We need a short summary on how was the project done. The summary is below: Google's Project
Question:
We need a short summary on how was the project done. The summary is below:
Google's Project Oxygen: Do Managers Matter? In November 2012, Prasad Setty, Google's vice president, people analytics and compensation, was smiling as he stepped into Union Square in San Francisco. Google executives had organized a celebratory dinner at one of San Francisco's top restaurants to recognize a cross-section of Google employees who had been working on manager-related initiatives over the past several years. Setty's people analytics team was well represented at the dinner, and he was particularly proud of the team's work on Project Oxygen. The initiative had contributed to what Laszlo Bock, senior vice president, people operations, had deemed ''a revolution in people management at Google.'' The Oxygen initiative had started with a fundamental question that Google executives had raised in the early 2000s: ''Do managers matter?'' The topic generated a multiyear research project that ultimately led to a comprehensive program, including surveys, feedback, training, and rewards, designed to help Google employees become better managers. By November 2012, the program, with its focus on eight key management attributes, had been in place for several years, and the company could point to statistically significant improvements across managers overall. Buoyed by the success of Project Oxygen, Setty was already thinking ahead. As he stepped into the Google bus that would take the group back to Google's Mountain View headquarters that evening, his attention turned to his goals for 2013. He wondered how the people analytics group, and the broader people operations group, could build on the success of this project. Now that they knew what made a good manager, what other types of training could they create? Should they focus more on senior managers who might require greater attention to leadership skills? Would it make sense to hire differently? Should they invest more in helping managers get up to speed? These questions flashed through Setty's mind as he settled into his seat. He was tempted to start talking to his team about plans for 2013, but then reminded himself that there would be plenty of time for those discussions back at the office.
Company Background Google was founded in 1998 by Sergey Brin and Larry Page, two computer science doctoral students at Stanford University who developed an algorithm that ranked Internet search results based on which sites had the most links directed at them. Within a year, Page and Brin raised $26 million from investors to turn the idea into a company, and in late 1999, Google moved into new headquarters in Mountain View, California, dubbed the Googleplex. Page and Brin established an informal culture from the very beginning: exercise balls were repurposed as moveable office chairs; desks were made from wooden doors; and dogs were permitted at work. In addition, the founders hired top chefs to provide free meals for all employees. The culture was designed to encourage collegiality and to break down barriers to the rapid development of ideas. Business Overview By late 2000, Google had established partnerships with leading websites such as AOL and Yahoo! and was responding to 100 million search queries each day. That same year, Google introduced a keyword-targeted advertising program called AdWords. Advertisers could buy keywords, and Google displayed the advertisers' text ads when users entered those words in the search box. Google found that its targeted advertising software provided click-through rates that were four times higher than the industry average at the time.1 Google's advertising services enabled the company to start making money, and by late 2001, Google was profitable. Page and Brin hired former Novell CEO, Eric Schmidt, as Google's CEO in 2001, and Page and Brin took the titles of president, products and president, technology, respectively. Google continued to attract paid advertising to its search services, generating $440 million of revenues and $100 million of profits in 2002.2 Headcount reached approximately 700, with nearly all employees based in Mountain View. Google went public in 2004 (ticker: GOOG), raising $2 billion. Revenues grew each subsequent year, and by 2011, Google had $37.9 billion of revenue and $9.7 billion of net income. (See Exhibit 1 for Google financials and headcount statistics.) New products such as mobile advertising and productivity software were becoming increasingly popular, complementing the core business of search and display advertising. After the IPO, Google made several large acquisitions, including YouTube and Motorola Mobility. There was considerable global expansion, and by 2012, approximately half of Google's revenue came from outside the United States. The company's share price on November 30, 2012 was $698.37, resulting in a market capitalization of approximately $250 billion. Organization and Culture At the time of Google's IPO, the company had about 3,000 employees and was run as a ''triumvirate'' with Schmidt, Brin, and Page at the helm. By November 2012, Google's headcount had risen more than tenfold to approximately 35,000 employees; a year earlier Page had replaced Schmidt as CEO. Employees were organized into three primary functional groups: engineering, global business organization (sales), and general and administrative (G&A). Approximately one-half of its employees were based in Mountain View, with the remainder spread throughout the world. Google had an engineering-dominated culture. Eric Flatt, software