Otis Elevator: Accelerating Business Transformation with IT
At Harvard the only thing considered duller, safer, and less adventurous than working for a bank was working for some old-line can't-miss industrial firm like Otis Elevator, which only needed caretakers. —Tom Wolfe, A Man in Full (New York: Farrar, Straus and Giroux, 1998), p. 31 At the northeast corner of 20th Street is the turreted red-brick Victorian Apartment house . . . the first cooperative apartment house in New York City. Peek into the luxurious lobby! The foyer is adorned with stained glass and Minton tiles, and the building is equipped with Otis hydraulic elevators installed in 1883, and among the oldest of their kind still in service. —Gerard R. Wolfe, New York: 15 Walking Tours (New York: McGraw-Hill, 2003), p. 261 In early 2004, Otis Elevator President An Bousbib (pronounced boozbe) finalized his presentation to United Technologies (UTC)—the parent company. The past 12 months had been a success for Otis, as highlighted in the 2003 Annual Report: Otis completed the second-largest acquisition in its history with the addition of Amtech Elevator Services. Amtech brings . . . a customer base spanning universities, hotels, hospitals, airports and convention centers throughout the United States. Overall, Otis revenues grew by more than $1 billion to $7.9 billion. Operating profit increased by more than $300 million to $1.38 billion, and operating profit margin grew by more than one point to 17.4%) Otis was the largest manufacturer, installer, and servicer of elevators, escalators, and moving walkways in the world. The operating landscape was shifting. "Today," Bousbib said, "we are no longer making things; we are moving things. Our differentiating focus is orders-of-magnitude improvements in logistics and service." No longer was Otis an old-line industrial company. Just a year earlier, when Bousbib held his annual leadership conference, he set forth a vision for Otis: "To become the recognized leader in service excellence among all companies—not just elevator companies—worldwide." In its quest for service excellence, Otis's future standards for comparison would be companies such as UPS that had institutionalized customer service and had developed

standard work, process flows, and metrics to govern every customer interaction and every internal activity. Achieving this goal would require transforming the culture at Otis. In Bousbib's description, the company did not simply provide elevators but total solutions to customers. This meant that every function of the business, from design to sourcing to manufacturing to new equipment installation to maintenance, needed to be run with a customer-focused mind, and all these functions further needed to be integrated into a single customer-centric business model. Otis Elevator was proving to be nothing like the company of Tom Wolfe's description.
Company Background Otis Elevator is named for the company's founder, Elisha Graves Otis, who invented the "safety-brake elevator" in 1853. (Exhibit 1 describes events leading up to the installation of the first passenger elevator.) Otis's core business was the design, manufacture, installation, and service of elevators and related products, including escalators and moving walkways. Otis's name had become synonymous with one of the most useful and dramatic inventions of the 1800s, the passenger elevator.
UTC's chairman and CEO, George David (who joined Otis in 1975 and became president of Otis in 1986), noted the recent accomplishments of Otis: "Otis Elevator.., had really good, powerful results in this last three years, even in the face of pretty tough economic conditiorts." 2 From 1997 to 1999, Bousbib was UTC vice president of strategic planning, and from 1999 to 2000 he was UTC vice president, corporate strategy and development. In 2000, at the age of 39, Bousbib became Otis's executive vice president and chief operating officer. Bousbib was elected president of Otis in 2002. Prior to coming to UTC, Bousbib had been a partner at an international management and technology consulting firm. Bousbib earned a master's degree in mechanical engineering from the Ecole Superieure des Travaux Publics in Paris and an MBA in finance from Columbia University. By 2004, Otis had 1.5 million elevators and 100,000 escalators operating throughout the world. Otis had elevators in 10 of the world's 20 tallest buildings and more than 1.4 million elevators and escalators under maintenance. Otis sold products in more than 200 countries and territories. Engineering headquarters was in Farmington, Connecticut with international facilities in Japan, France, Germany, Spain, Korea, and China. Of 60,000 employees, nearly 90% worked outside the United States. Otis operations were organized into seven regional businesses: North and South America, South Europe and the Middle East, North and East Europe, the United Kingdom and Central Europe, North Asia, South Asia, and Japan. Revenues climbed from $6 billion in 2000 to $8 billion by the end of 2003 with 80% coming from outside the United States. The company's return on sales, which historically fluctuated around 10%, reached 13% in 2000 and then increased by more than 500 basis points to over 18% by first-quarter 2004. As margins expanded, productivity soared and profits grew. The company had always provided solid and steady profit contribution to UTC since its acquisition more than 25 years earlier. In the period from 2000 through 2003, Otis's contribution to UTC's profit increased from 25% to 35%. Otis's sales and profit growth helped drive the strong performance of UTC, which had projected revenue of $35 billion for 2004, 14% over 2003's. (See Figure A for the stock performance of UTC.)

