The Wizard Is Oz

The high-tech transformation of UPS by CEO Oz Nelson demonstrates that the top guy needn’t be a “techie” to make a big? buck investment work.

March 1 1994 by Chief Executive


Sometimes it’s not enough to be the low-cost producer. Founded in 1907 by four young men who delivered parcels to local stores in Seattle, United Parcel Service nearly stood alone in competition to the post office in the ground package delivery business. But while UPS was tinkering with its existing system, competitors in the early 1980s delivered a rude shock. The success of Federal Express and other carriers proved that customers would pay twice as much for overnight service and the ability to track packages en route. Heavy users also wanted volume discounts. At the time, UPS didn’t give any. In short order, smaller ground carriers such as Roadway Package System began cherry -picking UPS’ better customers.

At first, UPS was slow to react. It didn’t introduce overnight service until 1982, and automation didn’t seriously begin until 1986. But since then, the Atlanta-based company has undertaken a technological and cultural transformation transformation led by two longtime UPS veterans: Kent “Oz” Nelson, 56, chairman and CEO, and Frank Erbrick, 54, senior vice president and CIO. Up to this point, neither executive had computer or IT experience. Before 1985, the company had less than 100 people in IT confined mostly to mainframe computing-maintenance of data bases and processing information.

“It was a back-office cost function,” recalls Erbrick. “Oz had a strong selling job to transform it into a marketing weapon.” Nelson sought Erbrick to help with the task, and the two worked closely together.

Beginning in 1986, they launched a fiveyear, $2 billion technology upgrade program. Each year since then, the company has spent an additional $500 million; though the smoke has cleared stateside, UPS expects to maintain this level because of expansion abroad. From the program’s inception, Nelson and Erbrick pushed for a bottom-up effort. Hundreds of teams feed ideas to cross functional steering committees, which in turn set priorities. This led to the development of UPS’ local area network and PC-based, multitier architecture, in which mainframes are slaves, not masters of the network. Erbrick oversees 3,000 IT employees using 56,000 PCs linked through 1,200 LANs. Two data centers in Mahwah and Paramus, NJ, have three mainframes each that crunch numbers and store heavy-duty data bases.

One of the company more ambitious projects is a mobile fireless data service, called TotalTrack which uses cellular technology and an alliance of 70 cellular carriers to immediately transmit package delivery information from its 50,000 vehicles to its mainframes. The service provides tracking information for all air and ground packages the same day they are delivered. Previously, this information was not available until the following day.

Having more acronyms than the Pentagon, not all UPS technology projects prove to be big winners. But those that win, win big. One example is the DIAD (Delivery Information Acquisition Device), a hand-held computer that enables UPS drivers to collect delivery information and scan in customer signatures that were previously kept on paper. The elimination of illegible signatures nearly tripled the number of error-free, next-day air deliveries. (UPS delivers 11.5 million parcels a day.)

UPS still must move quickly to be competitive in international markets. It continues to lose money in Europe and Mexico, and domestic carriers are nipping at Big Brown’s heels. Nonetheless, UPS air volume has been growing in double digits for the last two years with a fourth-quarter increase of 27.6 percent over the same period in 1992. This compares with an industry rate of 10 percent to 12 percent. Privately held UPS likely will top $17.5 billion in annual revenues for 1993, with profits approaching $800 million.

As Nelson reveals in the following interview at UPS’ corporate headquarters, sometimes the CEO doesn’t always see eye-to-eye with his CIO. “Frank breaks dishes now and then, but he has an amazing grasp of our business for one who isn’t a marketing or customer-service guy. I’ve gritted my teeth a couple of times,” says the CEO, who is nicknamed after 1940s bandleader Ozzie Nelson. “But I’ve never turned him down on anything he’s asked for. You can do that if you’re lucky to have a CIO with a good business sense.”

ONLINE DELIVERY

In recent years, UPS has used technology to transform it from an operations-oriented company to a market-driven business. Can you describe the transformation process?