engineer, commented, ''We are a company built by engineers for engineers.'' Decision-making was described as ''consensus driven,'' and the organization was very flat. Many employees had similar titles, since Google had relatively few managers (5,000), directors (1,000), and vice presidents (100), compared with other companies of similar size. It was not uncommon to find engineering managers with 30 direct reports. Flatt observed, ''Management in the engineering organization is highly constrained, by design. There is only so much you can meddle when you have 30 people on your team, so you have to focus on creating the best environment for engineers to make things happen.'' For much of Google's history, there were questions internally about the overall importance and contribution of managers. Nadav Eiron, engineering director, explained: ''Engineers generally want to spend their time coding and debugging. Many think that talking to direct reports gets in the way of getting that work done. And without training, some engineering managers have a hard time striking a balance between providing direction and micromanaging.'' Jennifer Kurkoski, people analytics manager, commented, ''There are many engineers?not just at Google-----who tend to think that managers are, at best, a necessary evil, and at worst, destructive.'' In fact, Page and Brin once asked Stacy Sullivan, director of people operations, if Google needed managers at all. Sullivan recalled an early experiment: Back in 2002, we had a few hundred engineers reporting to four managers. Seeking a collegial culture similar to what they had experienced in grad school, Larry and Sergey decided to create a totally flat organization?with no managers whatsoever. That experiment lasted two or three months. Too many people went directly to Larry for things like approving expense reports and resolving interpersonal conflicts. Google supported its rapid growth in headcount with a rigorous, data-driven hiring process. The company dedicated substantial resources to ensure that every person hired was top-level talent. Resumes were screened for markers that Google had identified as key success factors for doing well at the company, including extremely high levels of cognitive ability. Candidates who passed the resume screen were then assessed for other attributes such as initiative, flexibility, collaborative spirit, and ''evidence of being well-rounded''-----all components of what made a candidate ''Googley.'' Executives believed that Google's hiring filter was a critical component of the company's success, helping to create a cadre of extremely high achievers. Kurkoski described her group: ''We are hardworking, ambitious people. We are unrelenting perfectionists. We enjoy working together, and we never want to let each other down.'' Google had a culture in which good ideas were celebrated, and authority was derived from peer respect. Chris Loux, head of global enterprise renewals, noted, ''Managers here fail if they rely only on the authority of their position. Google has many young, high achievers who crave autonomy.'' Google had a relatively fluid organizational structure in which groups were created and modified in response to product innovation and market needs. Sebastien Marotte, vice president, enterprise EMEA, commented, ''Everything at Google moves so fast. It is a highly disruptive, unpredictable environment.'' This dynamic led to frequent shifts in reporting relationships. Michelle Donovan, director, people development, noted, ''People move all the time within Google. It is not uncommon to have three different managers over a two-year period.'' Google offered generous compensation packages, including base salary, bonus, stock options, and an extensive set of benefits and perks. In addition to providing free breakfast, lunch, and dinner, the company offered free use of workout facilities. Office hallways were lined with video games and foosball tables, and free courses were offered to all employees. Many employees likened the company environment to a college campus. Comfortable Wi-Fi--enabled buses were provided to help employees commute from the city of San Francisco to Mountain View. Google's compensation packages, perks, and retention levels were considered among the highest in Silicon Valley. Page commented:
My job as a leader is to make sure everybody in the company has great opportunities, and that they feel they're having a meaningful impact and are contributing to the good of society. . . . I don't think it's any of [the individual lifestyle perks]. It's important that the company be a family . . . We should continue to innovate in our relationship with our employees and figure out the best things we can do for them.3 People Operations Page and Brin envisioned a human resources function at Google that would extend well beyond administering benefits and overseeing performance reviews. Laszlo Bock was hired to head up the function in 2006 and named it ''people operations,'' or ''people ops.'' One reporter noted, ''At the heart of [people ops] is a sophisticated employee-data tracking program, and an effort to gain empirical certainty about every aspect of Google's workers' lives?not just the right level of pay and benefits but also such trivial-sounding details as the optimal size and shape of the cafeteria tables and the length of the lunch lines.''4 A substantial portion of the group worked closely with engineering and global business organization teams throughout the company in a role called ''HR business partners.'' People operations helped manage the performance review process, which included regular feedback to managers as well as annual 