Elevator Industry: Competition and Marketplace Otis's main "hardware" competitors included foreign multinationals Hitachi, Kinetek, KONE, Mitsubishi Electric, Schindler Holding, Tatung, ThyssenKrupp, and Toshiba. Industry competitors typically attained operating profit margins of 5%40%. For a small-building project, the elevator manufacturer was selected by the contractor, architect, or building owner. Larger projects often involved all three parties in the decision-making process. Buyers selected an elevator company on the basis of its ability to satisfy the elevator performance specifications and architectural requirements, price, reputation, and past performance on other projects. An elevator service company was selected by the building owner or property manager on the basis of responsiveness, quality, and price. As a building aged and competition for tenants increased, the cost of service often became the major consideration, and the lowest bidder typically received the service contract. Many elevator manufacturers offered discounts for long-term service contracts. Some had begun to bundle up to a year of service for free with a new elevator sale in an effort to attract and maintain customers. New elevator sales were directly correlated with building cycles, but the elevator service market had been very stable. As the installed base grew, service historically accounted for a much higher portion of profits than new-unit sales. The service market attracted many participants because of its steady demand, low barriers to entry, and high profitability. Thousands of elevator service companies existed, including both elevator manufacturers and many small companies devoted exclusively to elevator service. These companies serviced elevators from almost any manufacturer—particularly elevators made prior to the introduction of microprocessor-based control that used similar generic technology. The U.S. and Western European elevator markets were mature, and growth in new-elevator sales was coming primarily from emerging economies. China was a particularly hot marketplace and key area of growth: "62,000 units of elevators and/or escalators were installed in China in 2002."3

From 2000 to 2004, Otis was aggressively growing with the market by acquiring elevator businesses in China. Otis's rival, Kone, was quoted as saying: "China has become a growth engine for the world's elevator industry."4 Others competing with Otis and Kone for China included Schindler, Mitsubishi, and Hitachi. The market was built almost entirely on revenue and profits from new sales, compared to more mature markets where service accounted for as much as 75% of revenues and profits.
OTISLINE® Customer Service Center—Early Applications of IT In the early 1980s, at the urging of its then executive vice president and chief operating officer, David, Otis created a centralized customer service system to dispatch service mechanics. IT worked with many functional areas to implement this 24 x 7 concept, called OTISLINE customer service center. Previously, service personnel were dispatched from local offices, and there was no central view of service delivery, response time, or cumulative product issues in the field. With the OTISLINE customer service center running on a newly installed mainframe, a centralized Otis service dispatch group of 160 people operating 24 hours a day could respond to a customer in less than a second. Information from multiple Otis data sources was aggregated. Rapid response was an important design element; customer service experience had shown it was unacceptable for more than 2% of transactions to take longer than a five-second response time. OTISLINE customer service center improved visibility of the elevator service business performance to management, enabling it to provide more effective quality service to customers. Prior to this, service problems might be unknown to senior management until they had become critical situations resulting in customer complaints or in some cases service contract cancellations. OTISLINE customer service center allowed Otis to produce "excess" callback reports for various levels of management. For example, elevators receiving three or more callbacks in a month were reported to the district manager; those receiving eight or more in 90 days were reported to the regional vice president. Critical situations were reported to the president of the regional business. The expectation was that reducing callbacks and improving product reliability would improve customer satisfaction, thereby reducing maintenance contract cancellations. OTISLINE customer service center was first deployed in the United States and then was introduced to other Otis regions. With OTISLINE customer service center, managers had the information they needed to truly manage service business operations. Aided by technology, Otis was able to restructure the company, eliminating several layers of management and speeding communication between field mechanics, customers, and company management. Another important IT application of the 1980s was REM° elevator monitoring. This application enabled a microprocessor-based elevator to monitor its control system and log performance statistics directly onto a distant computer. Elevators communicated problems to a computer at headquarters. The computer analyzed problems and produced trouble reports used to dispatch service mechanics before the elevator went out of service.
With data from REM elevator monitoring, service mechanics adjusted running elevators to keep them operating at maximum performance levels and could handle problems before customers were even aware of them.