Our original mind-set was: Why do we have to give somebody information on a package we picked up at 5 p.m. and delivered the next morning? Federal Express can do that, but it would cost us maybe 50 cents a package, and to us, a tenth of a cent was a big deal.

In 1983, we charged half the price for air delivery that Federal Express did and waited for customers to beat down our door. That didn’t exactly happen, but within 90 days we did recoup the revenues we lost by cutting prices. However, the business just wasn’t growing as fast as it should have.

Finally, we realized we were going to have to do what Federal Express does-provide information on each package. So we began building an Information Systems group. At the same time, we recognized we had to familiarize our functional people with technology and let them call the shots on business matters. So I put someone in charge of the technology and R&D side and someone in charge of the computer side.

Today, we can do things other companies can’t, for example, produce electronic signatures. When we give someone a proof of delivery, it contains an electronic replica of the signature of the person who accepted the package. We can replicate that signature electronically and fax it anywhere. We also can provide customers and shippers with certain information, such as destination, insurance, and type of service, before the package is delivered. We don’t do that on all our ground packages, but we do on our air deliveries.

How much do you spend on IT?

We’ve spent millions on technology the last few years-$500 million a year in 1991 and 1992 on technology, computers, and R&D. 1992 was the first year we spent more on computers than on vehicles. Initially that scared me; but information is just as important as packages.

Early on, I didn’t track our spending as closely. I went to our chairman, Jack Rogers, and said: “We’re going to spend a lot of money in this area. And some of it’s going to be spent on infrastructure, though the payback is sometimes difficult to measure immediately. But our steering committee, both the IS group and the technology group, is going to look at the projects and determine the costs and the payback. If they are in line, the committee’s going approve them until we run out of money.”

It was hard on him, because he’s a real cost guy, but he replied, “Go do it.”

Nonetheless, all resources are finite. How did you prioritize these ideas?

The IS steering committee analyzed their feasibility. We determined the manpower needed and decided we could commit X number of people to seven or eight projects out of 20 or 30 this month.

The committee assigned priorities. Projects that strengthened infrastructure came first. Next were things to help our customers and improve our service. The third priority was projects that garnered pure payback and increased efficiency; for instance, we installed a tracing system and automated our telephone centers.

What was the biggest constraint to making projects work?

Getting people to commit to a project and follow it through to the end. Until people become aware of what’s possible, they line up a whole lot of projects, commit to them, and promise big returns. Then when you start doing them, they get diverted to other things. They don’t man the projects properly, and they don’t get their people involved. I kept telling them, “Once you assign a project, it’s your responsibility to see it through, not that of IS.”

How do you measure your payback?

It depends on the project. In some cases, the payback comes in the form of increased productivity: more pieces delivered per hour, or more phone calls per hour. In others, it’s cost savings: processing costs or logging fewer man-hours. In still others, it’s higher reliability. For example, we now have a computerized list of our customers and all their information, so we don’t have to re-enter the data each time we deliver a package. And with our DIAD, the error rate in information delivery dropped from 1 error in 100 package scans to 1 in 300 scans. Every one of those errors had to be corrected. We calculated the cost on these mistakes.

TRANSFORMATION AIMS

What was UPS’ single biggest weakness before you started down this road?

We used technology to evolve from a reliable, high-production, single-service company on a par with Parcel Post domestically to a multiservice company able to compete with Federal Express, Airborne, DHL, and TNT internationally on air deliveries, and Roadway Package System on ground deliveries.

I spent about three days a week in a hotel room for almost a year working out the strategies, repricing our products, setting up criteria for discounting, and establishing training programs for our salespeople. Their pitch was stalled at, “We offer a lower price than the post office.”

We were cost competitive and relatively productive. However, we hadn’t identified our true costs at any given time for any given package for any given customer. This required a lot of work, a lot of research, and a lot of assistance from technology.

