Metal fabrication through team collaboration

A metal fabrication shop manager might push a cross-training initiative, but that could be a temporary fix to a larger problem: the siloed nature of the entire manufacturing organization. Getty Images

When people think of cross-training, they often think of the cross-training matrix. When the matrix is full—that is, a group of employees know every skill needed in a particular area—you have a flexible workforce. Bill Ritchie isn’t a fan.

He is a fan of cross-training, however. Ritchie is a recently retired consultant who focused on quick-response manufacturing (QRM). Developed by Rajan Suri at the University of Wisconsin-Madison, QRM promotes the use of collaborative teams, or cells, both in the office and the shop floor, and cross-training within those cells is critical.

So what gives? If cross-training is critical, more is better, right?

As Ritchie explained, not necessarily. “Think about a 10 by 10 matrix where you have 10 people learn 10 jobs. There’s a lot of waste in doing that. One, it is rare that everyone needs to be trained in everything. And two, they typically forget skills not used after they haven’t done [them] in a while.”

Also, the matrix isn’t anchored to what really matters, what QRM calls the manufacturing critical-path time, or MCT. Slightly oversimplified, MCT is a measure of waste represented by how long it takes between accepting an order until the customer receives the ordered product or service. The shorter MCT becomes, the greater throughput an operation has.

Unfortunately, the way many modern manufacturers (and businesses in general) are organized actually can hinder MCT. In the cost-accounting, Henry Ford-esque world, everyone performs a narrow range of tasks at a certain hourly rate, then passes the work downstream. From this comes many of the persistent frustrations of modern manufacturing. Everyone focuses only on local, not global, efficiency, throwing work “over the wall” where the job is someone else’s problem. This leads to cluttered inboxes in the office and excessive work-in-process on the shop floor.

Inserting a cross-training matrix into this mess is a bit like a bad paint job. It’s bound to chip away soon, if not immediately, to show the original rough foundation. For instance, if a cross-training initiative requires a brake operator to train someone else on basic tasks, his or her first thought might be, “Great. I’m training my replacement.” That’s the rough foundation (“my knowledge is my sole source of job security”) showing through.

Ritchie prefers to change the foundation. Instead of performing tasks constrained by reporting silos, collaborative teams can solve problems quickly, accelerate orders, and help shorten MCT dramatically.

In this environment cross-training doesn’t occur arbitrarily, just to put a checkmark on a chart, but with obvious purpose. If something needs to be done, employees learn from their teammates and collaborate to finish the job. Cross-training becomes so natural that in this structure team members might not think they’re being “cross-trained.” They’re simply being trained, learning from their co-workers on how best to maintain or improve quality while accelerating a job along its critical path.

About Job Titles and Skills

When Ritchie worked with manufacturers to set up collaborative teams, “the hardest thing to accomplish was to get people to stop thinking about job titles.”

Team structures don’t abandon job titles entirely, and there remains a hierarchy in pay scales. A degreed engineer will make more than a data-entry specialist. But as Ritchie explained, this doesn’t mean the data-entry person can’t perform any task traditionally assigned to engineers.

Ritchie described one office team, or cell, that was put together to streamline order processing. Most jobs involved extremely intricate products that required several engineers on the team. “We made progress, but one week the engineers got really tied up with certain drawings and bills of material. So we ended up teaching CAD to the data-entry person, so she could go in and make changes to the prints.”

To make this happen, Ritchie said, the company first had to overcome its bias toward job titles and the corporate silo structure they create. Titles and their associated job descriptions focus on tasks. But customers don’t buy tasks from a fabricator; they buy fabricated parts and products. In this sense, everyone in the company has a specialty—be it engineering, press brake operation, or anything else—but their job is to accelerate work down its critical path and into customers’ hands.

Everyone also had to appreciate that not every task traditionally assigned to those with certain job titles requires the same level of skill. Engineering is a prime example. Sure, engineers might spend part of their day on highly technical tasks, but not every task they perform requires extensive training or four-year degrees. In fact, they perform a lot of clerical work, like making marked changes to a CAD drawing. This doesn’t mean they should look at such tasks as “below” them. It’s just something that needs to happen to move a job down its critical path, and if they have the time to perform the task, they should. But if they have other priorities, there’s no reason why others couldn’t be trained in CAD so they can make those marked changes.

About Customer Delays

Ritchie has heard the same complaint at hundreds of companies over his consulting career: We could process orders so much faster if customers would answer our questions in a timely manner.

Many of his clients have asked whether such delays should be added to MCT. After all, a company can develop quick-response cells throughout the organization to accelerate a job’s velocity, but what can a fabricator do when the engineering team has a question and the ball sits for weeks in the customer’s court?