The OTISLINE customer service center concept fundamentally changed the expectations of customers in the elevator maintenance industry and helped stabilize the cancellation rate.
Setting the Stage for the Next Transformation Wave Bousbib recognized that being a service company meant delivering more than flawless performance for units under maintenance. Bousbib liked the phrase, "We maintain elevators, we service customers." It meant that Otis needed to embrace a culture of service and to develop the processes needed to deliver world-class service. Otis had to redefine its processes beyond service support tools such as OTISLINE customer service center and REM elevator monitoring, and in fact beyond just the maintenance side of the business, to include new-equipment design, supply chain, new-equipment delivery, and field installation. Bousbib said, "As an engineering and manufacturing company, we did not have the same focus on customer service processes ... we needed to change the iculture and better reflect thatsweionesinsfact a service company. These new processes had to be embedded in every aspect of the company's operations." In early 2002, Bousbib decided to focus on the new-equipment business cycle and set what he described as a SHAG' for achieving fivefold improvement in the order-to-hand-over cycle within five years. Historically, this complete cycle had taken anywhere from eight months to 18 months or more depending on the customer and many other variables. Bousbib's goal was to reduce the cycle time and ensure consistent delivery of quality products and services to all customers. (See Table A.)

These gains were to be achieved through the coordinated effort of multiple ongoing change programs in engineering, supply chain, and sales and field operations. In addition, a major IT initiative, the e*Logistics program, was established to enable and sustain the quantum leap improvement.

Engineering
Otis had acquired more than 150 companies since 1995. Each brought with it its own products and operations. These diverse products had to be rationalized in order to fit into the Otis family. At the same time, Otis needed to move from a regionally driven product strategy to a global one. Traditionally the product strategy, product development, and production were managed regionally. This led to redundancy, overall inefficiency, and complexity.
In 2001, Otis began a program, the SIMBAT" program, to fundamental! chan e the r0t architecture and ultimately the design process; the goal was a standard:: _•:architecture defining modules and subsystems from which all Otis elevator rdeitte, The SHABA program reduced the number of modules in use and lowered project costs throughout the value chain. Distinct engineering programs (such as the design of new motors) fell from more than 500 to 50. Tom Saxe, who held an M.S. in engineering and an MBA, both from Stanford University, joined Otis in 1987 as an operations manager and held a variety of positions in manufacturing, branch management, and engineering before becoming vice preside:. t of engineering in 2003. Saxe recalled, "At one time, Otis held 72 types of motors in their inventory. As a result of the SIMBA program, that had been reduced to only 10 motors needed for future designs." (Exhibit 2 shows one of Otis's newest elevators; Exhibit 3 depicts the key subsystems used in the construction of a typical Otis elevator.) Part of this new process was adhering to a standard terminology that could be used with all customer settings, such as:
• Project—complete customer project (usually one contract)
• Contract—all elevators or escalators in a customer agreement
• Proposal—possible configurations of elevators or escalators
• Group—similar elevators, escalators (e.g., adjacent elevators in a high rise are a group)
• Unit—an individual elevator or escalator
• Field module—controller, machine, frame, motor, and so on (typically 20-60 depending on model)
• Subsystem—logical grouping of field modules (typically 10-15 per unit)
Supply Chain
Throughout the 1990s, Otis applied total quality management (TQM), lean manufacturing, root-cause analysis, and the teachings of quality guru Dr. Yuzuru Ito.6 A UTC proprietary program, Achieving Competitive Excellence (ACE), was designed around a standard problem-solving model and standard tools that empowered employees to identify and solve problems while at the same time practicing continuous improvement. Otis streamlined manufacturing operations from 52 factories in 1995 to 26 by 2003. Manufacturing would move to where lowest costs and highest quality levels could be achieved. Otis addressed two questions: whether or not to produce components in its value chain and, if they decided to produce, where to do the work. A number of previously manufactured items would be outsourced. In 2001,

