Need to do this human resource case analysis project

microsoft_vega_case.pdf

Home >Human Resource Management homework help >Need to do this human resource case analysis project

Harvard Business School 9-300-004 Rev. January 8, 2001

Professor Christopher A. Bartlett and Research Associate Meg Wozny prepared this case as the basis for class discussion rather than to illustrate either effective or ineffective handling of an administrative situation. Some names, data, and details have been disguised to protect confidential information.

Copyright © 1999 by the President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call 1-800-545-7685, write Harvard Business School Publishing, Boston, MA 02163, or go to http://www.hbsp.harvard.edu. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of Harvard Business School.

Microsoft's Vega Project: Developing People and Products

In March 1998, Jim Kaplan, product unit manager for Microsoft's Interactive Media Products Group, was deep in thought as he left one of his regular meetings with Matt MacLellan, a 26-year-old 5-year Microsoft veteran. MacLellan had spent the past three years managing the Vega team’s development of an innovative interactive learning system designed to improve student involvement and retention. Kaplan felt a great deal of pride in the project—“probably the most rewarding experience I’ve had as a manager”—not only because the product was based on his original idea, but even more important, because he’d been able to coach and develop MacLellan as a manager.

Now, however, his young protégé had just told him that he was not getting satisfaction in his management role and wanted to concentrate on writing software as a developer—or Software Design Engineer (SDE), as they were more formally designated. Walking back to his office, Kaplan tried to sort through the questions and concerns racing through his mind. As MacLellan’s mentor, should he support such a radical career shift when Matt had five years of project and program management experience but had never worked as a professional software developer? Even if he did, how could he manage such a move, not only to Matt’s satisfaction but also in the organization’s best interest? Of particular concern was the performance level he could justify for MacLellan as a developer. Despite his deep knowledge of the company’s products, technologies, and organization, Matt’s lack of applied software development experience would make it hard to support a rating that maintained his current compensation package. It was an issue Kaplan realized he would have to think about very hard.

Microsoft's Human Resource Practices: Making People Strategic Assets1

Throughout the meteoric two decades of growth that made it the world’s most valuable company, Microsoft’s leaders attributed its outstanding performance to one core capability: its continued ability to recruit, develop, motivate, and retain exceptionally capable people—“the best team of software professionals the world has ever seen,” as CEO Bill Gates liked to boast. And although the policies and practices that the company had developed during its brief life were often quite different from standard human resource approaches, within Microsoft they were part of the deeply embedded management philosophy.

1 This section summarizes “Microsoft: Competing on Talent,” HBS No. 300-001.

This document is authorized for use only by felipe montoya ([email protected]). Copying or posting is an infringement of copyright. Please contact [email protected] or 800-988-0886 for additional copies.

300-004 Microsoft's Vega Project: Developing People and Products

Recruiting the Best and Brightest

From the start-up days, Gates recognized that success depended on hiring exceptional people. “We’re in the intellectual property business,” he said. “It’s the effectiveness of our developers that determines our success.” Above all, he wanted to raise the bar through recruiting. Commented a human resources executive, “What Bill has always instilled in us is to hire people who are better than we are.” As one magazine observed, “Microsoft has been led by a man widely recognized as a genius in his own right, who has had the foresight to recognize the genius in others.”a

From the day he was hired as Gates’s assistant in 1979, Steve Ballmer became Microsoft’s first recruiting coordinator. His mantra was, “We want people who are smart, who work hard, and who get things done.” This combination of “horsepower and drive” was to shape Microsoft’s recruiting for the next two decades. Once the smartest and most driven were identified, they were pursued relentlessly. “There’s a standing policy here,” said Ballmer, “whenever you meet a kick-ass guy, get him. There are some guys you meet only once in a lifetime. So why screw around?”

Candidates were subjected to an intense interview process, involving up to 10 Microsoft employees. The recruiting process was particularly rigorous for developers, who were tested not only on their technical competence. Oddball questions like, “Why are manholes round,” were aimed at testing the candidate’s deductive reasoning, creative problem solving, and composure. As soon as the interview was over, each interviewer would send email to all other interviewers, starting with the words “Hire” or “No Hire,” followed by specific feedback and suggestions for follow-up by the next interviewer. (See Exhibit 1 for an interview feedback email.) To assure that only top candidates received offers, after the hiring manager made the final hire/no hire recommendation, a so-called “as appropriate” interviewer was also involved. A senior manager explained:

Very often, the “as appropriate” interviewer is a person who is outside the hiring group, someone solidly grounded in Microsoft culture and committed to making sure that we hire people who are going to be good Microsoft people, not just good people for specific jobs. That person has veto power, which puts a system of checks and balances in, because the hiring manager may feel a lot of pressure to fill a job, while the “as appropriate” interviewer doesn’t.

Microsoft’s tight control on headcount reinforced the pressure to resist settling for the merely satisfactory candidate. Even when the company was growing rapidly, Gates and Ballmer insisted that the company employ fewer employees than were required to carry out the work. The formula was n minus 1, where n was the number of people really needed. Said a senior HR manager: “[Beyond hiring smart, driven people] Ballmer’s second principle was that the default decision on a candidate is ‘no-hire.’ In other words, unless you can identify a very clear reason why we should hire someone, we shouldn’t hire him or her.”

As the demand for new employees grew (6,660 in 1998, of which 4,823 were new positions), the company’s recruiting practices became more specialized and formalized. But Gates and Ballmer insisted that all managers keep deep personal involvement in this activity to ensure the continuance of the high standard. For example, although full-time campus recruiters in the United States sorted and screened candidates from the annual crop of 25,000 computer science graduates—2,600 were targeted for campus interviews, yielding 800 candidates invited to visit the company—it was management’s responsibility to identify—and woo—the right candidates from that screened pool. Despite the large numbers, the process continued to yield an impressive rate of acceptance. Of the 800 college recruits invited to Redmond, for example, about 500 received offers and almost 400 accepted. (See Exhibit 2.)

