Quasi-Lean at a Call Center

I stumbledupon an interesting article on Lean implemented at a Call Center.  Their implementation of Lean is pretty basic, but I think it’s a decent start.  In addition to what they did, I would have taken the Lean For Service and Lean Provision approach.

Specifically, this is what they implemented:

• Use visual workplace to expose production problems
• Began to identify “waste” in the context of a call center environment.  In their case, this was an outbound call center that collects data via surveys.  When quota is met, then additional surveys after that is considered “waste.”

So, as you can see, they haven’t really done much in terms of their Lean implementation.  But, apparently, they see big improvements despite their piece-meal approach to Lean Manufacturing.

My biggest beef with this implementation was an over-emphasis on productvity.  Productivity is a worthy measure, but it’s only one.  A truly Lean approach would begin with questions around “Value” — what the customer is willing to pay for — which may or may not include issues of productivity:

• Is the customer willing to pay for this activity?
• What is the customer expecting from this activity?

Answering the questions above will highlight waste from value and help the firm to understand customer expectations for consumption-based activities.  For example, basic consumption activities can sometimes place an undue burden on the customer.  Take, for example, the experience of getting a car fixed.  A simple activity like that can easily place burden and can take many person-hours than it really should.

Specifically, my approach would be to walk the Gemba from the the customer contracting with the Call Center for a survey, through the completion of a call for that customer and the subsequent aggregate reporting back to the customer that contracted for the data: the provisioning piece would be the contracting and the reporting back to the customer, the consumption piece would be the activity of data collection.   A visual time-map like that, showing time, activities, and also value-add and non-value-added activities would be very helpful and would immediately expose waste that can be reduced or eliminated.

The article was originally published in the Oregonian, August 2006:

Lean, said Martyn, can more easily spot waste in manufacturing firms, where piles of scrap metal beside a production line, for example, can indicate problems. In service industries, production is less tangible.

“You are looking at setting standards for your system, looking at when you aren’t meeting those standards, and going after the root causes of the problems that keep you from meeting those standards,” Martyn said.

“Your goal is to permanently solve the problem.”

In RDD’s case, the first step was making production problems visible. The company uses fairly low-paid workers — average pay is $9.50 an hour in salary and commissions — to gather data, mainly by telephone, for research studies. Martyn needed a way of measuring productivity so glitches would be obvious.

Each call center is organized with about 15 callers, or “research associates,” overseen by a supervisor, or “coach.” Martyn devised a system of metrics, or measurements.

Data is now compiled for each team: “RPH,” or revenue per hour, measures how much money each team makes per hour, based on time spent on each study. “Run” measures the research associate turnover rate, or how long people stay on the job. Another measurement is the cost of labor as a percentage of revenue. The lower, the better.

Early numbers were dismal.

Under Martyn’s guidance, visual systems were established to help research associates increase their productivity by reaching the right kind of respondents. Each research associate, for example, has within reach a dial that is turned to indicate a respondent’s political affiliation. The associate also has a two-sided sheet to indicate the sex of the person: blue for male, pink for female.

Boards attached to the wall show, for each survey being done, how close to quota the company is for people in various categories. If they’ve met the quota for a particular category, the caller can end unneeded interviews.

“It seems like it’s simplistic, but that’s what makes it work,” said Christian Byrd, a coach at the Missoula, Mont., call center who supervises 15 research associates.

Gerianne Schmidgall, a site manager at the Montana center, said the visual system makes the job easier for research associates, and increases their commissions.

“I was around before we started using some of these visuals, and we were constantly going over quota,” Schmidgall said. “We were doing work that’s not needed, and that’s a waste.”

John Fries, a senior project director for Alan Newman Research of Richmond, Va., is a customer of RDD. Fries’ company designs surveys for customers, including advertising and government agencies. He uses RDD to collect data over the telephone.

RDD’s use of Lean makes it much better at doing work quickly or changing a job in midstream than another data collection firm Fries uses, he said.

“We have a lot more flexibility using John’s shop than with our other call center,” Fries said. “We’re impressed by measurements they are doing, down to the efficiency of each interviewer.”

RDD still wants to grow, Stepleton said, but in a controlled way. “Part of maturing is being more patient,” Stepleton said. “That’s a huge fundamental of Lean.”