360-degree reviews. The group also helped manage and interpret the Googlegeist survey, a comprehensive assessment that was completed by over 90% of employees and evaluated how they felt about career development, perks, and benefits, as well as company culture. In 2007, Bock hired Setty from Capital One to head up a group within the people operations called ''people analytics.'' Bock challenged him to ''apply the same discipline and rigor to people operations that we use to manage Google's business operations.'' Setty recalled, ''I didn't want our group to be simply a reporting house. Organizations can get bogged down in all that data. Instead, I wanted us to be hypothesis-driven and help solve company problems and questions with data.'' The group's mission evolved over time. Initially, it was ''all people decisions should be made using data''; later it became ''all people decisions should be informed by data and analytics.'' Setty explained the thinking behind this change: We want to use data to eliminate bias in decision-making, but we don't want data to completely erase the role of personal judgment. A few years ago, we ran an experiment in which we tried to use data to determine which engineers should be promoted. Managers weren't comfortable with that approach. They wanted data to help them make better decisions, but they didn't want a ''black box'' that used data to arrive at answers without human input. Setty built out the people analytics team with PhDs who brought rigorous research methodologies to the company. Three members of the people analytics group?Kurkoski, Brian Welle, and Neal Patel?formed a small team called the People & Innovation Lab (PiLab). Led by Kurkoski, their goal was to tackle questions related to the well-being and productivity of Google employees. In early 2009, the PiLab shared its initial set of research questions with Setty, including ''How can we encourage employees' savings behavior?'' and ''How can we improve the onboarding process for new hires?'' Setty noted, ''I remember meeting with the PiLab team in early 2009 to review all of their proposed questions. While there were a few we wanted to pursue, one stood out from the rest. The team asked, 'Do managers matter?''' Patel, former people analytics manager, and as of 2013 a technical project lead with Motorola's advanced technology and products group, elaborated: ''We hoped the research would help us determine whether managers impacted the performance of their teams. Like a lot of our work, it was an aspirational endeavor. We asked ourselves, 'What if everyone had an awesome manager?not just a good one or a competent one, but a truly amazing manager? What kind of results would we see?''' Project Oxygen The question of whether managers matter formed the basis of a broad research project undertaken by the people analytics group in late 2009. Under the direction of the PiLab and with a core team of Patel, Donovan, and Kurkoski, the project was code named Project Oxygen. Donovan and Patel were Project Oxygen's co-founders and co-leads. Naming internal projects was a time-honored engineering tradition at Google that was soon adopted by other groups, as Patel explained: I got the idea from a colleague who named one of his big projects after the X-Men. Soon, I began putting ''Project'' in front of the name, both as a reference to the X-Men and to the secret plans in the 1960's TV show Danger Man. Then I started putting an element in the periodic table after ''Project.'' So every project I worked on became Project Krypton or Project Argon or Project Palladium. As we began wrapping up the research phase of ''the manager project,'' Michelle thought we should give it a cooler name. She suggested the element oxygen, because having a good manager is essential, like breathing. And, if we made managers better, it would be like a breath of fresh air. Patel also recalled, ''We knew the team had to be careful. Google has high standards of proof, even for what, at other places, might be considered obvious truths. Simple correlations weren't going to be enough. So, we actually ended up trying to prove the opposite case; that managers don't matter. Luckily, we failed.'' Research The Project Oxygen team started by reviewing the data that had already been collected when employees left Google to see whether management issues were cited as one of the reasons for leaving. They found some connections between low satisfaction with one's manager and turnover rates, but given the low overall turnover at the company, they did not feel the data was robust enough to extrapolate to the general population at Google. Moreover, a correlation between satisfaction with one's manager and retention wasn't enough to prove that managers were the cause for attrition. Patel addressed this by examining a cross-section of high-scoring and low-scoring people managers based on a combination of Googlegeist ratings and performance review scores. ''High-scoring'' managers were those in the top quartile (top 25%) on both measures, and ''low-scoring'' managers were in the bottom quartile of both. Patel explained: At first, the numbers were not encouraging. Even the low-scoring managers were doing pretty well. How could we find evidence that better management mattered when all managers seemed so similar? There was really no reason to expect any differences in team performance or team member satisfaction between the two, ostensibly similar, groups. But since we didn't expect to find any differences, even small differences were impressive. It turned out that the smallest incremental increases in manager quality were quite powerful. Good managers do matter.