Increasingly, however, Microsoft was unable to recruit the bulk of its new employees from campuses as it had done in the early days. By the mid-1990s, over 75% of new hires were

Microsoft's Vega Project: Developing People and Products 300-004

experienced people attracted from within the industry, a trend that was reflected in the changing employee demographics. (See Exhibit 3.) A “strike force” of 300 full-time recruiters was set up to locate and build relationships with the industry’s most capable people and attract them to Microsoft.

Microsoft’s HR specialists also worked hard to make sure the new recruits were placed on the jobs for which they were best suited. Internal research had shown that both productivity and retention were very highly correlated to whether the individual was engaged and excited by his or her specific job. As one HR manager explained, “Our first priority is people quality, but the second issue is getting them to the right place, where they are going to have a positive impact on Microsoft and where they’re going to feel empowered to do great things here.”

Gates continued to insist that no activity was more important for him than meeting superior candidates to convince them that they should join Microsoft. He even targeted the summer interns, inviting all 600 in groups to evenings at his $60 million home. Gates also regularly telephoned wavering star prospects, even at levels well below those he directly supervised. “He still calls up college kids who are key people that we want to hire,” said Microsoft’s recruiting director David Pritchard. “That blows them out of the water. They never believe it’s him.”

At the Microsoft Financial Analyst Meeting in 1998, Microsoft’s executive vice president Bob Herbold said that the “people issues” were “the critical determinant” of the company’s success. “When people ask what makes Microsoft special, the number one reason I always give is incredible competency in regard to recruiting and training and developing people,” said Herbold.

Microsoft's Work Environment: The Caffeine Culture

Microsoft’s cultural norms could be traced back to the company’s start-up days when Gates, Paul Allen, and four programmers created a hot-house of innovation and hard work. Software developers dominated the company, and up until the early 1980s, Gates knew all their names, faces, and telephone extensions by heart. Said one programmer hired from MIT in the early 1980s:

The software engineers got the best of everything. From the day you started, you got your own office. That was really nice, really a motivating thing. It was about the only company where you could get your own office at that stage in your career. There was a lot of respect for a young kid right out of college.

Yet to many, Microsoft’s resource-constrained, intellect-driven management model was disorganized, even chaotic. A technical manager who joined Microsoft from Xerox in 1983 explained:

There were a lot of people at that time who were absolutely incompetent at managing people. It was the Peter Principle: very successful technical people would get promoted to management roles. You’d get 30 people reporting to one guy who was not on speaking terms with two-thirds of the group, which is inconceivable.

By 1986, Microsoft’s nearly 1,200 employees moved into new offices in Redmond, Washington. The low-slung buildings nestled into the 29-acre wooded “campus” were designed in the shape of an X to maximize the number of windows. Unlike in the open-plan buildings popular elsewhere, each employee still had a fully enclosed 9’ x 12’ office with a door, to ensure privacy necessary to “sit and think.” Numerous cafeterias, with food at prices subsidized by the company, facilitated social interaction. Microsoft spent more than $8,000 per employee each year on nonmandated benefits, with more than $715 a year per employee on beverages and food subsidies alone. As one employee noted, “Anything with caffeine is free.” In many ways, it had the feeling of a college campus and provided a comfortable post-college sense of familiarity and belonging. (See Exhibits 4 and 5 for campus and office pictures.)

300-004 Microsoft's Vega Project: Developing People and Products

Although employees’ average age moved above 30 in the mid-1990s, the culture remained remarkably unchanged: employees dressed informally, there were no status symbols, and the early ethos of thrift remained. There were no set work hours, but the culture attracted those comfortable with 14-hour days and working weekends. Yet motivation and morale—routinely measured in internal surveys—remained high. (“Our efforts are significant to millions of people—it’s the old change-the-world thing,” one explained.) But burnout was a continual concern. As one employee noted, “One year here is like three years elsewhere.”

Through all the growth, Gates struggled to keep alive the small-company “change the world” feeling. He continually restructured the organization into small units—with 30 to 200 people— further dividing these into working groups with responsibility for a product, project, or program. Much of the coordination as well as overall direction setting and control were still managed by Gates well into the 1990s. Using his legendary intellect, he managed through a wide-ranging “hands on” style that he employed to question, challenge, and cajole. Having little patience with the notion of “professional management” as a separate group, he told his organization:

I really don’t know the difference between a professional manager and anyone else. . . . We’re not here to say, “I’m a professional manager, give me something to manage,” we’re here to get the job done. So we don’t actively distinguish between professional and nonprofessional managers.

Historically, Microsoft had never experienced big retention problems in an industry where 15% to 20% turnover was common. Microsoft’s 1998 attrition rate (measuring voluntary separation) was below 7%. (See Exhibit 6.) (Dismissal was extremely rare, since most who could not meet performance standards or operate under stress sought alternative opportunities.) However, when its HR group delved deeper into the numbers, they realized that among employees who had been with the company for seven or more years, and for those in senior positions, the attrition rate came close to the industry average.

Reflecting its concern, management hired an outside organization to interview ex-employees and to administer a standardized survey to find out how current employees felt about working at Microsoft. The results were generally very positive, confirming the company’s own surveys. For example, 88% answered “agree” or “strongly agree” to the statement “I like the kind of work I do,” and 75% to the statement “My work gives me a feeling of personal accomplishment.” The consulting firm indicated it had never before recorded such strong positive results. (See Exhibit 7 for sample survey questions.) Still, management recognized that in an increasingly competitive market, it would have to work hard to prevent the loss of talent and experience.