Bousbib announced a reorganization of the supply management function and of the manufacturing function into a new single global supply chain and logistics management function that would become a key driver of the strategy. As in Tom Wolfe's description, many of the company's employees still thought of Otis as a manufacturing company. Bousbib wanted to change this mind-set and convince his organization that at a specified quality requirement, there were not any parts of the elevator that had to be made in-house for strategic reasons. "I have concluded that we never need to make anything ourselves," he said in a somewhat provocative statement that shocked many at Otis. "The only reasons why we make components are cost and logistics."7 Vincent Della Valle, a director in the supply chain organization, was responsible for the elevator passenger interface modules (including doors, cabs, carpets, lighting, and other features). His team was working hard to streamline the lead time from supply chain to manufacturing to field on a global basis. The simplest way to do this would be to produce a limited number of elevator models for sale globally. However, customer needs, code requirements, and aesthetic considerations made this impractical. Della Valle said, "We realized we could not sell a global elevator. The challenge of the supply chain organization became the coordination of multiple global suppliers located in various parts of the world. We had to work with the suppliers to ensure that aggressive delivery times were met. E*Logistics capabilities became a key requirement."
Sales and Field Operations Tony Black, a graduate of Florida Atlantic University with a B.S. in engineering and an MBA from Darden School of Business, joined Otis in June 1991, working in field offices as an account manager and later managing field operations. Since 2001, Black was responsible for quality and field operations worldwide. He initiated an improvement program called the Sales and Installation Process (SIP). Following lessons learned in the ACE program, Black's team gathered best practices from around Otis related to sales and field operations. An important objective he had was to rationalize the 65 million hours of annual labor—at an average cost of around $40 per hour—required to install, modernize, and maintain Otis products. Best practices were identified from the most successful sales offices, and these were made standard processes across the organization. For instance, the use of a prebid checklist became a requirement for all sales order personnel. This proved to be a key success factor in helping the customers define their needs, and it also helped bring sales and field installation teams together to discuss customer proposals early in the sales cycle. e*Logistics—The Information Transformation
The critical enabler of this information transformation was a project called the 'e*Logistics information transformation project. This initiative provided IT systems to facilitate business process re-engineering that was taking place throughout the company. The project team was made up of subject-matter experts in the areas of sales, field, and order management along with IT project managers who led process development requirements writing and system development. The team reported directly to a steering committee chaired by Bousbib that met every two weeks.

Giuliano Di Francesco joined Otis in the engineering organization. He held various positions in engineering, manufacturing, and logistics before becoming project director for the e*Logistics program in 2001. He was involved from the formal launch of the project. The first steps involved the merger of separate IT initiatives that had been under development for contract estimating, pricing, and online ordering. Di Francesco saw the e*Logistics program as the means for connecting sales, factory, and field operations through the Web. Technology in the project relied on open standards, Internet-based communication, workflow tools, and back-end integration with established enterprise systems. Virtually all of the technology was proprietary and built by Otis software resources and based on standardized data interfaces—such as Otis's own version of RosettaNet.8
Bousbib's view was that Otis would become "infinitely information enabled," and the e*Logistics program was the key facilitator of that information transformation. Bousbib described this vision: Traditionally we had focused on the management of physical assets. The next step in our evolution will be the management of information flows between all the participants both in the production of our elevator and escalator systems, and in the performance and delivery of our installation and maintenance processes. This is a remaking of the entire company where information flow has mastery over product flow. It is only possible because of the technology, network, World Wide Web, and new computer and software tools available today. Otis's "Global State of the Business Report 2001" highlighted two overriding objectives: "First, the continuous transformation of our business processes to improve innovation, quality, cost and speed; second, the migration of our core business competencies from product and service management to customer solution management and eventually to logistics and information management." According to Di Francesco: To achieve continuous transformation, the e*Logistics program makes sure the business process change sticks. At first, business executives saw process improvements from SIP, but after even just a few employees left, benefits fell off and became inconsistent. With the e*Logistics program, best practices from SIP are baked into the organization and institutionalized to achieve that continuous transformation.
Virtually everyone in the company would come in contact with the e*Logistics program, since it so thoroughly spanned the value chain. "That's the key thing that gets me and the team up every day," said Di Francesco, "the extent of the impact we're having on every area of the new-equipment business."
Each of the key processes, and the corresponding elements of the e*Logistics program, are described below.
Project Proposal At the very beginning of dealing with a new customer, SIP required a prebid checklist. Historically, these had been simple forms filled out on paper. With the introduction of the e*Logistics program, both sales supervisors and field-installation supervisors were required to review and approve the project as part of the prebid process. (Previously, field-installation supervisors might

