An Interview with Ross Smith by the American Journal of Play
Reprinted with permission from the American Journal of Play (Volume 5, Number 1, Fall 2012)
Over the last two decades, Ross Smith—Director of Test for Microsoft Corporation—
developed software for mainframe systems, PCs, and hand-held devices. As
a long-time member of Microsoft’s Test Architect’s Group, he helped create nearly
every version of the company’s Windows and Office products that appeared after
1995. A recipient of the Harvard Business Review/McKinsey M-Prize for Management
Innovation, Smith holds five patents and is coauthor of The Practical Guide
to Defect Prevention. He also devises games and social-networking tools to train
managers, and he writes a weekly newsletter for Microsoft employees worldwide.
The Management Lab (MLab) at the London School of Business has published a
case study of his “42Projects,” a corporate initiative involving productivity games.
In this interview, Smith describes how play and productivity games have altered the
relationship of workers to work and of work to management at Microsoft and the
dividends that these approaches can yield in trust, productivity, and satisfaction.
American Journal of Play: Mr. Smith, what in your background and experience
led you to Microsoft?
Ross Smith: I studied decision making and computer science at Rider University,
and after graduating, I worked in county government for a while and then for
a hand-held computer distributor. I joined Microsoft in 1991 from a small
hand-held computer start-up. I began at Microsoft in Charlotte, North Carolina,
working on a product called Works for DOS. About six months later,
after the release of Windows 3.1, I moved to a testing and quality-assurance
position for linking and embedding objects in Windows NT. I moved to the
Office-software team just before the release of Office 95. I remained there
until 2003 then went back to Windows for Vista and Windows 7. I am now
back in Office working on a product called Microsoft Office Lync, which is an audio-video chat and conferencing set of applications.
AJP: What does testing entail?
Smith: Our job is to assess and improve the quality of our products, including
testing across multiple devices and platforms—Windows, Mac, iPad,
iPhone, Windows Phone, and Android. Doing that, and ensuring that a
product will deliver a high-quality experience across such a diverse set
of end points, requires a talented and engaged test team and a sustained,
broad, and diverse set of testing techniques. My role as a Director of Test
is to help with the overall strategy to ensure quality, to recruit engineers
for our team, and to help keep the team engaged.
AJP: What is the 42Projects initiative and how did it come about?
Smith: In 2004 we had a small, research-like team called the Defect Prevention
Team charged with improving software quality. This is where much of our
thinking around the use of games started to take shape and from that we
eventually wrote The Practical Guide to Defect Prevention. At the start of
Windows 7, I became the Director of Test for the Windows Security team
and was responsible for roughly eighty-five people charged with assessing
the quality of the software.
As I began meeting with members of the team, it became obvious that
the level of talent we were hiring in testing had increased significantly,
and the type of skills that people brought to the workplace differed from
when I started. We wanted to explore how we might change our processes
to better accommodate these new skills and how we might apply them to
security testing. We explored ideas like 15 or 20 percent time, out-of-thebox
week, sprint runs, and other creative attempts to give people freedom
to experiment. Fifteen or 20 percent time is a practice in which several companies,
most notably 3M and Google, allow time each week for employees
to choose their projects. Out-of-the-box week and sprint runs are similar,
as they declare a period of time free of assigned work and allow employees
freedom to choose what they want to pursue. One of the challenges we
had in our project was that the ebb and flow of the project cycle impacted
every individual differently, so it was very difficult to find an initiative that
worked for everyone.
As we were doing this, we ran into a paper on trust from a couple of
researchers at the University of British Columbia—John F. Helliwell and
Haifang Huang. Their paper, “Well-being and Trust in the Workplace,”
equated the level of job satisfaction to the level of organizational trust,
suggesting that a 10 percent increase in trust felt the same as a 36 percent
pay raise as measured by job satisfaction. It dawned on us that the culture
we were trying to create and the behaviors we were trying to reinforce
were all rooted in trust. Freedom to experiment, empowerment, freedom
to fail, opportunity to suggest new ideas, collaboration, ability to suggest
new and different ways of doing things—all require a strong foundation
of organizational trust. We all agreed that a high-trust organization would
bring value, but we had no idea how to create one.
AJP: Where did you start?
Smith: We decided simply to ask the team. We got people together in a conference
room with yellow sticky notes and asked them to write down behaviors
they felt would influence trust—positively or negatively. Once we had the
list, we wanted to order and prioritize it, and we developed a simple pairwise
voting game that would allow players to choose between two alternatives
and help us stack rank the list. We realized quickly that prioritizing
the list wasn’t really that helpful, however. Because trust is situational and
relationship based, we then moved the behaviors into a wiki-based playbook
that could offer a reference for team members as we aspired to modify our
collective behavior.
AJP: The wiki format enabled all to contribute?
Smith: Yes. The bulk of the effort came from a few individuals, but everyone
on the team had the opportunity to contribute examples and findings.
We also realized that while trust could have a significant influence on our
team capabilities and that what we call productivity games could have a
significant impact on our work, we would need to continually innovate
in the way we manage to better accommodate the skills of the incoming
workforce. So we decided to wrap these three pillars—productivity games,
trust, and management innovation—into a single, quirky initiative we
called “42Projects.”