Development Through Stretch and Challenge

In Microsoft, there was a strong belief that smart, driven people (“hard core” in Microsoft terminology) were best developed through challenging and engaging assignments. The company’s rapid growth coupled with its “n minus 1” staffing philosophy ensured that people were thrown into stretching assignments early. A senior HR executive explained:

We have very limited educational and training opportunities for our managers. But I think that we have absolutely developed leaders. . . . You get people having to move from managing 10 people to managing 200 people overnight. That kind of stretch in the job will either create growth or death. Fortunately, we have such great people that most of them have just grown by leaps and bounds.

Because the company recruited primarily technical experts, it was important to allow people to develop along either of two career paths—one in the technical domains (such as development and testing) and another as a management track (in the product groups, for example, or at the corporate

Microsoft's Vega Project: Developing People and Products 300-004

level). The technical career paths were essential to retain skilled people and afford them the same recognition and compensation as those who advanced as managers. Although the titles did not always reflect roles accurately, the typical career path was to move from being a new hire to being a mentor, a team lead, then a team manager in one functional area of a product (e.g., development manager for Word, or test manager for Excel). Above these managers were senior level positions that integrated functional activities or cut across product units.

As this dual track process developed, management felt the need to clarify the incentives and bring some comparability across the functional disciplines. In 1983, the first “ladder levels” were established to assist in recruiting developers. For example, a new developer hired from college at Level 30 would typically spend 6 to 18 months before moving up a level after demonstrating the ability to write production code without a lot of supervision. Promotions from Level 31 to 32 typically occurred at 6-month to 2-year intervals, but moving from Level 32 to 33 was a major step. The equivalent of making partner in a law firm, it required that the manager describe the candidate’s contributions and potential to the director of development and an executive vice president. Promotions to levels 34 and 35 required Gates’s approval. (See Exhibit 8 for ladder level examples.)

The company encouraged its employees to switch jobs internally to develop themselves and keep their interest. Prior success was not necessarily the key qualification for the new job since Microsoft had a long tradition of promoting those involved in failed projects. An often-quoted sentiment was, “If you fire the person who failed, you’re throwing away the learning.” Beyond experiential learning, Microsoft put great faith in development through coaching and mentoring as John Agnew, director of Executive and Management Development, explained:

We believe that development is not about training. It’s about jobs. So the company has had a tendency to put bright people with high potential in really huge jobs and test the hell out of them. The next most important thing is about the kind of coaching you are getting while on the job by your boss and important others.

New hires understood it was their responsibility to learn from experienced people like team leaders, experts, and particularly their formally appointed mentors who, in addition to doing their own work, took on the primary teaching responsibilities. The practice was deeply embedded in Microsoft’s values, and every manager accepted it as an important part of his or her responsibility. Gates himself also played this vital role in the development of thousands of employees. Throughout the year, project teams, from billion-dollar businesses to start-up ideas, presented their progress, current status, and future plans to the CEO in the legendary “Bill Meetings.”

In order to bring greater efficiency and consistency to employee development as Microsoft grew, the corporate HR group began to formalize the process of identifying current as well as future leaders. Such initiatives got a boost when Bob Herbold joined Microsoft as chief operating officer in 1994. After 26 years with Procter & Gamble, where leadership development was taken very seriously, he convinced Gates to focus the organization’s attention by reviewing key people with Microsoft’s top executives on a more regular and formal basis. By the late 1990s, VPs were required to identify their high-potential people who were then tracked through a companywide inventory.

Review and Reward: The Options-Driven Engine

Reflecting Gates’s belief that shared ownership motivated and retained employees, even in the days when Microsoft was structured as a partnership, key employees were given equity in lieu of high salaries. Equally well established was the linkage between individual performance and reward—primarily in the form of stock options, to conserve cash in the fast-growing start-up. Gates’s style was to give employees frequent and typically brutally honest performance reviews, a norm that became institutionalized in companywide semi-annual reviews tied to pay raises, bonus awards, and stock option grants.

300-004 Microsoft's Vega Project: Developing People and Products

Gates’s belief in setting specific quantifiable objectives also became part of the process, with each individual committing in writing to measurable performance objectives every six months—for example, a developer might agree to complete three modules of code or reduce the number of bugs from 1,000 to 50. Eventually, the acronym SMART was applied to performance objectives—Specific, Measurable, Attainable, Results-based, and Time-bound. Finally, the reviews reflected Gates’s obsession with learning from mistakes—often referred to as “the disease model of management.” Review sessions were routinely punctuated with questions such as, “What did we learn?” or “What could we have done better?”

In the late-1980s the performance review system incorporated a 1-to-5 performance scale tied to a forced evaluation curve in which 25% of employees received evaluations of 3.0 or lower, 40% received 3.5, and 35% received 4.0 or higher. At the end of each six-month period, every employee filled out a performance review form, describing what he or she achieved and what did not go well and providing a self-evaluated score. Then, in a face-to-face discussion, the manager provided the employee with his or her rating. A score of 3.0 or lower was regarded as undesirable and a 2.5 rating or below usually meant the employee was on the way out the door. On the other hand, a score of 4.0 or above was good news; there were very few 4.5 and only two or three 5.0 scores each year—an honor that warranted a personal visit from Gates. (See Exhibit 9 for sample performance review form.)

Special software allowed the manager to link the employee’s rating to a suggested merit increase range (typically 4% to 6%) but in keeping with its historical roots, Microsoft’s salaries were typically lower than at rival companies. In contrast, bonuses to 15% of salary were common, and stock options were generous. (In Microsoft, salary was recognized as reflecting historical value, bonus was for current performance, and options were awarded on the basis of one’s future worth to the company.) Performance-review surprises were rare since most Microsoft managers reviewed performance against objectives every month or so prior to the formal review.

Once a year, managers rated their subordinates on their eligibility for stock options on a scale from A to D—with A being reserved for employees viewed as tremendous long-term assets to the company. After a manager proposed the amount of shares an employee should get, a manager one level above reviewed the proposal and submitted it for a VP’s approval. The 15% of employees who did not receive stock options during any review period were given an implicit message that the company was not interested in investing in them. (See Exhibit 10 for compensation ranges.)