As Ritchie explained, “When people ask me this, I ask, ‘Do customers still blame you if a job is late? If so, you’d better leave [those customer delays] in the MCT.’ But the better question is, ‘How do you ensure that when you take an order, you have all the information you need?’”

Some details are bound to be unforeseen early; supply chain or material availability constraints might crop up, for instance. But how can a fabricator set up a system that at least minimizes the need to contact customers for further detail after a job is awarded?

To start, a shop might analyze the nature of its customer base, including its principal contacts, as well as a fabricator’s existing sales talent. For instance, some customers (as well as some inside salespeople) might prefer a point of contact for sales, but then a direct connection with an order processing team that delves into the technical details. Others might prefer a single point of contact, usually the salesperson, who can answer customer questions as jobs progress from quoting and through order processing.

In either case, salespeople work in a collaborative team comprising everyone with the necessary expertise to move a quote and order forward, including engineers and estimators. Team members naturally learn other order-processing steps and gain expertise as they work to push a quote or order down the critical path. Through this, they learn to refine their order-processing steps and ask customers the right questions to minimize later surprises.

Sales also could work under an incentive structure that encourages closing orders accurately and completely. This, Ritchie said, ties the sales performance metric not just to top-line growth (closing orders), but also to the MCT (attaining all the necessary details upfront to streamline order flow to the shop floor). For example, 75% of a commission might come from landing the order, while 25% could come from attaining all the necessary information before closing the order.

“Ideally, whoever does the quote also owns the order when it comes in,” he said.

Of course, Ritchie has heard the counterargument: Why go through all the trouble to gather details if the job hasn’t been won yet? Here, an analysis of quoting practices might be in order.

“I worked with one manufacturer that had a bid-win rate of just 9%,” he said. “After looking at the process, they realized that they were quoting a bunch of budgetaries.” That is, buyers intended to stay with their regular suppliers, but were also required to get at least three qualifying quotes for every order. So, to fulfill that policy, they called on a fab shop that they knew would quote anything that came in the door.

If a shop qualifies its quotes and avoids obvious window-shoppers as well as quote requests far outside the operation’s sweet spot, then it has a better chance at increasing its bid-win rates. A higher win rate makes quoting less wasteful. All the effort spent to prepare a detailed quote, setting the stage for smooth order processing, has a better chance of paying off.

Resource Management

“A collaborative group gives you a unique way to know what’s happening in your business,” Ritchie said.

This includes an undistorted view of capacity and available resources. For instance, say a fabricator has a constraint in engineering. In the traditional silo structure, the shop might just hire another engineer. But as any custom fabricator knows, the flow of work isn’t consistent. When it rains it pours, and during such storms the shop might have enough work to support another full-time engineer. But during slow times, that same engineer might spend most of the day scrolling through social media.

Such inefficiencies can spread when a workplace’s organizational structure focuses on tasks (engineering, order entry, etc.) versus MCT. What if the previous scenario occurred in collaborative teams? In the first example, the team could have helped the engineer with some basic tasks during the crunch times, eliminating the need to hire another engineer.

Conversely, if the entire team were overwhelmed, upper management would then know for certain that, yes, the team needs additional resources so that it can continue to shorten MCT. This could be a degreed engineer or others with varying levels of expertise. But because they all work in a team, you can bet that new person won’t spend half his days on social media during the slow times.

Note that not every team or cell needs to have a direct connection to MCT, though an indirect relationship will still be there. To illustrate, Ritchie described a situation in which finding employees was a company’s primary constraint to shortening MCT. So it put together a cross-functional team dedicated to the hiring process. Team members visited trade schools and developed the right relationships to find the people the business needed to shorten its MCT, increase throughput, and grow.

How a Team Improves

As Ritchie explained, a well-functioning collaborative team has a self-monitoring and improving nature. Going back to the prior example, working in a siloed structure, no one notices the bored engineer in his cubicle, scrolling through his Twitter feed. In a team environment, communication is continual. Everyone knows what others are working on; indeed, they depend on each other to push work down the critical path. They’re quick to recognize a bad apple and are apt to talk about it. One bad apple is likely to be pruned before it spoils the bunch.

At the same time, top performers spread best practices and, in doing so, improve the performance of the entire team. “What I’ve found,” Ritchie said, “is that really capable people start showing other people how to do things better. So, just as one bad apple spoils a bunch, one good apple makes the other apples tastier.”

Spreading that good taste builds a company culture, making it so that people don’t come to work just because they need to, but because they choose to. In this age of seemingly never-ending hiring challenges, an organization with engaged employees, those who are there because they choose to be, has a better chance of accomplishing great things.