have never seen sales order information until the elevator components were delivered to a job site for installation.) The new process was completely electronic and project scope and cost seen by end users as easy to execute. The new automated process solidified commitment for the estimates, a notable improvement in an area viewed as critical for two organizations that had to work together throughout the life of a customer account. Other elements of the project proposal included: 1. Gathering account information: Most Otis companies lacked sophisticated customer relationship management (CRM) systems. North America had a mainframe system; France used Siebel; many country organizations relied on simple databases running on PCs or nothing at all. 2. Determining elevator configuration (the act of putting together the proposed elevator specifications): Otis entities in Europe, Asia, and North America relied on configuration systems that ran on either mainframes or PCs, programmed into a simple spreadsheet. In fact, a number of branch locations still submitted requirements on paper and relied on the factory to prepare a configuration for the customer proposal. 3. Proposal preparation: Some countries developed front ends for creating proposals, including elevator and installation engineering drawings. Generally, there was a work template, but none of the countries had integrated to a point where they could easily push a button to prepare the proposal or readily communicate it to the factory. Prior to the e*Logistics program, the first two steps used minimal pieces of automation, and the third was mostly manual. Data was dispersed across separate systems and in most cases had to be re-entered to be shared. With the e*Logistics program, all of this became automated and the data integrated; technology handled the basic steps and allowed easy management of changes to orders. The technology had been developed over two years by Otis, with a Web-based front end, a back-end database, and workflow technology that allowed each individual user to interact with the system based upon his or her assigned roles. Finally, the e*Logistics program fed the proposal's information directly into Otis financial systems.
Sales Processing Once the customer accepted a proposal, the order then needed to be booked, validated, and scheduled. These steps had previously been handled manually, but the e*Logistics program automated the workflow of all these activities—electronically circulating key documents to all the appropriate supervisory personnel. This culminated with booking the order and feeding the data directly into the various Otis financial systems using standardized interfaces. (Hyperiont9 was used to roll the data up from around the world into corporate systems.) Once a contract was awarded, the field-installation supervisor was required to meet with the customer and review the field-installation terms of the contract. Most crucially, before the order was released to the factory, the field-installation supervisor was required to visit and assess the quality and readiness of the job site and formally note his concurrence with the timing of delivery. This