In 1999, analysts estimated that at least 10,000 current Microsoft employees had each been granted options worth more than $1 million, while those around since the 1986 public offering had seen a 750-fold stock runup, creating several billionaires. What surprised many was the number of employees who continued to work at Microsoft even when they were financially secure beyond their wildest dreams. When people did leave, Microsoft’s research suggested, it was usually not because they were rich enough to give up work but because they were burned out or the sense of personal growth or challenge had run out for them.

It was in this corporate environment and operating with these human resource philosophies, policies, and practices that Jim Kaplan gave birth to an idea that would evolve into the Vega program. The rest of the case focuses on how he developed the product idea and the people supporting it.

Flying Vega: Microsoft's HR Policies in Practice

In the summer of 1996, Jim Kaplan, product unit manager for Microsoft's Interactive Media Products Group was itching to launch a new project. His idea was to create a truly interactive toy

Microsoft's Vega Project: Developing People and Products 300-004

that could talk and respond to a child’s voice and touch as well as become a medium through which the child could interact with a home computer. Kaplan’s intention was to not only create an interactive toy but a learning system designed specifically for an under-served market: girls ages 3 and up. However, because he was responsible for developing Microsoft’s children’s interactive encyclopedia which required his undivided attention at the time, he started looking for someone who could focus on his new product idea full-time. His boss, Vikram Puri, suggested that he talk to Matt MacLellan, a young program manager who had worked on Microsoft’s massive Oasis program and who was ready to move to a smaller project where he could learn how to build and ship product. Recalled Kaplan:

Vikram said to me: “You’ve shipped a dozen products. You’ve got this great product idea, so why not get it started with Matt. This is the challenge he’s been looking for, and it would be a great way to teach him product development discipline.”

Wooing the Candidate

Puri’s intuition about MacLellan needing a change was well founded. In fact, at the time, Matt was seriously contemplating leaving Microsoft. Since joining the company out of Harvard in 1993, he had been working on Oasis, a huge advanced operating-system technology project. Unfortunately, the group he worked with for three years ended up having problems building a product, which proved frustrating. Commented MacLellan:

It was a fairly futuristic project, a bit over-ambitious, and we all got kind of mired down trying to put it all together. . . . I learned a lot about technology and about designing, but I hadn't had the experience of actually shipping something to customers—really building a product—which is what Microsoft is all about. I realized I wanted a change.

The situation presented Kaplan with a double challenge. For the first time in his career, he was about to step back and let someone else take charge of all stages of the product lifecycle. But since MacLellan was on the verge of leaving, an even bigger problem was the issue of retention. Kaplan felt he could manage the situation. He had a well-established record of managing morale and creating an exciting work environment and considerable experience in managing bright young people who needed a challenge. He explained his approach:

I don’t necessarily try to keep an employee who wants to leave. Instead, I try to learn what it is that the employee’s seeking and what the right next step in their development is. I then try to enable that. The first possibility is to see if the employee can stay in the role that they are currently in. The next possibility is to stay within the organization, then within the company, and then within the industry. Finally, to stay within life—not “drop out.” I take an expanding radius approach.

After talking to MacLellan, Kaplan concluded that his desire to escape the frustration of being stuck in the design phase of product development was stronger than the desire to find the right project to work on: “The advice I gave him was, ‘Make sure that whatever you do, you’re running towards something and not away from something.’” He outlined the new interactive learning system product idea he wanted to develop and asked MacLellan to think about whether it was something that could really excite him. At the same time, he encouraged him to talk to other groups within Microsoft and even look at opportunities in other companies. In short, he told him to create “a landscape of opportunities” for himself, then decide which one was most attractive.

The two agreed to meet again a week later. In the meantime, MacLellan talked to several groups within Microsoft. But he also spent about 40 hours researching the new interactive toy

300-004 Microsoft's Vega Project: Developing People and Products

product idea Kaplan had presented to him. The more he thought about the possibilities, the more excited he became. He recalled the process:

Speaking with Jim got me excited about the product. I could see that he had a vision, passion, and smarts, and he made it feel like a situation I wanted to be in. He didn’t hard-sell me, but he got me excited about the idea. He actually had a couple of people from the team come and talk to me, and really was active in getting me a lot of information and making me excited about the possibility.

Having done extensive research, MacLellan realized that there were no truly interactive, PC- compatible learning toys on the market. A newly minted father of a four-month-old baby girl, he was not surprised to find that of thousands of children’s software titles on the market, only a handful were created exclusively for young girls. He instantly recognized it as a great opportunity: “That was really kind of the spark for me. It was clear that we could build something new and exciting aimed at girls, something that would compensate in an arena in which most computer games and electronic toys—with their correlated technological proficiency—seemed targeted predominantly toward boys.” By the end of the week, MacLellan created a five-page document in which he reflected the original product vision, along with his own ideas. He took it to his meeting with Kaplan.

At the meeting, MacLellan reported that he was excited and ready to start working on the interactive learning system project. The fact that Kaplan did not pressure him to take on the job without exploring other opportunities seemed to be what he needed. “I created an option for him, but I didn’t put the screws on him. . . . That relieved him of the burden immediately,” said Kaplan. He offered Matt a job as a program manager on the Interactive Learning System (ILS) team.

Coming on Board

With Kaplan’s encouragement, MacLellan used his five-page vision document as his starting point. He was excited about the new role and explained his responsibilities:

A program manager isn’t necessarily a manager of other people. They manage a product or a piece of a product not from the marketing side, but from the side of actually designing, developing, and building it. An analogy might be a director in a play or a general contractor at a building site. They are in charge of making sure other people are doing the right things and everything works together. They play a lot of different roles. They need to understand the technology but they also need to work with the marketing people and think about a wide range of issues.