AJP: Why the name 42Projects?
Smith: The number forty-two references the magic number in Douglas Adams’s
The Hitchhiker’s Guide to the Galaxy novels. His hugely powerful Deep
Thought computer posited forty-two as the ultimate answer to the ultimate
question of “life, the universe, and everything.” Finding out the ultimate
question, though, proved to be the tricky part—and we are still looking
for ultimate questions in our design of productivity games, our quest for
trust, and our continuing interest in management innovation. In addition,
forty-two is represented as 10101010 in binary, is Jackie Robinson’s
number in Major League Baseball, and is the angle at which a viewer of
light deflected from water will see a rainbow.
AJP: Did your own early play experiences affect your present work at Microsoft?
Smith: Absolutely. I grew up an avid gamer, not just video games, but also board
games like Risk and Monopoly. I have a particular affinity for—perhaps
addiction to—Robotron 2084, the 1980s game to save the last human family
from robots. I think the attraction is that the game looks impossible
at first glance, but after a significant investment in tuition—one quarter
at a time—player dexterity improves, and the game becomes playable. I
also spent a lot of time around pinball machines as a teenager, and I enjoy
cartoons and cartooning. However, I didn’t really make the connection
between game play and work until after we started seeing the success of
games in the workplace on the Windows Defect Prevention team. I was
not trained as a game designer or a game programmer, but as I started to
read more on good game design and motivational psychology, it seemed
obvious to me that using games in the workplace was a natural extension
of the human affinity for play, games, and gaming.
AJP: Are workplace productivity-boosting games a new idea?
Smith: Using games to get work done is not a new idea. Mary Poppins is famous
for singing, “In every job that must be done, there is an element of fun. You
find the fun, and snap! The job’s a game!” Also, there are plenty of stories
of back-room competition at used-car dealers, where leader boards track
sales. A leader board is perhaps the most common game mechanic, and,
of course, they are used in golf as well. In colonial times, playing games
during harvest time was commonplace. My favorite story, though, is about
the work gangs that built the great pyramids.
AJP: How did games figure into pyramid work gangs?
Smith: According to what I’ve read, ancient Egyptian pyramid builders organized
and subdivided workers into teams or gangs, some of which identified
themselves with names such as Friends of Khufu, Drunkards (or celebrants)
of Menkaure, Endurance, Perfection, and so on. Apparently, these teams
competed with one another to carve and haul the massive stones to build
the pyramids. In certain monuments, you can find the name of one gang
on one side of the pyramid and the name of another gang—I assume the
competing gang—on the other side. It’s as though these gangs were pitted
against each other. According to some Egypotologists, the gangs, or phyles
competed for beer and bread to see who could do the job faster. Carving
their team names in to the stones helped them keep score.
AJP: How do you see that ancient example applying to today’s workers?
Smith: While the twenty-first-century worker has a lot more freedom, most
of us still don’t go to work because we want to. We show up each day and
do our best work in order to earn a paycheck. In a very abstract sense,
most of us share the condition that the motivation to perform work is
extrinsic. And that’s where games and play apply in both societies. Games
and play offer alternative or additional extrinsic rewards to make things
more fun—whether you’re being directed to haul a huge stone up a ramp
to build a pyramid or to haul a huge budget request upstairs to an angry
finance director. The respective workers might not choose to do either
task, but sprinkling in gaming elements and rewards can make the job
more satisfying.
AJP: Where else did you find inspiration for productivity games in particular?
Smith: In several publications. The 1977 classic novel Ender’s Game, by Orson
Scott Card—where Ender Wiggins plays a series of games then ascends
to Battle School to play more games that, unbeknownst to Ender, prepare
him for fighting for real in the Third Invasion—is a great illustration of
the potential use of games.
Byron Reeves and J. Leighton Read, in their book Total Engagement:
Using Games and Virtual Worlds to Change the Way People Work and Businesses
Compete, suggest that “successful businesses in the future will redesign
work from the gamer’s point of view. Businesses will create a workplace
that accommodates employees (‘players’) who want to know the rules,
advance frequently, partner quickly, and nurture reputations in a narrative
that aligns their own objectives with those of the organization that pays
their salary” (p. 8).
At IBM, Li-Te Chang, Phaedra Boindois, and Osamuyimen T. Stewart
have done great work on the use of games at work. And the 2008 book,
Changing the Game: How Video Games Are Transforming the Future of Business,
by David Edery and Ethan Mollick, shows how such ideas have become
commonplace.
Our productivity games at Microsoft build on these earlier works. We
have the advantage of working in software—a malleable medium—and
this gives us a lot of flexibility in allowing the work to become the game,
as it did for Ender Wiggins.
AJP: Can you draw some parallels between elements of good game design and
elements of motivational psychology?
Smith: Yes. I see similarities in structured goal setting, rewards, and positive reinforcement. A review of Abraham Maslow’s Hierarchy of Needs, for example,
illustrates how game mechanics can support the top three levels—Love/
Belonging (team-based play), Esteem (leader board), and Self-Actualization
(avatars, level achievements, helping others). Motivational concepts include
reward and reinforcement, as well as negative reinforcement, punishment,
and coercion. Simple game mechanics, such as a leader board, provide a
structured alternate reality playground where these needs can be addressed.
