell what was, what was pending, what status it was. Um, so when our team was unsure about the answer, their default position was just not saying anything, right? And that's just not a great, uh, formula for success in any business, of course. Um, and what it ultimately led to is dragging out conversations, so we talk about lengthy communication trails or escalation to me or to our president, right? Uh, the phone calls from the long-term customer that says, "What the heck are you guys doing?" Right? So that kind of escalation is not what anybody wanted to see. So what we needed to try and get out there was, like we talked about in a lot of these things, having that live status on the jobs was huge, right? So now, everybody on, uh, our leadership team is able to see the status of any job that we've got going on, whether that's through work boards or whether that's through Watchtower. Those are huge functionalities for us. Our guys in the production side deal with that every day. That helps them plan out their day in terms of what's gonna be happening on the shop floor, who's gonna be working on what. Um, that, that visibility has been, uh, a huge boon for our production management folks. Um, and now also when a customer calls, it's very easy for us to be able to say what the status of a job is or if it's on hold, right? We can give them that information. For me, for myself, even as somebody that's not necessarily out on the shop floor every day, if a customer does happen to phone me or if I'm out visiting them, I can pull that information up and I can tell them what's going on with their job immediately, right? Or show them that as an example of how we can, uh, we can better suit- uh, support their business, right?

[Drew]
It's a incredible, uh, turnaround in, in the on-time delivery and dock time. What... Like, what do you think it... from, from your perspective, uh, or from your customers' perspective rather, like, what is the, the decreased response time? Like, so a customer used to call you, did it use to be an hour until they got a, a status on a job and now it's 10 minutes? Like, what do, what do those actually look like at Western Hard Chrome today versus last year?

[Glenn]
Um, I would say you're probably right. It would be something like when they originally would call if they... if we actually got back to them, which is bad to say. Um, but if we did that, we'd be talking about, say, three hours, right? You'd be... The guys would have to go out to the shop floor. They would talk to somebody on the shop floor or they would have to talk to a couple different folks to know what was going on, and then they would get sidetracked with something else because somebody else would talk to them. And so that spiraled to the point where you weren't getting any response to the customer or very del- very slow, right? So, uh, like I say, now, the guys can pull it up on Steelhead. They can be... They're on the same device as they're looking it up. They can, they can give a response to the person as soon as they want, right? So-

[Drew]
Without hanging up the phone.

[Glenn]
Without hanging up the phone. Right? The other thing that's been big for us is the use of the customer portal, right? So now, all of our key customers, I don't know how many of you are using it in your Steelhead worlds, but, um, that's been big for us. Every ka- every key customer, we haven't opened it up to everybody yet because some of our customers are quite small. It really isn't rele- relevant to them. Um, but now, our customers, the ones that are tech-savvy or just shy about calling us can get all their information all in one shot without ever having to actually talk to us, right? So, um, and the other sort of benefit to that is it keeps our own team honest because now they know that the customers are seeing that data. So anytime we thought we might want to shine somebody on about something, guess what? They can see that data too, right? So it helps to kind of, you know, keep everybody on the same page as it were. And as you can see, uh, definitely decreased the response time. Our on-time delivery numbers have drastically improved. We, uh, had a target in our business of 85% on time as sort of a normal benchmark for people. Uh, we were historically always running about 45%. We've been h- now over the last, uh, uh, say six months been running, uh, well over 85%, if not higher than that, right? And dock to dock time, so our internal processing time has dropped dramatically as well.

[Drew]
Just to recap, uh, Glenn, with, with the transformation from Q2 of '24 to Q2 of '25, obviously, we've, we've done some significant measures in customer service quality and, and, uh, the margin departments. What is, what is next for Western Hard Chrome and, and how do you want to continue down the path of continuous improvement?

[Glenn]
The things that we wanna see and, uh, that we wanna continue to work on with you guys, uh, big one, a couple... well, three big ones for us. Scheduling and capacity planning, um, trying to better understand, so that means more data input from us, but a better understanding of what we actually can provide to a customer and when, right? So we're still at a certain stage where we know what's going on in the shop floor, but we don't have a good, uh, idea of what our capacity truly is.Um, so once we have that piece in place, it's a lot easier to say to a customer, "Yes, we can give you a four-day turnaround because we know that those, those production areas aren't at full capacity yet," right? So that'll be a big one that we're already working with Steelhead on, and Drew and, and myself have been going through. Uh, quote, turnaround time. A couple of folks already mentioned that. Um, we've seen some, uh, new work that the, the Steelhead folks have done on RFQs. Um, and I had... Even just talking to a friend of mine that works in a different business, he's talking about how his sales team can't turn quotes around fast enough for their customers and they lose. Um, and that's... It's just not what anybody wants to see. So the faster we can get a quote to a customer... For us we're, we're in a good position because we do have a lot of those long-term guys that just come to us. But we're trying to get new business and come in. So if we can respond to them quicker and give them a more, uh, defined answer about what's going on, it kind of ties to capacity planning. That's really what we wanna see. And then the last one is kind of tying together with our, uh, uh, work on labor efficiency, is to try and look at the volume of chrome or thermal spray that's removed per hour per grinder. So this is another one of these ways to further enhance our ability to assess productivity in our business that isn't driven by, uh, widgets out the door, right? So that's

