PHONE: +1 936.588.1130


Your comprehensive source of information using behavioral technology to improve safety and performance. Sign up to have new articles delivered straight to your inbox!


Successful Management at Tamko Roofing Products

If at first you don’t succeed...

Each year for the past decade a new safety program has been launched at the Frederick, VA, Tamko Roofing Products plant. And every year the program would last about a month before beginning to fade. But last September employees started training for an intense safety observation process, and it’s still working and improving behaviors, says Gene Burrows, corporate safety director.

What’s the difference? Based on the theory of continual improvement, this process depends on employee interaction. Managers don’t pass down instructions. Communication cuts across the work force, emanating from peer observations.

Continual improvement isn’t a new idea at Tamko. For the past dozen years, the manufacturer has applied it to quality and production at all seven of its plants. But no one realized the connection to safety until Burrows received some mailings and attended a seminar on the principles of behavioral science.

The Frederick plant was picked for a pilot project.

Stuck on a plateau

But first Burrows had to convince corporate management that continuous improvement would work for safety. He gave a presentation, including hand-outs and an overhead that showed accidents numbers, to get buy-in. Safety numbers were stable – 10 OSHA recordables a year for 1994 and 1995 – but not improving with the management-oriented safety programs Tamko had used for years. Burrows emphasized that the way to significantly reduce accidents – and associated costs and liabilities – was to start at the grassroots. Setting an accident number go was not the answer either, he explained. Nobody is smart enough to know what ideal accident numbers are. Anyone could pick a number too high or low, he told managers.

It took five months to plan the process. Consultant Terry McSween of Quality Safety Edge, based in Houston, TX, began employee training, 15 at a time in half-day sessions in September. A safety-knowledgeable group of 12 supervisors, leadsmen and workers was formed to lead continual improvement. Known as the design team, they have the responsibility to make the process successful and help train all other employees.

The team began with a mission statement that also defines plant safety values: “Our purpose is to create a safe workplace for all employees in a way that promotes continuous improvement in open and honest communication, teamwork and leadership in safety so that we reduce or eliminate accidents and injuries.”

A list of 22 operationally defined behaviors most likely to cause accidents was drafted. Items range from intricate safety actions to simple use of PPE. Inattentiveness caused 36 percent of Frederick’s accidents in 1995 so “eyes on path” is on the list.

  • To give the program substance and ensure management and employee support, the team drafted six rules that management agreed with and signed:
  • To avoid employees’ fears of punishment, the names of those being observed are not recorded.
  • Results of worker observations are not used to discipline. This helps keep communication honest-a necessity for continuous improvement.
  • Supervisors and managers are not evaluated based on the results. Leaders need to feel confident that they won’t be fired due to observations.
  • All employees will be trained in the observation process. For it to work, employees need to know how and why safety must be done correctly.
  • Supervisors and managers will be encouraged to support and maintain the process. They are vital to keeping it alive.
  • Employees will be allowed to immediately review the observation data and receive feedback. This gives workers the chance to quickly address safety problems, says McSween.
  • Employees learn how to work safely by acting as both observer and observee, and by getting feedback from other employees, not higher-ups, says Henry Thomas, the production superintendent who McSween and several design team members think of as a champion of the process.

Thomas works hard to keep things rolling. Skepticism grew as behavioral training progressed. Employees had seen too many safety programs crash and burn to get excited about this particular approach, he says.

Confronting concerns

Motivation waned when scheduling got in the way of planning and communicating. Team members are scattered throughout 3 eight-hour shifts at the Frederick plant. Thomas pulled them together for a “clear the air” meeting where everyone discussed their concerns about safety, scheduling and accomplishing goals. Employees really opened up and talked about their worries, says Thomas. Some said they couldn’t see safety improving. Others felt the work demanded too much of their energy.

The team decided to divide responsibilities into subcommittees, such as training, problem solving, and data gathering.

This strategy got the team rolling again, but motivation still could be better, says Tim Jellison, leadsman and design team member. It’s a struggle to keep everyone involved. About four of the 12 design team members continue to feel slightly discouraged by scheduling problems and have been slacking off, he says. Jellison relates this to the history of failed safety programs. Still, he thinks the mood will change when it’s apparent that the program is working.

And there are signs that it is. Tamko finally rolled out the process on February 1 of this year. Observation data collected so far reveals that the Frederick plant is working safe 93.8 percent of the time when employees are being observed (5,000 safe actions and 327 concerns have been recorded.) Burrows notes that data is usually on the more positive side in the early going as observers are apprehensive about criticizing their peers.

Employee observations have definitely improved the overall safety culture at the Frederick plant. In fact workers mention safety and point out risks to each other even while not observing, says Jellison.

Tamko was going to wait four or five years to review observation data before taking the process to its other facilities. But the starts-up is going so well at the Frederick plant that implementation begins this month at sites in Dallas, TX, and Phillipsburg, KA.