The entire team became excited when the additional research found evidence suggesting that significant differences between high-scoring and low-scoring managers did, in fact, have a significant impact on job satisfaction, retention, and performance. For example, data showed that Googlers on teams with higher-scoring managers were happier. In addition, higher-scoring managers in 2008 had lower turnover rates in their teams than low-scoring managers. In fact, Patel's research indicated that retention in 2008 had a stronger link to manager quality than to seniority, performance, tenure, or promotions.5 Furthermore, the data indicated that better managers had higher-performing teams. 6 Overall, the research team found that Google employees with higher-scoring managers had consistently higher scores on all Googlegeist dimensions, including innovation, work-life balance, and career development. (See Exhibit 2 for details.) Based on this research, the PiLab team concluded that managers do matter, but that the finding was actionable only if they could answer the next question: ''What do our best managers do?'' During the summer of 2009, the PiLab team and a few interns set out to conduct double-blind qualitative interviews with managers from both the highest- and lowest-scoring categories. All interviewers were given a script to follow with questions like ''How often do you have career development discussions with your direct reports?'' and ''What do you do to develop a vision for your team?'' Data was gathered from managers in all three major functions of the company (engineering, global business organization, and general & administrative), and also from managers at multiple levels and from diverse geographic locations. The PiLab team also reviewed the Great Manager Award nominations as well as thousands of qualitative comments collected through the Googlegeist survey and performance reviews. (The Great Manager Award was an annual program where Google employees nominated approximately 20 managers from across the company as outstanding managers.) The PiLab team spent the next several months coding the information from these sources. After much review, they identified a set of eight behaviors that were common among high-scoring managers throughout the company. The attributes, which were presented in order of the number of times they were mentioned in Googlegeist comments to give some idea of their relative importance, included being a good coach, expressing interest in members' success and personal well-being, helping team members with career development, and being a good communicator.7 Kurkoski noted, ''When we pulled together the list of eight behaviors, we were actually concerned because they looked so familiar. We wondered if we would get traction with a list that might seem obvious.'' Donovan added, ''We hoped that the list would resonate because it was based on Google data. The attributes were about us, by us, and for us.'' (See Exhibit 3 for a full list of the eight behaviors and a quotation illustrating each behavior.) Action Plan Upon completing the two lists, people operations employees developed an action plan for 2010 to develop and support managers in a more comprehensive manner. Kurkoski explained: ''We all believed that our study was not just about collecting data, it was about driving change. What we did with the data was just as important as gathering it. We knew that we needed to make the Project Oxygen findings relevant for all parts of the company.'' The Project Oxygen team worked with other parts of people operations to develop a broader people management initiative that included plans to increase company-wide awareness of effective manager behaviors, and create curriculum, tools, and programs to improve manager quality.