MacLellan’s first priority was to develop an overall product vision—to make an interactive toy that would introduce girls, ages 3 and up, to computers and provide exciting, interactive learning via games—then develop a design to meet that vision. “That’s the part that I really enjoy. I get a lot of excitement out of doing that design work that takes a high-level idea and breaks it down into some architecture of software needed to do it,” he said. He contacted a California-based company that had conducted research among 1,500 girls, 700 parents, and various experts over a period of several years and found out that girls were significantly less interested than boys in games where the key character died and started over. Instead, they preferred games with higher social interaction. MacLellan came up with an idea of creating an animatronic toy that would be not only an exciting learning buddy, but also a great role model. An avid amateur aviator, Matt envisioned a game built around adventures with Amelia Earhart, the first woman to fly solo across the Atlantic.

The animatronic doll—Amelia—would have two modes. As a freestanding unit, she was a cuddly plush toy—by squeezing her in different ways, a child could have simple conversations, play games, or hear her sing children’s songs. Plugged into a computer and powered by the accompanying edutainment software, focusing on math and reading skills appropriate for ages 3-5,

Microsoft's Vega Project: Developing People and Products 300-004

4-6, and 5-7—Amelia would engage the child in fun, interactive learning, offering encouragement, advice, and on-screen tutorials. The educational content would be energized by creative games, inviting the child to help Amelia on her adventure-studded flight missions by solving age- appropriate puzzles. The software bundle was designed to offer content reviews in different formats to facilitate learning in a variety of skills from simple math and sentence concepts to geography.

However, Matt’s initial euphoria was soon brought back to earth by the fact that there were not enough people to build the product. Feeling some frustration, he took the problem to Kaplan, telling him, “We have this neat stuff and the time is running out.” In response, his boss managed to get an addition to his headcount and started looking for internal candidates to support the project. However, he told MacLellan that he wanted him to be deeply involved in the interviewing process and the hiring decisions for the program.

About the same time, Kaplan began investigating Purple Ostrich, a small company based in Houston, Texas, which had developed a unique animation technology called Truemotion. Kaplan asked his newly appointed program manager to check it out as a possible acquisition. Although Kaplan made the final decision to acquire the company, MacLellan felt good that he was involved in the process. “I flew out to Houston to meet with the people there and to get a sense of them. I also had to explain what we were working on and sell them on some of the ideas,” he recalled.

As the Houston team was relocated and integrated and the project picked up momentum, MacLellan grew into his new responsibilities. Kaplan was particularly impressed with how MacLellan built a real identity for the project and an esprit de corps among those working on it. An avid amateur pilot with a goal to replicate Charles Lindbergh’s solo flight across the Atlantic, Matt called the project “Vega,” after Amelia Earhart’s Lockheed Vega plane on which she completed her solo transatlantic expedition. As the project progressed, the Vega theme became more prominent. The team created T-shirts with the picture of Earhart’s Vega on them, featured Vega on the programming credits, and even made the pass code to the software test lab 52032—the date of Earhart’s solo Atlantic flight, May 20, 1932. Kaplan recalled the impact on team motivation:

Matt created this incredible team environment. Since we couldn’t actually go flying together, Matt came up with this idea to go bungee jumping in Issaquah. We all went. He got this team of about 12 or 15 people that he wasn’t actually managing—but he was a spiritual leader and a very effective one at that—and he just turned it around. And my coaching was really about the spiritual leadership that came very naturally to him. My real value added was in helping him develop discipline and in being a program manager who could actually make it happen.

For the first time in his career, Kaplan realized that his role had become that of a coach rather than program manager. “Normally, I’m kind of the evangelist—the one who stands on the stage and mixes the Kool-Aid,” he recalled. “But this was my first opportunity to go into a true coach mode. I had to take a big step back and give the spotlight to someone else. I became Mr. Behind-the-Scenes.” For MacLellan, whose previous experience as a program manager did not go beyond the design stage, the new situation of getting a product out also presented a big challenge. “I was coming from a position of never really building the product or having the whole team working on building things,” he said. “I had to go from a place that I was comfortable—designing things—to this new world of actually building things.” This role involved a lot more coordination and integration of activities and all the people issues that came with it. He recalled:

Seattle is a nice place to live, but if you’re coming from Texas, it’s an adjustment. There were issues with the move and with integrating into a different corporate culture. Even though I was not directly overseeing those people, I was aware of some problems. One person who really got burned out ended up quitting.

300-004 Microsoft's Vega Project: Developing People and Products

About halfway through the project, in one of the regular conversations Kaplan had with MacLellan to find out how everything was going and how Matt felt about his new role, the young program manager reported that he was very satisfied with the extent of his professional growth, appreciated the freedom he’d been given, and acknowledged that from a career standpoint, the experience had greatly increased his marketability. At the same time, however, he reported feeling somewhat emotionally disconnected from the product since he was not the one who physically designed, built, or tested it. He recognized his role as a catalyst and acknowledged that his presence had a positive impact on the team, but was confused about his value added. Mostly, however, he told his mentor he was lacking satisfaction and fulfillment in his job. Kaplan felt his role was to bring that sense of reward back to MacLellan:

I said to him, “Look at the product. Now, close your eyes and pretend you didn’t exist. The name of the product wouldn’t be the same. The concept wouldn’t be the same. The team wouldn’t be as enthusiastic about it. We would probably be three months behind schedule. The quality of the product wouldn’t be the same. From a competitive standpoint, we wouldn’t be as aggressive. And the team wouldn’t be having as much fun. That’s how you have to build your reward system, every day. Look at where we are.”

As the exchange continued, MacLellan began to wonder if he really wanted to be a “general contractor,” or whether he didn’t enjoy more the challenge of “building stuff, getting dirty, and hammering nails.” But his boss pushed him to think about finding satisfaction by taking on the challenge of becoming a true manager. He offered him a job as ILS’s lead program manager. As Kaplan explained it, this would require MacLellan to move beyond a coordinating role to actually take responsibility for managing people. “You’re going to get even less dirty now,” he said. You are going to have people who get dirty and whom you tell what to do.”