[Drew]
Tha- that was one interesting point that I do want to touch on that, that Glenn has, has, uh, mastered, uh, as an engineering mind, is how do you normalize across your plant floor? Some parts are big, some parts are small. If you measure department A to department B, what if this department only produces big parts? They get the, the... They have less part throughput, but maybe their, their revenue is higher. Well, we've gotta find the statistic or the KPI that normalizes all production. And for Western Hard Chrome, that's, uh, what we found to be, is cubic inches of material added or material removed so that we can compare relatively equally operator A to operator B, or department A to department B.

[Glenn]
So that's it. I hope that, uh, we've given you some value here today and, uh, yeah, if you've got any questions...

[Audience]
So we're a, we're a batch operator. Uh, we don't have a line. And part of the challenges I've had trying to do the costing is that, one, I don't know where to put the fixed overhead on the process. Uh, and two, we have a blast room, right? So you have an operator. We load him with one hour's worth of work, which could be three jobs. So the timers are... Sometimes they work, sometimes it's on a sheet of paper. We allocate a blast buddy to go in and help him and mind him, make sure he's okay, and help him move. Don't know where to allocate that. You know, when we're on the powder coat line, we could have two, three guys working a job for 15 minutes, and then they go off and do something else. So the timers are kind of all over the place, um, on an individual level. How did you... Or do you have... anyone has any insight as how I can kinda look at that so I can get better margin numbers?

[Glenn]
Um, so for us, we're fortunate that we don't have lots of cases where we're running multiple timers at a time. So it's more linear that way for us, so it's a little easier. So I don't have a good answer for sort of how to do that. What we've done based on the amount of produc-... Like now that we have a better handle on our productive labor, we've established a more accurate labor rate that we have within the system. So that's how we've dealt with it. So we look at... And that's how we baked in our fixed cost as well into that labor rate that we're running in the shop. So... And then it's just a matter of looking at the jobs that we're got- we've got going through and where we're exceeding. Like, as we look at that margin, where's our, our labor exceeding what our expectations were per job, right? So that analysis is happening as we're looking through every job. Fortunately, we've... You know, we'll have, um, 100 work orders on the shop floor at any given time. So it's not, it's not an overly onerous way of, of doing it. But we can see that in a snapshot of going through the, the jobs that are working on the shop floor to see where we maybe have to make some adjustments.

[Drew]
Uh, just one tidbit is y- y-... We can either do, um, within Steelhead what we call station costing, where operators that are working in a particular booth could clock into a station. That way, their time is, is split or allocated across the jobs that are in one, uh, station, even if there's multiple. Or if we don't utilize racking and we're a batch operation, maybe look into utilizing racking.

[Speaker 3]
Okay. So talking about the timer, I'm curious that, uh, when you release a ch- new job to the floor, you have a traveler or job card with, as a specified, uh, operations. So when you measure the, say, the efficiency, do you have a planned, uh, say, time for each operation so you can see the operator, how long they spend booked there? Or is it, uh... It's kind of, as a experience from the management? So how you are, say, set up the standard, uh, put it in Steelhead, then how to measure the actual performance?

[Glenn]
On our... When we cost the job in the first place, we actually break that down by operation. So we've got the hours and then the labor, you know, so then that's why the calculation of the cost gets done. So we give an allocation of time. If there's some-... So that... And that works well for jobs where it's relatively the same kind of product that we have. If it's something that's different that we've have maybe not done before, then that's where we'll engage the shop floor and we'll actually talk to them about what they think it's gonna take. So that one is a little bit of a shot in the dark in the first place, right? Um, but you get a good sense of whether you're accurate or not fairly quickly when you do that, right?

[Drew]
The, the feedback loop is critical, certainly, of the, the price versus cost or estimated versus actual.

[Audience]
How did, uh, Steelhead help your on-time delivery? It's a pretty significant increase.

[Glenn]
I think the big thing for us with, uh, from that was the visibility of what was going on. Um, when we were doing it before and not being able to... Part of it was we knew what was going on in the shop floor. So if we had something that was lagging behind, we could move stuff ahead. And then also that visibility helps us, even though we're not fully where we want to be with scheduling and capacity planning, um, we can also say now with more confidence that, "Hey, customer, I know you want a 10-day turnaround, but we need to give you a 12." So that helps, obviously, right? Because you can be more accurate in hitting your dates, right? So that's, that's where we've seen some of those improvements.

[Audience]
Round of applause for, for Glen and Drew [clapping].