A key part of the Project Oxygen team's plan was to modify the annual upward feedback survey that was already being used at the company, and to then collect responses to the revised survey a second time each year. The updated survey would focus specifically on the eight attributes of great managers, and each manager would receive personalized results that represented an average of the scores from his or her direct reports. The report would also include anonymous comments that were meant to be helpful. Setty recalled, ''While redesigning the upward feedback survey, we had lots of conversations about the intent. Should we link the findings to performance reviews? We made a conscious decision to keep it separate, particularly because this was the first time we were telling managers about the importance of these eight attributes.'' Mary Kate Stimmler, people analyst, added, ''Project Oxygen was always meant to be a developmental tool, not a performance metric. We realized that anonymous surveys are not always fair, and there is often context behind low scores.'' Although the Project Oxygen team was excited about the research and implementation plan, they knew it would be critical to get ''buy-in'' from managers across the company. Donovan explained, ''We have a consensus-oriented culture. Commanding employees to do something doesn't work.'' The Project Oxygen team planned to ''socialize'' its findings through a combination of grass-roots meetings with junior and mid-level managers and solicit sponsorship from Bock and other senior executives. The team brought in Judith Hoban, who worked in the people development space, to help socialize the findings with the company and think about how the findings should be taught. Kurkoski noted, ''Although it was important to have executive support, it was even more critical that we talk to people in the trenches. I think Michelle and Neal must have done over 50 presentations around the company where they really got into the details, including the research methodology. They made the data relevant for each audience.'' The presentations not only included the list of attributes, but also described best practices to help managers improve. (See Exhibit 4 for examples of best practices associated with two attributes.) Donovan reached out to managers throughout Google, seeking invitations to staff meetings and other gatherings where she could share Project Oxygen findings and the proposed action plan. Setty explained, ''In the engineering organization, we started with what we call the 'tech advisors'-----the engineers who have been at Google a long time and who are vocal. We wanted them to share the findings, because we knew that tech managers would be more likely to listen to their peers than to folks from people operations. Early feedback was positive. People latched on because it was Google data.'' Flatt commented, ''The list substantiated what I already knew anecdotally about management, but it was nice to have the eight attributes listed in one place.'' Eiron, one of the tech advisors, started a small grass-roots group (called a ''grouplet'') within the engineering organization to focus on management development; he also encouraged the adoption of Project Oxygen within his own group. Project Oxygen Implementation Surveys To measure the eight attributes, the Project Oxygen team partnered with people operations to create two variations of the survey. The Upward Feedback Survey (UFS) was developed for employees in global business organization and G&A, while the Tech Managers Survey (TMS) was developed for members of the engineering group. Both surveys had 16 statements, spread across the attributes. Statements included, ''My manager regularly gives me positive feedback'' and ''My manager does not micromanage.'' Respondents were asked to answer with ''strongly agree,'' ''agree,'' ''neutral,'' ''disagree,'' or ''strongly disagree.'' (See Exhibit 5 for the UFS statements.)
The first upward feedback survey with Oxygen attributes was sent out to employees in June 2010. The timing was specifically intended to be different from performance reviews, which took place in April and September. (See Exhibit 6 for the 2012 calendar showing the timing of Google's surveys and evaluations.) The people operations team communicated frequently about the survey as a developmental tool. Confidentiality was stressed, and managers did not get reports if they had fewer than three employees. The survey was sent out again twice in 2011. In the time between the two surveys, the New York Times profiled Project Oxygen. Donovan noted, ''The article drew internal attention to the project, which helped us gain even more momentum.'' The surveys were not mandatory, but the vast majority of employees completed them regularly. Feedback Reports Managers received their results several weeks later in an online report that gave numerical scores for each question in the form of ''% favorable response'' for each question as well as specific comments. Managers were encouraged to share their reports with members of their teams and to talk openly about areas for improvement. The initial reports referenced the eight attributes (dubbed the ''Oxygen 8''), and included links to learn additional information about each attribute, including what success looked like. (See Exhibit 7 for a sample results page and Exhibit 8 for Oxygen tips that were shared with managers.) Reactions to the first set of reports were generally positive, although managers were sometimes surprised by particular findings. Stephanie Davis, director of large company sales based in San Francisco, reflected on the first time she received her UFS results: I was not surprised by the list of Oxygen attributes, even though I had never methodically thought about them. I had received positive feedback as a manager before, but my first report pointed out things of which I was not aware. For example, I was surprised that one person on my team didn't think I had regularly scheduled one-on-one meetings. I saw this person every day, but the survey helped me realize that just seeing this person was different than having regularly scheduled individual meetings. My team also wanted me to spend more time sharing my vision. Personally, I have always been inspired by Eric [Schmidt], Larry, and Sergey; I thought my team was also getting a sense of the company's vision from them. But this survey gave my team the opportunity to explain that they wanted me to interpret the higher-level vision for them. So I started listening to the company's earnings call with a different ear. I didn't just come back to my team with what was said; I also shared what it meant for them. Training By the end of 2010, a group within people operations started to build management training courses around the Oxygen findings. A few months later, when managers received their survey reports, they were given specific suggestions about actions they could take to improve. For example, a manager who received unfavorable scores in coaching would get a recommendation to take a particular coaching class and could sign up directly from a link in his or her online report. In addition, the development team arranged panel discussions with high-scoring Google managers from each functional group. People analyst Stimmler explained, ''We realized that engineers don't necessarily want to hear about management from people in HR. But they are willing to listen to engineering managers whom they respect.'