The Management Experience

As soon as he assigned MacLellan to his new management responsibilities, Kaplan moved several people on the team to work directly for him. But he also enacted one of his management principles which held that newly appointed managers should be responsible for their own hiring decisions. He reflected on the origins of the philosophy: “When I was a manager, I was often given people that weren’t the people that I would have necessarily hired. One thing I struggled with was really reshaping the roles and getting the people on track. I felt like I inherited problems due to others’ mistakes, and that frustrated me.” As a result, whenever he could, Kaplan gave newly appointed managers an authorized headcount to fill.

With two headcount positions open for him, MacLellan ended up hiring one internal candidate and one outsider, Mark Reyner, a fresh graduate from Duke with a degree in computer science. Reyner was one of the first candidates MacLellan interviewed, and almost immediately struck Matt as an exceptional individual:

I wanted to make sure I was hiring someone really strong and compatible. Mark was really sharp, really motivated, and he had gotten things done. And he had a real understanding of different technologies. I felt as if I were talking to someone already pretty senior at Microsoft. I knew that this guy was a star.

The previous summer, Reyner had been one of seven interns selected from a very competitive field at Duke (more than 80% of the computer science department applied for the job). He was one of five of that group interviewing for positions as full-time employees beginning in June 1997. Reyner explained the process of being recruited:

Microsoft's Vega Project: Developing People and Products 300-004

I have the utmost respect for the abilities of Ron Padgett [the manager of Campus Recruiting for Duke]. He was just excellent at finding the right fit. He asked me about my interests, then took me through a list of groups whittling it down to eight cool ones. He told me, “I don’t want you to turn us down because we didn’t interest you. Pick four and in two days you’ll find the position you want.” Even though Microsoft hires more people than the other companies I interviewed, it was the individual attention I received that was far and away more than anything anyone else had done.

As the official “hiring manager” on the ILS headcount, Kaplan had the final decision on making an offer, but his strong bias was to let MacLellan make his own choice. Yet when he received a strong hire recommendation after Matt’s first interview with Reyner, Kaplan felt Matt was showing premature excitement to fill his first headcount. After interviewing Reyner, however, Kaplan said, “Matt, I totally support you. When you find an A team person, you hire them on the spot.”

Knowing they were competing with three other teams internally as well as other outside offers, the two managers began plotting how to win the bidding battle. Responding to Kaplan’s urging to “get creative,” MacLellan came up with the idea of sending the baseball fanatic Reyner a baseball glove signed by the Vega team. Then in a stroke of inspiration, he decided to ask Bill Gates to sign it for the team, inscribing it “Mark—Welcome to the ILS team—Bill Gates.”

Reyner reported that after two days of interviewing he was elated but confused—all four Microsoft teams wanted him and all seemed interesting. He described the process:

About three days later, Ron kept calling urging me to choose. I was impressed by Jim Kaplan who talked to me until about 10 o’clock one night, but I was still trying to find what group fit the best. What really sold me was when I picked up this box at the post office. The fact that someone remembered I was interested in baseball and acted on it, I thought was pretty special.

But some aspects of managing people proved more difficult and even unnatural to MacLellan. He explained one such difficult situation that arose with Gary Soto, a transferee from Texas, who was still trying to settle in to the Microsoft culture, and experienced some “ownership" problems after MacLellan was named lead program manager. At first, the young manager tried to set up very specific concrete things for Soto to do. But when these things did not get done, he found himself getting frustrated and spoke with Jim. “I realized that what I needed to do was to give Gary space and get him passionate about things again,” said MacLellan. “I realized I needed to change my approach.” As MacLellan backed off and gave Soto more freedom to set his own agenda, the young Texan started taking ownership of some parts of the project. Eventually, he became “a really huge asset to the team,” according to MacLellan, acting as a strong customer advocate and championing several product improvements.

Right on schedule, in August 1997 Microsoft made its first shipments of Amelia just in time for the Christmas orders secured earlier in the year at the Toy Fair in New York. Just prior to shipment, MacLellan organized a “ship party” for the whole ILS team. Like the rest of his team, he felt a real sense of satisfaction in creating a product that had almost immediate impact. He remarked:

It was really amazing how hands-off Jim was throughout this whole process. He was giving me feedback, but I was designing and day-to-day running the project. That was really exciting for me. . . . But the real thrill came when the product started to be used. I’d talk to people at Microsoft whose kids market tested Amelia, and they would tell me how their kids and their friends’ kids love the product. Do you remember Furby from Tiger Electronics? It was a hit, but our Amelia is something truly great. It’s such a thrill to know that I have had a part to play in that

300-004 Microsoft's Vega Project: Developing People and Products

development—that it’s had a real impact and that kids, not only girls it was primarily designed for, are using it to learn about technology and math and stuff.

Second Thoughts

After the release, however, some of the project’s excitement dissipated. Over the following six months, the team spent most of its time and energy working on follow-on releases that incorporated incremental improvements on the original product. For MacLellan it was a frustrating time:

My time was spent almost entirely being a manager, dealing with logistical issues, smoothing ruffled feathers over bugs or differences about the next generation, and dotting the I’s and crossing the T’s. When I stopped and reflected, I realized I wasn’t getting a lot of satisfaction from that. . . . It was a bit of a quandary because I was moving forward as a program manager and the natural career track is to continue managing more and more people and teams. . . . I was wondering if I wanted to stay on that track.

During his regular discussions with his boss, MacLellan expressed his re-emerging doubts. Kaplan offered suggestions and advice, encouraging his protégé to think like a manager, not a developer, and to put his energy into growing his team members. When the frustration continued, however, Kaplan suggested that MacLellan take some time off “to get perspective.” On his return from a week-long trip to Hawaii, MacLellan’s mind was made up:

I really wanted to design and build things: snapping the Legos in place to create something interesting. I told Jim that I wanted to become a developer. The problem was that, although I had written software on my own growing up and had studied it in school, I had never done it at a professional level.