Just as easily as games can attract effort by using rewards, they can also use
the threat or the shame of not being on the leader board as a way to coerce
players to participate.
There are also differences. Games tune the environment and employ
design elements like accidental success to reengage or keep players moving
towards a goal. Games educate the player and offer an alternative point of
view more effectively than real life. And games offer structured multiplechoice
options, the chance to undo, and practice modes with real-time
assessment that in certain situations even help address Maslow’s safety
level in real life.
AJP: How can games build trust and inspire innovation?
Smith: The ability to innovate is a key component of successful companies;
innovation requires experimentation and risk taking; and creating a culture
of risk taking is difficult. It is insufficient to encourage or command
employees to take risks. Organizational culture must support employees
who experiment. Many organizations claim they want employees to take
risks, but performance-evaluation systems reward only success—or even
worse—penalize and punish employees who experiment, fail, and learn.
Risk taking and other behaviors that support innovation—freedom to fail,
willingness to collaborate, and experimentation—all require significant
organizational trust.
AJP: So game play is key to building trust, and trust is key to inspiring innovation?
Smith: Yes. In his classic work Homo Ludens: A Study of the Play-Element in
Culture, Johan Huizinga calls play “a free activity standing quite consciously
outside ‘ordinary’ life as being ‘not serious’, but at the same time absorbing
the player intensively and utterly” (p. 38). Using productivity games and
play in the workplace is a successful technique to build organizational trust
and, by extension, create a culture of innovation. Game play can provide
structure and rules to support experimentation, risk taking, and failure.
It’s hard to fail at work, but it’s culturally acceptable to lose the game. In
Man, Play, and Games, Roger Caillois says that “play is essentially a separate
occupation, carefully isolated from the rest of life, and generally is engaged
within precise limits of time and place. There is a place for play: as needs
dictate, the space for hopscotch, the board for checkers or chess, the stadium,
the racetrack, the list, the ring, the stage, the arena, etc. Nothing that
takes place outside this ideal frontier is relevant” (p. 6).
The “place for play”—or magic circle as some people call it—for members
of an organization is a place to learn trust-building behaviors. Experimenting
with new ways of working is acceptable within a game—in a place
for play. If things don’t go well, the game is, as Huizinga suggests, outside
ordinary life and not serious. Games offer a framework to support risk
taking and experimentation, and in a game, someone can learn new skills
by trying them out on their own. If they fail, well . . . OK, they lost the
game, but there was no long-term impact, no risk to their careers. They
just played a game. If I’m a computer programmer, I probably won’t ask
my manager for permission to take a class in marketing, even if that’s my
passion. I might spend hours in my spare time making marketing videos
and showing them to no one, or posting them anonymously on YouTube.
However, if you put together a game for the best ad campaign submission,
I might be the top contributor.
AJP: What are the costs of a low-trust work environment?
Smith: I’ll answer in terms of our team’s work on software testing. In software
testing, it gets progressively harder and harder to find defects. If we run the
same test or use the same technique over and over, the effort becomes less
effective each time we do it, because we find only the easy-to-find bugs. The
software is trained to pass the test. However, operating the way we do, with
risk taking and variations, guarantees against what we call regressions—bugs
that recur or slip back in to the product. To put it another way, the cost of
the low-trust work environment for us is that we run the chance of missing
bugs because we do the same thing over and over. Conversely, in a hightrust
environment, people will experiment with new types of testing, and
as they do, they will uncover new bugs.
AJP: Do low-trust environments also discourage innovation?
Smith: Yes, that’s true mainly because a low-trust environment discourages
independent decision making, collaboration, empowerment, and process
improvements. People stick to the status quo, thereby reducing or eliminating
the chance to apply new and creative thinking to existing business
challenges.
AJP: Can you give us other instances where games are better than other means
for using employee skills?
Smith: Yes, there are two workplace scenarios where games are, in my opinion,
better than other techniques. The first is in areas where employees can
develop or expand skills that help with regular work. The second is in areas
such as organizational citizenship where new skills might help the team
but are not part of the regular job. By adding games and game elements,
we can make both types of training more attractive and more rewarding,
thereby encouraging and attracting effort.
AJP: So games teach?
Smith: Yes. There are many success stories about games in education. Microsoft
Research is partnering with the Rochester Institute of Technology to do
fascinating work to gamify elements of the overall educational experience.
The school’s Just Press Play game integrates game mechanics with
desired student behaviors to encourage participation and thus tie together
academic and social experiences in ways that enhance both. Also, there
are some great studies on the VGALT [Video Games as Learning Tools]
website showing that games are an effective way to teach. In particular, a
piece on learning through the Dance Dance Revolution video game offers
some interesting lessons around the role of identity and engagement. As
I mentioned earlier with regard to motivational psychology, games can
provide an alternative playground where players act differently. Of course,
it’s only natural to extend these findings to employees as well as students.