Classes were designed to be interactive and immediately actionable. Coaching classes offered detailed recommendations on how to deliver personalized, balanced feedback. They also provided opportunities for employees to role-play and practice giving and receiving feedback. ''Vision'' classes used a series of images to encourage creativity. Participants were encouraged to bring ideas to life by telling a compelling story and practicing writing their own vision statements. They also did an exercise in which they had to write down answers to questions such as ''What is the unique value your team adds to its users?'' and ''What are your team's core strengths?'' Kathrin O'Sullivan, manager, people development, explained, ''Google has turned a corner and become passionate about management development. We have a team of instructors, and we are piloting online 'Google Hangout' classes so managers can participate from around the world.'' Eric Clayberg, a Google software engineering manager, noted: I am a serial entrepreneur. I had been managing teams for 18 years before my latest start-up was bought by Google. Yet when I started taking some of the classes offered by people operations, I learned more about managing in six months than I had learned in the previous two decades. For instance, I was worried about the flat organizational structure at Google; I knew it would be hard to help people on my team get promoted. I learned in the classes about how to provide career development beyond promotions. I now spend a third to half my time looking for ways to help my team members grow. Engineers hate being micromanaged on the technical side, but they love being closely managed on the career side. In 2011, the people operations group built and launched the ''Start Right'' class, a two-hour introductory course for new people managers that was built around the Oxygen attributes. In addition, people operations promoted its Manager Flagship series, which included ''greatest hits'' from each of the key topics. The series covered areas such as giving and receiving feedback, building a vision for your team, and managing change in three two-day modules over six months. People operations also started to send new managers automated e-mail reminders with recommendations on what new managers should do to be successful at Google. The recommendations included links to Oxygen findings and information about additional courses they had not taken. Great Manager Award In 2009, Google instituted the Great Manager Award. Managers had to be nominated by Google employees, and approximately 20 recipients were selected each year. The prize included a week-long trip to a destination like Hawaii, where winners had the opportunity to spend time with senior executives and other recipients of the award. When the award was first created, selection criteria were based on general feedback from Google employees about their managers. In 2010, the people operations group modified the criteria so they reflected the eight Oxygen attributes, and employees were encouraged to refer to the attributes when submitting nominations. The managers who were selected served as role models for the company, sharing best practices through a variety of channels. Clayberg was one of the recipients of the Google Manager Award. He believed that the career development skills he acquired through his Project Oxygen training was a key reason he was nominated. He noted, ''It was amazing to spend a week in Hawaii with cool, fantastic people. Since then, I've been spending time coaching other managers. Recipients of the awards are expected to give back.''