Although not really surprised, Kaplan was now confronted with the difficult decision of how to respond. The problem was that MacLellan’s experience made him a Level 33 program manager, whose demonstrated capability and future potential made him a clear candidate for continued bonuses and significant stock options. Indeed, in late 1997 his exceptional 4.5 rating meant that, in addition to his 5% salary increase, he received a $14,000 bonus for his performance in delivering the Vega project, and an allocation of 2,800 stock options reflecting his perceived potential as a program manager. But despite MacLellan’s understanding of the company, its products, and technology, Kaplan felt his limited experience writing code might realistically classify him as a Level 31 developer, where his base salary already exceeded the top of the range. Even if he had confidence enough in MacLellan to stretch him to a Level 32 or 33 as a developer, Kaplan knew that his semi- annual review ratings against such a high standard were unlikely to be good, at least in early years. The bottom line was that Matt’s early performance was unlikely to qualify him for bonuses or stock options—at least not until he climbed the long learning curve of others at a similar level on the technical track.

Kaplan’s head was spinning. As Matt’s mentor, was this a line of thinking he should encourage? Matt had acknowledged he did not want to “throw away six years of a career,” but clearly did not find fulfillment in management. What was right for Matt? What was reasonable from the company’s point of view? How could Matt make a change that required him to essentially “reset the clock”?

Microsoft's Vega Project: Developing People and Products 300-004

Exhibit 1 xxxA Sample Interview Feedback Email with Hire/No Hire

Candidate’s CONFIRMED Interview Schedule

NAME: Drew Thompson POSITION: Software Design Engineer DATE: Tuesday, October 19, 1999

TIME INTERVIEWER

8:30 a.m. Mary Seybold 9:30 a.m. Keith Miller 10:30 a.m. Gary Morgan

11:30 a.m. Gyan Prakasha

1:00 p.m. Leslie Means

2:00 p.m. Ben Allenb

aLunch bAs Appropriate Manager

From: Gary Morgan Sent: Tuesday, October 19, 1999 12:20 PM To: Mary Seybold; Keith Miller; Gyan Prakash;

Leslie Means; Ben Allen Cc: Nancy Ciregna; Mike Snead Subject: RE: 10/19 Drew Thompson CONFIRMED Interview Schedule

HIRE.

First we tried a riddle. He was calm and poised and was able to analyze the problem in a structured way. Responded well to hints.

Then I asked him to code itoa. He spent a few moments thinking, and then described some of the approaches available, including recursion, and the tradeoffs he expected to hit for each. He did very well here.

He then coded up an itoa solution. It took him longer than I expected (35 minutes), but his initial solution was quite good. Although he did think to handle negatives, he had one bug that broke his negative handling. After I pointed out that he had a problem, he fixed that, and had a nearly textbook example of itoa.

As far as coding, I would rate him above entry level but below what I would have expected after two years. I attribute this to the type of work he was doing at his previous employer.

When I talked to him, he sounded highly motivated and very interested in improving his coding abilities. He also understands the big picture well and where he would fit in. I think he would be a good addition at an entry level and would expect him to advance fairly quickly.

Source: Microsoft Corp. (Disguised Data)

300-004 Microsoft's Vega Project: Developing People and Products

Exhibit 2 Recruiting/Rates of Aceptance

1998

Puget Sound 3,319

Worldwide 6,660

Acceptance Rates—U.S.

1996 1997 1998

Campus 61% 68% 70% Non-campus 88% 89% 88%

Source: Microsoft Corp.

Exhibit 3 Employment Headcount: By Geography and Age

Year Worldwide Headcount

USA Headcount

Average Age (years)

1975 3 3 1980 40 40 29.2 1985 1,016 870 30.4 1990 5,635 4331 30.5 1995 17,801 12,193 33.5 1998 28,052 19,487 34.4

Source: Microsoft Corp.

Exhibit 4 xx Microsoft Campus

Microsoft's Vega Project: Developing People and Products 300-004

Exhibit 5 xxA Typical Microsoft Office

Source: Microsoft Corp.

Exhibit 6 xx Attrition Rates

Microsoft Worldwide

1994 1995 1996 1997 1998

9.3% 8.5% 7.5% 7.6% 6.9%

Software and IT Services Industry

1994 1995 1996 1997 1998

N/A 13.7% 16.4% 17.2% 15.3%

Source: Saratoga Institute—Software and IT Services.

300-004 Microsoft's Vega Project: Developing People and Products

Exhibit 7 xxx Sample Survey Questions

Your Job

1. I work toward clear goals. 2. I am appropriately involved in decisions that affect my work. 3. I have the resources I need (e.g., tools access to information, people, etc.) to do my job effectively.

Your Immediate Manager

1. My manager helps me determine priorities for my work. 2. My manager sets high but achievable standards of performance. 3. My manager is good at planning.

Your Professional Development

1. I have significantly enhanced my skills in the past year at Microsoft. 2. I pursue opportunities to take on new, challenging assignments. 3. I know what skills I will need in the future to be a valuable contributor at Microsoft.

Your Work Group

1. My work group works toward clear goals. 2. The atmosphere in my work group helps me to be productive. 3. The people I work with cooperate to get the job done.

Cooperation Between Groups

1. Other groups keep my work group informed about decisions and actions that will affect our group.

2. When my group interacts with other groups, every group has a clear understanding of its unique roles and responsibilities.

3. There is good coordination between my work group and other groups with related goals.

Rewards

1. How do you rate your total compensation (base pay, bonus, stock) on your job? 2. How would you rate you rate the total cash compensation (pay/bonus) on your job? 3. How do you rate the amount of bonus opportunity and/or other incentive pay on your job?