AJP: What evidence is there that game-driven training or testing beats traditional
methods?
Smith: The notion of playing at work is an oxymoron to some. For them, work is
work, and using games to get real work done is a gimmicky idea. Naysayers
propose that a paycheck is sufficient motivation, and games are a distraction.
Why would we need games when we already have a reward system in
place? Gamification—as a term—has a bit of a reputation as a cheap way to
attract attention. The criticism from game-design professionals is that while
adding a leader board might provide a short-term boost, sustained effort
requires deep thinking around game design and motivation. We’ve heard
all those criticisms, so we are very meticulous and scientific in collecting
our supporting data. We instrument every aspect of game play so that we
have real data to support our investment. Software testing benefits from
the diversity of users, and so the more diverse our beta-user population,
the more likely we are to deliver quality products. If we can get marketing
managers, executives, engineers, and human-resources professionals to use
our software and walk through the same scenario, their collective feedback
proves paramount. The problem becomes how to motivate each of them
to do a little extra work for us—to try something new and to spend time
giving us their feedback. That’s where games work great.
AJP: Are there other ways in which games solve problems better or more efficiently
than traditional management methods?
Smith: Games are excellent at attracting volunteer effort—encouraging organizational-
citizenship behaviors (OCBs). OCBs are best thought of as going
above and beyond the call of duty—things individuals can do to help the
organization be a better place. A simple example is cleaning the coffee pot
before people leave for the day. It’s a task that helps the organization—
makes a better workplace—but requires some effort from someone. From
a game-theory perspective, there’s a condition known as the volunteer’s
dilemma—where any single individual can offer personal time to solve a
problem or anyone can take a free ride. Everyone benefits from anyone’s
willingness to volunteer, so using game mechanics to invite participation
solves the issue and improves the quality of life in the organization while—
most importantly—making everyone feel good about it. Games and game
mechanics motivate players to make an effort towards a goal, and the organization
benefits.
As I mentioned earlier, we have learned that games are also incredibly
successful at encouraging risk taking. The rules of the game are different
from the rules of the organization. The stakes in a game are much lower
than those in the workplace. Games are by definition voluntary, so whether
players decide to take risks in the game or not doesn’t matter in the context
of work. People are less fearful of losing (or not winning) a game than they
are of failing at work.
AJP: Can games encourage collaboration in the workplace?
Smith: Yes. Team-based games help people collaborate. We have two great examples.
In the Windows Security team, we wanted to get some additional
work done to meet stretch goals—goals above and beyond our required
deliverables. We didn’t need the entire team to change and do something
different. We only needed about 20 percent of the members to participate,
but we also needed the remaining team members to pick up some of the
work of that 20 percent to keep things moving. One member of our group
designed an Olympics-themed game with multiple teams and with a team
make-up of coaches and runners. The coaches became the 80 percent who
picked up the extra work of the runners who were doing the stretch-goal
work. Points were awarded for continuing the regular work, and judges
assessed the stretch work. The scoring was team based, so each team could
decide which players they needed in which roles. The game play took place
over about a one-week period and was supported by a couple of parties and
food fests. As a team, we were able to maintain the regular work through the
extra effort of the coaches and achieve some of the stretch work through
the extra effort of the runners.
AJP: Can you give us another example?
Smith: Here’s one from our prerelease testing of Microsoft Lync, a product that
replaces the phone and adds instant messaging, online meetings, and video
calls. To make sure we could ship a better product to our high-usage customers,
we needed to execute a large-scale beta program inside Microsoft
with tens of thousands of employees. In trying to do this, we faced three
problems. The first addressed how to ensure that our beta users would
come away with a positive impression of our product, even if they found
bugs. The second concerned how to get users to pair up to try different
scenarios. If we wanted someone to try a video-calling feature, for example,
how would we find someone willing to be on the other end? The third was
how to motivate the users to take the extra time to write up and send us
their feedback. It’s hard enough to get users to take the time to learn about
any new features; getting them to take even more time and write up their
impressions is even more difficult.
AJP: So what kind of game did your team design to solve these problems?
Smith: It was a game we called Communicate Hope—A Benefit for Disaster Relief.
Our users chose to join a team representing one of five real disaster relief
charities. All their work for us earned points for their team—the agency
they selected. We had team captains, videos, and related charity materials,
and as the users performed tasks with our product and gave us feedback,
they earned points for their team. At the conclusion of the game,
we donated money to the agencies in proportion to the number of points
earned for that charity. We saw tremendous results in the volume of feedback
and in the responses of gamers versus nongamers. In fact, in many
situations, those on a team representing a charity gave as much as sixteen
times more feedback than nongamers.
AJP: What impact do games have at work—for example, do they create greater
uniformity, or do they expand creativity?
Smith: I believe they expand creativity. Just as play helps kids pretend, experiment,
and learn skills they will use later in life, games in the workplace help
build a culture that is ripe for creativity and innovation. Again, I think it
comes down to risk taking and a freedom to fail, which games and play
facilitate. In 1996 about six years after he published his famous book on
flow, Hungarian psychology professor Mihaly Csikszentmihalyi wrote a
book on creativity. In it, he talks about the influence of environment on
creative capacities and how many cultures—from the Chinese sages to the
Hindu Brahmins to the Christian monks—sought out places of natural
beauty in which to create. He goes on to talk about the influence of the
macroenvironment—the broader context in which people work—ancient
Athens, the Arab cities of the tenth century, Florence in the Renaissance,
Venice in the fifteenth century, and so on. Obviously, we don’t operate at
that level, but I’d like to believe that the spirit of freedom, fun, and whimsy
surrounding our application of productivity games contributes to a creative
atmosphere.
AJP: You said earlier that low-trust work environments discourage risk taking
and innovation. Are there similar costs in a play-deprived environment?
Smith: The costs of a play-deprived environment are challenging to identify.
We have data on cost savings resulting from introducing games and play
into some of our business processes, but I don’t know if we could assume
our experience would transfer to all environments. There are certain areas
where games work well, and so depriving those areas of play and games
could result in missed savings. An area that’s easy to quantify is employee
morale and retention. On teams that encourage play and games, people
generally enjoy their work more. Csikszentmihalyi’s concept of flow—a
state where people move so deeply into their task that nothing else seems
to matter—is more likely when play is present. People do better work when
they are happy, engaged, and motivated, and play and games can increase
the likelihood that people enjoy their work.
AJP: Can games go so far as to combat workplace alienation and disengagement?
Smith: I believe they can. Well-designed productivity games can bring people on
teams together or invite friendly competition that paves the way for people
to work better together. Not everyone is motivated exactly the same way,
and so traditional employee morale improvement efforts—bowling nights,
trips to the movies, team dinners—are hard to tailor to unique preferences.
Game play, however, can be tuned to appeal to those who respond to leader
boards, those who respond to puzzles, those who want to collaborate, and
those who want to beat their own high score. Games leave room for ties
to certain affinities such as sports teams, hobbies, and charity work, and a
well-designed game can adapt to the individual user’s competency, interest
level, and engagement. In his The Anatomy of Melancholy, the seventeenth century
English scholar Robert Burton quotes Spanish humanist Juan Luis
Vives as saying that “mirth purgeth the blood, confirms health, causeth a
fresh, pleasing, and fine colour.” Then Burton adds that mirth also “prorogues
life, whets the wit, makes the body young, lively, and fit for any
manner of employment” (p. 119). I agree. Using games in the workplace
brings an element of fun, whimsy, and color. They help coworkers find
common interests, and they present a foundation for mirth. This combats
alienation and disengagement.
AJP: So, then, you equate the play-averse workplace with the risk-averse workplace?
Smith: For two reasons, yes. First, since play is typically unstructured and
optional, a workplace willing to entertain the idea of play is, by default,
willing to take risks. Just being open to the introduction of play at work
implies a tolerance for risk. Second, and most importantly, play can provide
a loose structure for experimentation and risk taking, so an organization
that is averse to play, games, or fun does not offer the flexibility of outcome
or tolerance for imperfect results. An organization that is amenable to
play is likely to be a high-trust organization willing to show tolerance for
experimentation and for provisional and imperfect results and, therefore,
have creative behavior and innovative breakthroughs.
AJP: Are you suggesting that game strategies can ease tensions between management
and workers?
Smith: Yes. Because games require agreement on rules and scoring, they essentially
force management and worker interests to coincide. Imagine a manager
describing a task or assignment to an employee with the same degree
of detail that a game offers. “Please sweep that section of floor with this
broom in that direction resulting in that dust being cleaned in one hour,
and I will pay you five tokens as you progress through the assignment. I
may even give you a bonus of additional tokens, but I won’t tell you my
decision about that until you finish.” The level of effort that must go in to
good game design forces agreement on goals in a way that’s not normally
done. However, everything is done in the context of fun, and so the work
doesn’t seem as meticulous or arduous.
AJP: So, have you found that games are generally more effective than traditional
management methods—for example, salaries, bonuses, and time off—of
providing worker incentives?
Smith: No, not exactly. For the right tasks, game mechanics can be more effective,
but only as a supplement to traditional rewards. People do their jobs—
the in-role portions of their work—for their salaries, bonuses, and other
regular benefits. Our work with productivity games attempts to motivate
people to apply part of their discretionary time to citizenship behaviors
that help the organization in other ways, and our experiments have shown
that expensive prizes do not work well in these situations. We try to tap
into intrinsic motivations. Perhaps an elaborately designed framework
that incorporates traditional rewards and incentives with game mechanics
could work, but our experience has been focused on OCBs, which makes
a direct comparison to traditional rewards hard.
AJP: Do productivity games work well under all types of managers, for example,
hands-on managers as opposed to more hands-off managers?
Smith: Productivity games will work with both styles. The flexibility and self monitoring
capabilities of games is another important dimension of games
as a business process. They help lead or direct in a very adaptive, flexible,
and responsive way. Employees who respond well to certain management
styles find direction in games—through their method of play—even if the
manager does not operate in that fashion. Games are flexible and willing
to change their style, and so they have a capacity to adapt to the actions of
the player better than a human can. Employees who perform best under an
autocratic leader and employees who look for a more democratic or laissez
faire–style manager can both find what they need in a well-designed game.
AJP: How do you convince bottom line–oriented, quality managers that workplace
games are not just for playing around?
Smith: We found that skeptical managers who doubted the potential of games
and labeled them gimmicky needed solid facts—objective and subjective
data—before they could concede that games help. As a result, we have taken
an academic approach to data collection. We have gathered a ton of data
over the years. My advice to those who might be thinking about bringing
games into their own organization is to measure everything, collect as much
data as possible, and follow up and interview everyone. The importance
of data—both objective and subjective—cannot be overemphasized. The
answer to your question is, quite simply data, data, data. It is not fair or
reasonable to expect anyone to react or change for anything other than
hard facts.
AJP: How would you respond to critics who might charge that gamelike work
processes reward workers psychologically but cheat them financially?
Smith: Jane McGonigal has a great comment relative to this in her book Reality
Is Broken: Why Games Make Us Better and How They Change the World.
She says that compared with games, reality is pointless and unrewarding,
and when we play games, we feel more rewarded for putting forth our best
efforts. I agree. Productivity games are designed to offer an alternative—to
help an organization recapture discretionary time in the workplace for the
good of the group, not to change a fundamental work process or offer an
alternative reward system. There are too many factors outside our influence
to be able to redesign entire work processes. Our work so far has been
to give people a forum where simple incentives, such as solving a puzzle,
are available to reward them. Success means that once in a while, people
eat lunch at their desk and play one of our games to help the organization.
Again, our experience has been that applying game mechanics to get
people to work more at their regular job does not work for exactly the
reasons you note—people feel as if the game in those situations is trying
to take advantage of them and their desire to play. Remember that play is
a voluntary activity.
AJP: Are you usually pleased with the results you get in encouraging organizational
citizenship behaviors?
Smith: Yes. Most of our games are very simple. We don’t have a dedicated development
team or production budget, so we rely on a lot of volunteer development
effort. Our typical game may have some product instrumentation,
a website, and a database. Our Communicate Hope game was probably
the largest effort so far, as we needed to coordinate charity donations and
secure a budget to donate. Our expectations for that game were pretty
high, but we were unsure how the disaster-relief theme would influence
game players. We had several areas where game participation exceeded our
expectations. The volume and quality of feedback, player satisfaction with
the experience, and engagement with the set of active players also exceeded
expectations. On the other hand, we were expecting more team-based play
than we saw, and the absolute number of players was a little lower than
we expected.
AJP: Are there any workplace outcomes that lie beyond the reach of games?
What limits are there to the usefulness of games?
Smith: Yes, there are definitely limits. We have seen some games fail miserably
when used to try to increase productivity in people’s core, in-role tasks that
they do as part of their regular job. If I am a grocery check-out bagger, then
bagging groceries and carrying them to a car are in-role tasks, and helping
my coworker fill out information on the benefits website probably is not.
If a game is produced and deployed to get me to do more of my regular
job, I might get confused. How do points relate to my paycheck? If I win
points or the game, do I get a promotion? If I lose, am I in trouble with my
manager? Also, there may not be many players who can do my job, so the
amount of additional work that gets done might be minimal. People can
get very emotional when they feel that game mechanics are put in place to
trick them into doing more work. And that has a negative impact on trust.
Games should not be used to motivate people to apply core skills to do
more of their in-role tasks. Use the paycheck and traditional compensationreward-
evaluation system for that.
AJP: And this is because play is usually voluntary?
Smith: Yes, exactly! As I said before, play is by definition a voluntary activity.
Games and play are more fun when we can choose to participate or not.
A game that has me doing more of my job is not voluntary; therefore, it’s
not really even a game. We have learned from experience that using game
mechanics to attract effort to core work is not received well and usually
does more harm than good.
AJP: When you employ game mechanics at Microsoft, how many people can
play the games?
Smith: We have had thousands of players for some of our more broadly deployed
games. Our Language Quality game had over forty-five hundred players.
However, we tailor portions of our games to specific groups. We respect the
notion that people are motivated by different things, and we try to account
for that in our games. Take a leader board versus a puzzle, for example.
Some people want the world to see them at the top of the leader board.
Others don’t care about being at the top but might be embarrassed by not
showing up at all. Others scoff at the whole notion of public competition
but are motivated every week to finish the crossword puzzle in the Sunday
New York Times. Some people don’t care what the world thinks. They just
want to know that they beat their personal best. It goes back to the motivational
psychology aspect. Some people are motivated by achievement or
arousal, some by humanistic tendencies (pride in country, for example),
some by stress (I earned enough points so everyone will see that my boss
is two spots beneath me). It’s different for each person, and we want to
build games that appeal to or motivate, in some way, everyone. We have
also themed some of our games to target a specific audience—college basketball
fans, for example.
AJP: Microsoft is an international corporation. Do you try to account for differing
cultural assumptions about play in other areas of the world?