Once we knew the cost of providing specific services, we changed our billing systems. We used to bill everybody at the same rate; now we have different rates.

Are you satisfied yet that you have all the information you need?

Generally, yes, but we still haven’t gone far enough in recording billing information electronically-such as scanning every package to get the name and address of the receiver, the kind of service (air, ground, second-day), delivery confirmation, and insurance. We have it for some shippers, not for all.

What’s your ultimate IT goal?

We will lower our cost of operations. We will be able to provide information internally to ourselves and make adjustments to personnel and equipment without losing a step on service. Customers will have a better idea of when packages will arrive or when they aren’t going to arrive and why.

POWER OF IT

What new projects are you looking at for the future?

Almost any Prism project excites the hell out of me. Prism refers to a way of approaching international communications projects that satisfy customer needs. For example, under the Prism concept, you can send a package from here to Europe, and if you want to pay the origination charge, but you want somebody else to pay Customs duties, and somebody else to pay the delivery charge, we’ll bill all three.

Does being private provide any particular advantage?

It helps a lot. That and being owned by our managers and supervisors, the ones who know the business and appreciate the money being put back into it. They have stuck with us. We could have had great difficulties sustaining the losses we had in international operations if we were a public company. It would have taken a lot of dancing and a lot of explaining, and somebody could have replaced me and dramatically cut our losses overnight by bailing out.

If I were going after short-term profits, I might have chosen another course. But international is going to be one of our winners, possibly as early as next year.

How much are you losing overseas?

About $250 million a year. But by about the end of this year, we should be at the break-even point on deliveries from the U.S. to other countries. Deliveries within foreign countries is a different story. That will take a little longer to turn around.

What advances has UPS made on the international side?

The problem with international deliveries is the paperwork. In the U.S., we just slap a label on a package and deliver it. No documents or invoices are involved. The shipper puts a Pitney-Bowes stamp on the package, gives us a sheet at the end of the day that shows how much is charged, and sends it to the billing department. It’s a beautiful system for us and the shippers.

But if you send a package from Indianapolis to Toronto, the law requires an ungodly amount of documentation. So we started that service and charged an arm and a leg for it.

Then we began building systems. We used to get long-distance calls on 70 percent of the packages we delivered, reporting them missing, asking us to fill out additional forms, etc. Our service was awful. It’s a sin to have packages sitting around because of incomplete forms.

Now we gather all the information beforehand. We get a package, send the data electronically ahead through Customs, let the officials peruse the list of a thousand shipments going to France that day, pick out the 10 they want to inspect, and tell us.

When the packages arrive, the ones Customs wants to inspect are laid out, and the others go right through. We’re trying to do that everywhere. If the customer omits something, we either get back to him immediately or the next morning. He gives us the additional information, and we send it electronically to the border. When the package arrives, someone adds the information and keeps it moving.

If a customer sends five packages to five countries in the world, and he says his receiver or consignee won’t pay the bill at the centralized office until he proves it’s been delivered, we can provide verification immediately. Before, somebody had to write a letter to verify delivery. You can’t make any money with all your people writing letters.

VISION QUEST

How would you characterize your relationship with your chief information officer?

We have a good relationship, although we don’t always agree. Once in a while, he forms an opinion too fast, and I try to change his mind diplomatically.

Were there any moments during the five- or six-year transformation period when you had a dark moment of doubt and asked yourself, “Do I really have this right?”

No. If you knew me better, you would not even ask. I don’t look back. I don’t believe in it. I’ve never doubted it was the right thing to do.

What role should the CEO play in information technology?

The bottom line is that as a CEO, you have to create a vision and coax your organization to embrace it. Bring your functional people to the table with your technical people and work shoulder-to-shoulder with them. You can’t bring in technology and expect your employees to manipulate it without guidance. It’s a sure loser if you do that. Only your employees learning technology, using technology, and reaping the payback of technology will make it work.