helped manage lead times, reduce inventory levels, and eliminate waste throughout the Otis value chain.
During the long sales cycle, Otis personnel routinely helped customers reengineer their orders. The e*Logistics program made the changes easy to record and visible to the entire supply chain simultaneously.
Order Fulfillment Bousbib told his supply chain team, "Ultimately, our long-term vision is to shift the center of gravity of our business from manufacturing to logistics and information management." At the heart of this shift was the creation of several regional contract logistics centers (CLCs). CLCs were responsible for accepting orders from the sales organization and delivering on-time, complete systems to job sites. CLCs did no manufacturing—they managed the supply chain including Otis factories, suppliers, field feedback, and the product improvement process. They were also responsible for market analysis, identification of customer needs, and creating product unit configurations.
Prior to CLCs, sales representatives submitted their orders by faxing or mailing documents directly to factories all around the world—the process had become unwieldy and error prone. CLCs handled all logistics and information flows between subsystem integrators (SSIs)—which were initially just the manufacturing facilities that existed at the time CLCs were first established. The SSI of the future would be the low-cost provider of a given subsystem (see Exhibit 3), which might be an Otis-owned entity or an outsourced provider. There was an emphasis on reducing internal lead times in the complete value chain of a CLC or SSI. In some areas progress happened quickly: Japan reduced its internal lead time to less than one week for some products. In other regions several weeks—and sometimes months—were the norm and became the subject of great management attention. Otis was edging closer to a just-in-time model for delivering product to job sites. At the same time the e*Logistics program was being rolled out, CLCs were also being rolled out around the globe. CLCs were built with knowledgeable resources from the manufacturing arm of the business, often staffed with personnel displaced as manufacturing operations were restructured. CLCs became centers of expertise around markets and their respective product models—such as Otis's Next Step'" escalator or Otis's Gen2TM elevator. At first, CLCs placed orders only with a single SSI, but in time the supply chain became more flexible and globally connected by the e*Logistics program. CLCs eventually could order from multiple SSIs—wherever they could find the lowest cost for required quality and delivery times. Under the direction of CLCs, the SSIs shipped their components and modules to distribution centers (DCs), which were strictly cross-docking locations on the way to job sites. Network technologies (intranets and the Internet) made it easy for CLCs to see all orders across the supply chain. Because of this, CLCs could source supplies and components from anywhere in the world. None of this would have been possible even five years earlier. Major responsibilities of the CLCs included communications with field-installation supervisors, customers, and sales representatives. Other tasks included estimating, specifying, and processing; product administration; contract engineering and layouts; customer service and contract control; contract consolidation, storage, shipment, unloading, and placement; and overall field quality.

Staffing in CLCs and DCs varied considerably by region, product responsibility, and countries/languages served. (Exhibit 4 details the CLC functions. Exhibit 5 depicts the central role of CLCs in Otis's overall value stream.) All of these activities had been enhanced through the network connectivity of the e*Logistics program. Cisco Systems in their online publication IQ Magazine noticed the transformation of Otis in these terms:
During the festivities celebrating his company's 150th anniversary in April 2003, Otis elevator's president, An Bousbib, made an unusual announcement. "We're not really a manufacturing company." Even though the company sells approximately 80,000 elevators and 8,000 escalators annually, Otis's real business, Bousbib says, is maintenance.... Otis is part of a vanguard of manufacturing firms that is beginning to look at the processes of how they build, sell, and service products not as individual processes but as one integrated process that is as collaborative as their enterprise software systems or computer networks.10

Field Installation
Historically, there had been no automated globally standardized project management tools for field installation. Site conditions were impossible for Otis to control, yet materials delivery scheduling depended on predicting when the site would be ready for elevator installation. Field coordination with order processing had been poor in the past. If equipment was ordered too early, orders could sit on a job site for weeks and be damaged by weather or vandalism. If an order was placed too late, major construction delays occurred. Otis field-installation supervisors were responsible for as many as 100 construction locations. They were responsible for communicating when sites were ready for orders to be delivered. With the e*Logistics program, they were now prompted to check site progress by workflows and could communicate job status by e-mail. Before the e*Logistics program, a number of process improvements had been put in place to improve overall sales cycle time, but application of the process was inconsistent. Black noted, "The e*Logistics program will take out the inconsistency—by automatically e-mailing reminders and updates it forces critical steps of SIP and makes good process part of the way you do business." To address cost inefficiencies and compress time to delivery, Otis moved from a push system (where product was sent into the field as soon as it was manufactured) to a pull system (where the field-installation supervisor determined when the product was needed and requested it, using electronic correspondence remotely and directly from the job site, based on its expected readiness). With the e*Logistics program, all the involved parties were asked to focus on managing to a delivery date based upon ideal site conditions—this meant holding orders until the last possible time and avoiding building slack into order and delivery schedules. This also allowed SIP to maintain lean manufacturing flows and low inventory levels. Throughout the process, the e*Logistics program introduced timed e-mail reminders for the field-installation supervisor to visit the job site. This helped the company to maintain contact with the general contractor. The e*Logistics system required status information to come back from the job site before the company shipped any items to the field.

