I was named vice president of AT&T network operations in 1997. At the time, we had seven regional network oaperations centers—one in each of the regions representing the seven Baby Bell companies that had taken over AT&T’s local telephone services 13 years earlier.
Running seven operations centers across the U.S. created problems for AT&T. Let’s say a Bell company’s network went down somewhere between New York and San Francisco. Staff in one regional center didn’t want to talk to staff in another center. Employees were more concerned about protecting their regions than taking care of customers.
If a business wanted to restore long-distance or other service, it might have to go from an operations center in New Jersey, to another in Chicago and then another in St. Louis before someone took care of the order.
Those situations were unacceptable. That’s when I came up with the “concept of one,” which I define as: “Do it right. Do it once. Use it everywhere.” Take the best processes. Emulate them across the organization. We replaced the seven operations centers and created a single one in Bedminster, N.J., within 14 months.
Zero can be as great as one. When I was promoted to chief technology officer of AT&T and president of AT&T Labs in 2001, Dave Dorman, then president and now CEO, asked me to select the best processes and technologies for the company, focusing on customer care. My goal: create a structure that demonstrated AT&T was really trying to make it easier for customers to conduct business such as ordering services, reporting complaints and checking their bills.
I physically tracked a ticket, or work order, from a customer who reported that his service was down. I followed that ticket’s path by flying to six different network maintenance and customer care centers over three weeks. Although some complaints—back in 1998—were resolved in 12 to 24 hours, that customer’s service was not restored for 18 days.
What took so long? There were defects, or errors, in the information on the ticket, and those errors needed to be addressed at each center. Let’s say a customer’s address is 10 East 54th St., New York, N.Y. The customer gave that address to an AT&T employee, but the staffer who typed the information into our database left off the “East.” We then don’t know if the address is East 54th St. or West 54th St., requiring an employee to contact the customer to get the correct address.
Defects slow you down from doing your normal work. And every time you have a human involved in trying to get a defect fixed, it takes minutes. That’s a long time.
That’s when I came up with the “concept of zero” to reduce the number of defects to zero and to similarly reduce cycle time—the amount of time from when a customer requests service to when AT&T actually delivers the service. My model for this concept is to eliminate human-to-human and human-to-computer interactions by automating processes wherever possible.
Where did we begin? I asked my team to count the ways that customers came to AT&T. Our inventory showed there were 500 different ways—via phone, Web, fax, e-mail, voice mail or other means—that customers could use to order services, report complaints and manage their accounts. That would be comparable to going to your corporate Web site and having to enter your personal ID and password each time you wanted to use a different section of the site. That’s unacceptable.
Using the “concept of one,” our corporate customers are now directed to one portal, called BusinessDirect, when they want access to our ordering, provisioning and maintenance systems. The Web site handles about 30 million transactions a year by 600,000 users at 400,000 companies.
Our savings? At least $10 a call and sometimes as much as $25. AT&T BusinessDirect has helped save or avoid several hundred million dollars in costs over the past three years.
Which is why “one” and “zero” really are big numbers.