Business Strategy

1. I can see a clear link between my work and Microsoft’s objectives. 2. I have a clear understanding of how my division contributes to Microsoft’s overall success. 3. I believe we are heading in the right direction as a company.

Source: Microsoft Corp.

Microsoft's Vega Project: Developing People and Products 300-004

Exhibit 8 Non-Executive Ladder Levels: Representative Positions on the Managerial and Technical Tracks

Typical Job Titles Non-Executive

Ladder Levels (Range 24-35)

Software Design Engineer (SDE) 29-35 Development Manager 31-34 Software Test Engineer (STE) 31-33 STE Manager 31-33 Program Manager 29-33 Group Program Manager 32-34 Product Manager 31-34 Group Product Manager 31-34

Source: Microsoft Corp. (Disguised Data)

300-004 Microsoft's Vega Project: Developing People and Products

Exhibit 9 Microsoft Performance Review Form: Key Content Sections*

PART 1. PERFORMANCE REVIEW AND PLANNING

A. Evaluate Performance Against Objectives • List each performance objective in priority order • Beneath each performance objective summarize and rate results for this Review period • Discuss specific reasons for the level of performance achieved on each objective, for example:

o Personal factors that helped or hindered performance o Situational factors (e.g., resources, people, events) that helped or hindered performance

• Give constructive suggestions for how performance could be improved

B. Identify Performance Plan for Next Review Period • List 5-7 specific, measurable performance objectives in priority order for the next Review period • Identify keys to success for achieving each objective, for example:

o Resources, tools, or other kinds of support o Training or development needs

• Performance objectives should be mutually agreed upon by employee and manager • If you are a manager, objectives should cover your contribution to your group or organization, as well as your

individual contribution • If you are a senior manager, include steps you are taking to understand and value diversity in your organization

PART 2. COMPETENCY AND CAREER DEVELOPMENT

At Microsoft, each employee is responsible for owning and driving his/her own development. The employee’s manager is responsible for providing appropriate mentoring and guidance. This section of the Performance Review process provides a framework for a useful employee-manager discussion. Ratings are not used in this part of the Review.

A. Identify and Discuss Strengths and Weaknesses • In this section, the employee should briefly evaluate his/her competencies:

o Strengths or personal assets (e.g., attributes, skills, knowledge, experience) that can be leveraged for career development

o Current weaknesses or personal liabilities (e.g., attributes, skills, knowledge, experience) that may limit career development

• The Microsoft Competencies can be very helpful in identifying and articulating strengths and weaknesses.

B. Identify Development Plan for Next Review Period • Identify 1-2 development objectives for the next Review period—strengths to be leveraged, weaknesses to be

addressed • Identify keys to success for achieving each objective, for example:

o Resources, tools, or other kinds of support o Training or personal development needs

• Information provided for each of the Microsoft Competencies can be helpful in developing objectives.

C. Discuss Career Interests and Goals • This section is for discussion only. Written comments are not required.

o In the Review meeting, it is important to have a brief discussion of the employee’s longer-term interests, goals, and concerns. This discussion could cover a variety of issues such as: things that are motivating or de-motivating to the employee about his/her job and working at Microsoft; perceived opportunities for learning, growth, and contribution; jobs or assignments of interest to the employee; support or assistance the manager can provide.

PART 3. GENERAL COMMENTS A. Employee Comments:

• Feel free to comment on work assignment, the Review process, or the company as a whole. B. Reviewer Comments:

• Note any additional comments regarding employee’s accomplishments and/or performance trends.

*Both the reviewer and the employee were required to complete the review form which became the basis of at least two one-on-one feedback sessions. The on-line form was also linked to other resources and help such as Microsoft Success Factors/Competencies, Giving and Receiving Effective Feedback, and Managing Employee Performance.

Microsoft's Vega Project: Developing People and Products 300-004

PART 4. OVERALL RATING AND SIGNATURES

Rating Definition

5.0 Exceptional performance rarely achieved. Marked by precedent-setting results beyond the scope of the position. Demonstrates the highest standards of performance excellence relative to individuals with comparable levels of responsibility.

4.5 Consistently exceeds all position requirements and expectations. Accomplishments are highly valued and may be well beyond the scope of the position. Demonstrates higher standards of performance excellence relative to individuals with comparable levels of responsibility.

4.0 Consistently exceeds most position requirements and expectations. Accomplishments are often noteworthy. Overall performance is consistently above levels of quality and quantity relative to individuals with comparable levels of responsibility.

3.5 Exceeds some position requirements and expectations. Successfully accomplishes all objectives. Overall performance consistently matches levels of quality and quantity relative to individuals with comparable levels of responsibility.

3.0 Meets position requirements and expectations. Accomplishes most or all objectives. Some aspects of overall performance may require additional development or improvement to match levels of quality and quantity relative to individuals with comparable levels of responsibility.

2.5 Falls below performance standards and expectations of the job. Demonstrates one or more performance deficiencies that hinder acceptable performance relative to individuals with comparable levels of responsibility.

1.0-2.0 Does not meet minimum requirements in critical aspects of the job and has numerous performance deficiencies that prevent success at Microsoft.

Employee Overall Rating (employee’s opinion of the overall rating): _______________ Reviewer Overall Rating _______________ Signatures

Source: Microsoft Corp.

Exhibit 10 xxxSalary Ranges and Options Ceilings, by Level.

Job Level 1998 Base Salary Range (Set on 50th Percentile)

Stock Option Application Maximum

29 42-53,000 1,000 30 50-62,000 2,000 31 60-75,000 2,500 32 72-90,000 3,100 33 85-110,000 3,800 34 100-135,000 4,600 35 125-180,000 5,500

Source: Microsoft Corp. (Disguised Data)

Endnote

a Randall E. Stross, “Mr. Gates Builds His Brain Trust,” Fortune, December 8, 1997.