Smith: We haven’t had to look too closely at that, because our games are predominantly
used inside Microsoft, and the culture of the company is very
similar across regions. The biggest area where we’ve tried to understand
and respect different cultures is the use of leader boards. Many cultures
value teamwork and collaboration over individual or personal recognition
and therefore do not advocate or respond to a public tally of achievement.
In some cultures, these competitions are very popular and motivating, and
in others, they are not. We’ve also seen some differences in how managers
perceive their employees playing our game in different cultures. In some,
managers encourage employees to participate, and in others, they discourage
participation. This often has to do with an individual manager, but we
have seen some trends across geographies. In the future, I believe it will be
important for game design to account for cultural diversity and a global
player population.
AJP: How does your Language Quality game relate to such diversity?
Smith: The Language Quality game was built to enroll native-language speakers
in helping us assess the linguistic quality of Windows translations. The
linguistic quality looks at the accuracy of the translations across languages.
The geopolitical ramifications of translation errors can be significant, so
accuracy, broad acceptance, acknowledgement, and review are important
aspects of the pursuit of high quality. We didn’t do a lot to tailor the game
play specifically to the individual languages or countries, but we did try
to encourage a sense of national pride across languages. We used country
mailing lists to invite employees to play, and we kept score by language as
well as by player. Many employees spoke of the importance of quality of
language translation to perception of code quality. Relative to a game like
Halo or even Angry Birds, it was a very simple game but effective because
it lies squarely in the realm of organizational citizenship. People wanted to
ensure that their language version of Windows 7 was high quality.
AJP: Do traditional games such as card games, checkers, and others influence
the design of productivity games at Microsoft?
Smith: Yes, absolutely. The simple games often contain the greatest lessons for
how to keep play interesting. We have a great card game, designed by Adam
Shostack in Microsoft Security, called the Elevation of Privilege. It’s inspired
directly by traditional card games, and we use it to help build threat models.
A threat model is a representation of a set of possible attacks against
the functionality or features of software, and it helps us build more secure
software because potential attacks are evaluated and discussed as the software
is being built. Elevation of Privilege is for three to six players. The
deck contains seventy-four playing cards in six suits: one suit for each of
the STRIDE (Spoofing, Tampering, Repudiation, Information Disclosure,
Denial of Service, and Elevation of Privilege) threats. Each card has a more
specific threat on it. As game play progresses, the threat model is built. As
players take turns in the game, they imagine new attacks in certain categories
dictated by the game play, and they develop the model by taking turns
throughout the game. This is a very specific example of how card games
can influence productivity game design.
AJP: With the considerable number of games you use, how do you keep players
playing? How do you make the play itself sufficiently rewarding?
Smith: One of the key goals in game design is to attract and engage as many
players as possible for as long as possible. Productivity games are slightly
different from games for entertainment, as we can’t count on genre to
appeal to a niche player. The target audience for a productivity game may
spend their discretionary time playing chess, Farmville, Halo, Ms. Pacman,
solitaire, Jenga, NASCAR, Angry Birds, crossword puzzles, or Madden 12.
Because we are trying to encourage beta testing and feedback, we benefit
tremendously from a great diversity of players. So we have to design games
that have appeal to a broad demographic. We rely on standard game-design
techniques and try to ensure that all our games appeal to players who enjoy
player versus player, player versus self, and player versus environment.
AJP: You said earlier that play is voluntary, so may we assume you do not require
employees to play your productivity games?
Smith: Yes, right. No one is required to play our games. In The Grasshopper:
Games, Life, and Utopia, Bernard Suits states very clearly that playing games
is different from working. For a game to be fun, or for play to be truly
voluntary, a player must be able to choose to stop playing at any time. We
expect that in our productivity games. We don’t require or anticipate that
anyone will play for a certain period of time. We believe that if we required
that, these would not be games anymore, they would be work.
AJP: Does competitive drive encourage wider participation in workplace games?
Smith: Yes, it absolutely has a place. It is one of the top four or five considerations
in appealing to a broad group of employees. In game design, this is known
as the player type who is motivated by elements in the player-versus-self
category. This is an important category, particularly in a number of cultures
from around the world. As I noted earlier, many cultures frown on or discount
the idea of a leader board. They believe strongly that no single person
should be better than another and that work as a team is what comes first.
Without a player-versus-self category, many people would simply ignore a
leader board, go about their business, and ignore the game. Adding game
mechanics to appeal to those who want to beat their personal bests allows
each individual to improve without stealing the spotlight from the team.
AJP: How does success at productivity games relate to career growth for those
who play them?
Smith: The traditional career ladder is often considered a zero sum game—like
cutting a cake—the bigger your piece is, the smaller mine must be. I would
propose instead that career growth over the long term is fueled by skills
growth. As an employee, there are many different ways I can grow my
skills. My manager might assign me a new task that stretches my ability, or
I might get to take a job-related training class, but that’s about it. In both
cases, I’m not choosing the skills I personally want to develop. To grow,
I need to stretch. Many organizations have stretch goals or extra time to
allow employees to spread their wings. But, for employees whose companies
don’t provide that, any attempt to think out of the box is a risk. We
talked about that—and about trust—earlier. I would like to come back to
it in this context. Many employees are comfortable taking risks and are willing to explain
to the boss why things did or didn’t work out, but others are not. In a down
economy, with high unemployment rates, many of us are just thankful
to have a job. An individual’s willingness to make bold bets and take big
risks on the job is reduced by fear—fear of not doing well at performanceevaluation
time, or fear of being reprimanded, or even fear of being fired.