Project Oxygen Impact People operations analyzed the impact of Project Oxygen by examining aggregate data from the surveys as well as input from individuals. A comparison of UFS and TMS scores from 2010 through 2012 indicated that median scores rose from 83% favorable to 88% favorable. Welle, people analytics manager, noted, ''These improvements hold true across the three main functions within Google, across survey categories, and across levels of managers. We've seen the least effective managers improve the most over time.'' Setty noted that the HR business partners had been able to work closely with the lowest-scoring managers, helping them develop new skills or transition into nonmanagement roles. Career development and coaching were two of the areas that showed the most improvement. The improvement findings were also consistent across geographies, and with varying numbers of direct reports. In an environment of top achievers, low UFS scores tended to inspire action. For example, Marotte (vice president of enterprise EMEA) had come to Google in 2011 from a senior sales role at Oracle. Based in Europe, he managed a global team of 150 people. During his first six months at Google, Marotte focused on achieving his sales numbers, which he met consistently. Then he received his first UFS scores which ''were a shock to my system.'' Marotte recalled: My first UFS scores were a disaster; I asked myself, ''Am I right for this company? Should I go back to Oracle?'' There seemed to be a disconnect, because my manager had rated me favorably in my first performance review, yet my UFS scores were terrible. At Oracle, all that mattered was hitting my numbers. My first reaction was that I had the wrong team. I thought that they didn't understand what we needed to do to win. But then I took a step back and met with my HR business partner. We went through all the comments and came up with a plan. I fixed how I communicated with my team and provided more visibility on our long-term strategy. Within two survey cycles I raised my favorable ratings from 46% to 86%. It's been tough, but very rewarding. I came here as a senior sales guy, but now I feel like a general manager. Other Google managers were equally surprised after receiving their first set of scores. Loux (head of global enterprise renewals) recalled receiving his initial scores and feeling frustrated because his group was in transition. He had recently inherited a number of employees, and a few had moved in and out of his group during the previous six months. Nevertheless, he found the low UFS scores he received in several categories to be ''very impactful.'' Loux recalled, ''I had received a performance review indicating that I was exceeding expectations, yet one of my direct reports said on the UFS that he would not recommend me as a manager. That struck me, because people don't quit companies, they quit managers.'' At the same time, Loux struggled with the question of just how much to push the low performers on his team. He explained, ''It's hard to give negative feedback to a type-A person who has never received bad feedback in his or her life. If someone gets 95% favorable on UFS, I wonder if that manager is avoiding problems by not having tough conversations with their reports on how they can get better. I think 75%--80% favorable might be the sweet spot.'' The question about the relationship between UFS scores and performance reviews was raised multiple times during the roll-out period for Project Oxygen. At one point, the people analytics group did a formal analysis to find out whether there was any truth to the perception that if a manager gave a direct report a low rating, he or she would receive a low UFS score from that employee. They found after looking at two quarters' worth of upward feedback data from Q2 and Q3 of 2011, that changes in performance ratings (both upward and downward) accounted for very little variability (less than 1%) in corresponding manager ratings across all functions at Google. At the start of Project Oxygen, people operations had considered creating a formal mechanism to link UFS scores with performance reviews, but then decided that doing so was likely to cause Project Oxygen to lose its impact. Bock explained, ''People who exhibit strong management behavior seem to be getting promoted at a higher rate than other managers. For example, in the last round of promotions to vice president, 10% of the directors promoted were winners of the Great Manager Award.'' Reflections The people operations group believed that nearly all managers saw value in Project Oxygen; however, a few had questions about aspects of the initiative. One question that arose was whether management was becoming formulaic: Were some managers using the Oxygen 8 as a checklist, engaging in recommended behaviors without a genuine interest in improving their management skills? Stephanie Davis noted, ''I have heard people ask whether our scores have gotten better because we now manage to the test. They wonder if management has become less authentic over time.'' Bock reflected on this issue, ''We are not trying to change the nature of people who work at Google. That would be presumptuous and dangerous. Instead, we are saying, 'Here are a few things that will lead you to be perceived as a better manager.' Our managers may not completely believe in the suggestions, but after they act on them and get better UFS scores, they may eventually internalize the behavior.'' A second question related to how Googlers felt about being told, with a high degree of granularity, how they should behave. Welle noted, ''In many ways, Google is an individualistic culture. We let engineers choose what language they want to code in, and employees are given a great deal of latitude about how they manage their jobs. In that context, Project Oxygen might be seen as very prescriptive, which is a big change.'' At the same time, Google managers seemed to appreciate the importance of being a manager and they took the role seriously. Eric Flatt explained, ''Managers here realize that they