Orders were not allowed to go to CLCs early. While the order was still sitting in a sales queue, the account representative could sell additional product options. As a building project evolved, new elevator features could be added (e.g., mirrors, lighting fixtures, and a range of premium finishes) before the order was released to the CLC. This simplified and lowered the cost of change order management, both for customer billing and for production. With more visibility to the order process and more control over shipping dates, the field-installation supervisor had increasing control over the financial success of each project. As one person described the shift, "We need to change field-installation supervisors from being firefighters to being business managers. That's a huge cultural change." The new processes helped eliminate many hidden costs of lost, stolen, damaged, and misplaced material. As Black said, "SIP tried to get the field-installation teams to manage the delivery date based on job-site progress, but e*Logistics helps to bake good process in."
Closing Activities The principal activities in this phase were transitioning the unit from Otis installation to Otis service and handing the elevator over to the customer. This stage included reviewing any engineering change requests and processing final billings. The process varied greatly by geography and even by contract but generally involved meeting with the customer and presenting the final bill to review and settle all change orders. With e*Logistics, as soon as the supervisor confirmed the job was complete, a workflow was triggered prompting a series of customer contacts and billing. This was expected to result in more accurate billing of change orders, higher conversion of new equipment to maintenance contracts, and faster collections.
Information Technology at Otis Ron Beaver, CIO, joined Otis in 1989. He initially worked in sales and marketing and became CIO in 2001. Beaver held responsibility for global IT, with seven regional IT leaders around the world reporting to him on a dotted-line basis. Each regional IT director reported directly to the president of one of these seven Otis regions: North and South America, Southern Europe and the Middle East, North and East Europe, the U.K. and Central Europe, North Asia, South Asia, and Japan. Regional IT leaders were responsible for all local systems and operations, supported by an outsourced infrastructure agreement with CSC. They had their own local development team and technology support services. Beaver noted, "We're a very diverse, $8 billion company with operations in a thousand branch offices. As a company, we grew through acquisitions. We have 50 financial systems and 15 different manufacturing systems operating all around the world." The Otis infrastructure included nearly 1,000 local-area networks with 600 wide-area networks (WANs) in the U.S. and 400 WANs internationally. There were 30 types of routers and switches supporting 3,000 applications. The Otis global network processed 60,000 orders each year—resulting from approximately 300,000 proposals. Otis did not break out IT investment for the e*Logistics program from other process-engineering expenses such as those for ACE or SIP, nor did it account for return on investment on IT separately from returns associated with other business-process change initiatives. All these were considered

together because the belief was that the benefits were tightly interrelated and could not be decoupled as easily as the costs might be. Beaver had dotted- or direct-line responsibility for 450 Otis IT employees, in addition to 250 contractors. He had budgetary responsibility for all major IT expenditures. About 70 people from Beaver's organization were dedicated to the e*Logistics project, reporting to Di Francesco. Beaver described the integration of e*Logistics with legacy systems: "The goal of having a single finance and/or manufacturing system in Otis is not achievable in the short term. Therefore, building the e*Logistics system to be ERP [enterprise resource planning] system-agnostic was essential to obtaining the business benefit quickly." Beaver's staff was also responsible for a middleware broker used on the e*Logistics program—much of it custom coded for Otis, with some development taking place in an Otis-owned, India-based software company. Internal sales modules were built with Microsoft .NET development tools. The supply chain and field systems were built around JD Edwards ONEWORLD software. Systems access, security, and connectivity relied on the UTC employee portal.
Remaining Challenges for the e*Logistics Program There were a number of technological challenges with the e*Logistics program. According to Beaver:
The first is the delivery of the e*Logistics program to the desktop. We have more than 20,000 PCs in the company and over 1,000 different major locations. So our ability to standardize the infrastructure and then understand how this application can perform across wide-area networks, local-area networks, etc., was very important to us. In the last two years we've obtained emerging technology that allows us to simulate applications across our global infrastructure. If one of the switches is wrong, or an Internet browser configuration setting is wrong, it impacts the performance of these applications. Now we can simulate this, sitting anywhere in the world but operating systems as though we are in Korea, Japan, China, or Russia. There were also many personnel challenges. There were 1,500 sales representatives and 1,500 field-installation supervisors around the world who would use this system. Substantial technology and process training was needed for all users. Pilot projects in the Netherlands, France, Germany, and the U.K. during 2003 had been successful. The U.K., which had a particularly long sales cycle, was able to cut its average sales cycle time for orders processed through the e*Logistics program from more than six months to less than three (corresponding to the first half of the complete cycle described in Table A). While the success was due to a number of factors—a new country manager, an involved management team, effective SIP deployment, local initiatives, and other programs—everyone recognized the critical role of the e*Logistics program in making and sustaining this reduction in cycle time possible. From the pilots, Otis learned that it took approximately six months for benefits to become visible after an e*Logistics installation. Data entry and systems initialization would take two to three months, technology and tools training would take a month, and three to four months would be required for staff to fully adapt to the new business processes.

Full-scale rollout in Europe and China was scheduled for 2005. The system would be deployed for the rest of the company by 2006. Regions were prioritized based upon need and return on investment, with areas with the least developed systems infrastructure in their business coming first. Adoption and success rates of the new system would be tracked by several metrics, including the number of units proposed per salesperson using the e*Logistics program and the ratio of sales support people to sales representatives. The supply chain showed improvement in factory turns and lead time as a result of the e*Logistics program. Field-installation times had been materially reduced for regions using the e*Logistics program, but scaling the system to be used by all of the field-installation supervisors remained the biggest challenge.

The Next Transformation
The transformation at Otis required a tremendous amount of effort on the part of management personnel and every individual employee of the company. A number of important improvement programs had been running in parallel, and they worked together to multiply the benefits. Black observed:
The rollout of the e*Logistics program is based on the maturity of SIP. If we hadn't done SIP, and just did the &logistics program, it would have taken much longer. Some of the rollout is dependent on people's adaptability to change. In North America, we've put laptops into the hands of field-installation supervisors. They are salaried, nonunion personnel who have usually come from the union ranks. About 30% of them are familiar with computers, and this only works when people know how to use the technology. Bousbib was pleased with the progress of the e*Logistics program rollout to the pilot countries in Europe. There had been approximately 300 person-years of effort invested in e*Logistics , and Bousbib realized how much work remained to gain full benefit from their global investment in IT. By early 2004, all of Otis's major products could be ordered through CLCs in certain regions. The global deployment of the e*Logistics program was proceeding according to schedule. A simple measure marking deployment progress was the number of new orders that went full cycle through the e*Logistics program. In the U.K., for example, that amounted to 5% of all the units sold in all of 2003. But the U.K. was expected to rely solely on the e*Logistics program by the beginning of 2005 for 100% of its orders. Otis had not yet decided on a global cut-over date when all orders would flow through the e*Logistics systems. Between 2000 and 2003, Otis revenue and operating profit had grown. But at the same time, working capital turns nearly doubled. Improvement in information flow had contributed significantly to improvement in cash flow. Bousbib took stock of his time at UTC and Otis: "We have outstanding people with outstanding values, and we can become an ever-greater company. What we have undertaken requires a remaking of the entire company. When we launched our CLCs and SSIs and deployed the e*Logistics program it was a mini revolution, but today's results are proving the success of this approach." Over time, Otis realized that there was not a direct correlation between product quality (as evidenced by callback rate) and cancellation rate. Beyond a certain minimum expected level of product reliability, Otis concluded that the customer canceled maintenance agreements because of dissatisfaction with the quality of service, not the quality of the product.

Even as the e*Logistics program was still rolling out, Bousbib challenged his executive staff to launch a new initiative that would help make Otis the number one service company in the world—not just in the elevator business but compared against the greatest service companies ever: "Our quest for service excellence will now require another mini revolution—to become number one in service—but this will also make us an ever-greater company." United Technologies Company (UTC), the parent company, was interested in hearing more about the progress of the e*Logistics program and also about the challenges inherent in Otis's new quest for service excellence.