Generally, most organizations do not reward failure, and it takes a hightrust
environment for people to take big risks. So productivity games are an
important way for management to encourage risk taking and experimentation
that lead to innovative output. Many successful employees think back
over their careers to examples of coworkers they competed with in ways
that turned out to be great for their careers.
AJP: Do employees from the gamer generation view productivity games in significantly
different ways than other employees?
Smith: There are some fascinating differences. For anyone interested in this, I
recommend Got Game: How the Gamer Generation Is Reshaping Business
Forever by John C. Beck and Mitchell Wade. They point out that gamers
view the world differently. They tend to have an increased appetite
for risk, are more willing to persist in finding a way out, are more open
with their communications, and tend to bring more creative solutions to
problem solving. These are generally our younger employees, of course.
However, the recent success of social games like Farmville and Angry Birds
has influenced everyone. One key thing I have seen and tried to work with
is the importance of immediate feedback—gamers’ tendency to canvass
the surrounding environment for hints of how to improve, their willingness
to press reset and start over, and how they value reputation. For me,
I’ve learned that all this is about more than games. There are some lifestage
and generational differences that include, but also transcend, gamers.
The Framework for 21st Century Learning—put forth from P21, a
national organization advocating twenty-first-century learning skills for
all students—highlights the four Cs: Critical thinking and problem solving,
Communication, Collaboration, and Creativity and innovation. Some
interesting research on children shows that more young children know how
to play a computer game (58 percent) than swim (20 percent) or ride a
bike (52 percent). While almost 70 percent of children ages two to five can
operate a computer mouse, only 11 percent can tie their own shoelaces.
According to the Entertainment Software Association, 64 percent of parents
believe games are a positive part of their children’s lives. The digital
media firm eMarketer expects the number of U.S. social gamers to grow
to 68.7 million in 2012, with almost 30 percent of the Internet population
playing social games.
AJP: Will this change work as we know it?
Smith: Yes, it will. Shifts in global, societal, technological, economic, and sociopolitical
trends will shape the future of work. Work and life will become
blurred. Already almost half of U.S. employees work beyond normal hours,
and one-third do personal tasks at work. Emerging economies, globalization
of the workforce, and smart and connected technologies that enable
mobility and flexibility will lead to an increased use of game mechanics
in the workplace of the future. Games can help businesses cross cultural,
generational, language, and geographic boundaries, and as the gamer generation
enters the workforce and as society shifts towards a more plugged-in
global village, it’s inevitable, in my mind, that games will be a big part of
the shift.
AJP: How does the fast pace of change in communications outside the workplace
compare to the pace of innovation inside the workplace? Is business
the crucible of innovation that most people think it is? If not, how can the
workplace keep up?
Smith: The two seem hard to compare. There are significant pros and cons, or
strengths and weaknesses, in each environment. Business communication
has always been rapid, reputation based, and reliable. Brands, job titles,
and marketing efforts help shape reputation, which influences the quality
of the communication. Social communications have grown by leaps and
bounds as a result of the transparent, open social network. I think there
are opportunities for each—the workplace and the world outside it—to
learn from the other. Businesses can benefit from crowd-sourcing and
social-networking tools. I believe that the changing demographics in the
workplace will bring new technologies and communication capabilities
for everyone.
If I use the dictionary’s definition of geek as “an enthusiast or expert
especially in a technological field or activity,” most of our employee audience
is geeky, and the games are targeted at them. I would expect that any
game built to attract organizational citizenship efforts from a group of
employees in any industry would be targeted at enthusiasts or experts in
that particular field. As productivity games move outside organizational
boundaries, I think they must become more mainstream and less geeky.
Game designers and game masters must know their audience. For example,
our games use acronyms that are familiar to our audience, but those would
require more explanation and context if our games were distributed to a
broader audience.
AJP: Summing up, you believe that productivity games translate into measurable
gains or reduced costs for business and other organizations. Correct?
Smith: Yes, absolutely. As Reeves and Read discuss in Total Engagement, games
can help with employee engagement, and whether that shows up in cost
savings, additional productivity, crowd sourcing, employee retention, or
recruiting, games can deliver explicit cost savings to an organization.
AJP: Finally, what is your fondest hope for the future use of play and games in
the workplace? How would you like to see play and work grow to more
closely resemble each other?
Smith: I believe productivity games will be viewed as a business process—a
twenty-first-century business management strategy—and applied widely
across a variety of industries. We’ve already seen more companies start
to pilot the use of game mechanics as part of their work, and with the
success of social games, it’s only natural that games and play will permeate
the workplace over the next few years. There are distinct areas where
games work tremendously well in the modern organization, and there’s
an opportunity for everyone to start experimenting. The future world of
work will be a better place by incorporating play and games as part of the
daily experience.
To read the original interview, please click here.
To go to the American Journal of Play, please click here.