are responsible for other people's careers.'' A third question related to the sustainability of Project Oxygen. Google employees were asked to fill out evaluation surveys multiple times a year. Marotte commented, ''It was a shock to my system to be evaluated so frequently and with such intensity.'' Some employees wondered if it was possible that they would reach a point of evaluation overload. Others were more sanguine. Nadav Eiron noted, ''I find it useful to get performance feedback. People who work at Google today realize the importance of management and want to invest in it. The composition of the company has changed.'' Looking Ahead As Setty rode the bus down Highway 101 toward Google's campus after the celebratory dinner, he reflected on the company's distinctive culture and values: We have a grueling recruiting experience and make people jump through all kinds of hoops before they get here. But once people join the company, there's this feeling at Google we've created a family. I don't know how to quantify it, but it's a value that is important to us. Treating people like family means nurturing them, developing them, and making sure they reach their full potential. Those that can do that best are our people managers. They don't just direct work, but actually help others improve. Setty was considering a number of possible new projects for the people analytics team. One was to expand Project Oxygen to develop a deeper understanding of the specialized attributes of the most senior managers of the company such as directors and vice presidents. Davis commented, ''Now that I'm a director, I need to think differently about my role. I spend time thinking about how to work with other stakeholders around the world. The management attributes are the same as other management jobs, but the delivery process is different.'' Bock added, ''Senior leadership often has to inspire. We are looking at the issue of how inspiration can be taught.'' The people analytics team was discussing whether a broader study of those management and leadership attributes specific to senior managers could be done with the same approach and research methods that were used for Project Oxygen. Setty also believed it would be helpful to learn more about the complete lifecycle of managers at Google, including hiring, lateral job transitions, promotions, and departures. Could Google, for example, do a better job of finding and hiring great managers from outside the company? How might Google identify the best management candidates among its current employees? His team was planning to analyze the job transitions for weak managers, while also reviewing the company's ability to reward and retain its top managers. The goal was to create a virtuous, reinforcing cycle. Donovan explained, ''When we hire, we want to ask potential managers about the Oxygen behaviors. Once they become a manager, we want to onboard them by training them on these behaviors. After they've managed for a few months, we want to raise their self-awareness by giving them a UFS or TMS report so they can see their team's assessment. Then, we should make it easy for them to attend manager courses targeted at their lower-scoring areas. This is the cycle we envision?we're part of the way there.'' In addition, Setty was interested in the possibility of applying people analytics to the study of teams. He elaborated: We have great teams at Google, but are wondering if we can take the same approach we used in Project Oxygen to make teams even more productive by providing greater guidance on their composition. Right now, teams are staffed very organically; we tend to look primarily at skills and availability. Should we be more prescriptive about diversity of representatives, variety of experience, and mix? How do we find the right combination of people who have worked well together in the past versus adding new blood? And then, once teams are established, what kinds of processes can we lay out to make them more successful? With Project Oxygen, we were looking at managers?people who had others reporting to them. Teams, on the other hand, are composed of many talented people?project managers, tech leads, and so on?who do not necessarily manage others. We have hundreds of people in these roles, and the focus of this work would be not just on the individual contributor but on the entire composition of the team. Finally, Setty believed it was important to continue to pursue further incremental change and improvement on the management front. He explained: When we started Project Oxygen, the data suggested we had a wide variance of managers and management skills. We've done a very nice job moving the average up quite significantly, particularly for the bottom quartile. We continue to work on reducing the variance across the population and trying to raise a very high mean.
Soon, however, we will have to start thinking about what else there is that drives people to go from good to great. We are starting to do some work looking at managers by personality traits to see which ones had the greatest improvements in their scores. With Project Oxygen, we didn't have these endogenous variables available to us; they weren't part of our data set. Now, we can start to tease them out, using more of an ethnographic approach. It's really about observations?staying with people and studying their interactions. We're not going to have the capacity to follow tons of people, but what we'll lose in terms of numbers we'll gain in terms of a deeper understanding of what managers and their teams experience. As Setty stepped off the Google bus, he was reminded of a comment that he had heard from Larry Page. Page had said that Google should be ''growing the leaders that the world needs.'' Setty thought about this insight, and it pushed him to an even more expansive agenda. He observed, ''My biggest fear is that Project Oxygen is constraining. Are we boxing people in? The Project Oxygen attributes are extremely important, but they are not aspirational. How do we create truly amazing managers?''
Expert Answer:
