Cultural change accelerates as the children of the Atari generation get their pudgy little hands on media we never thought of. Technology literacy evolves into a subconscious expectation that things work and are fun. After decades of thinking we knew what that meant, game designers suddenly find themselves facing an explosion of diversity in terms of both the technological landscape and the potential audience for interactive entertainment. This talk will explore some approaches to research and design that will help us grow both our audiences and our ideas.
Hi, I’m Cmdr. B. Laurel. I’m a Navy test pilot, and I just rode a modified F-14 into the desert floor. I’ve made the 60,000-foot downward spiral in some of the F-4 Phantoms and F-18s of the industry - Atari, the old Activision, Epyx. This time it feels a little different - I think I was fonder of the plane. My F-14 was Purple Moon and it closed its doors on February 18th. The characters should be able to walk away from the wreckage, if somebody acquires the assets. Actually, that cell phone call just now was Purple Moon’s CEO, letting me know that Mattel is going to buy what’s left of the company. If Rockett starts wearing pink, just kill me.
I promised to talk to you about some approaches to research and design that will help us grow both our audiences and our ideas. Because of what’s happened, I feel that I have to alter my course a little bit. Since Purple Moon did not make it to the big IPO or a lavish acquisition that made everybody rich, it would be understandable for you to conclude that the methods I’m advocating don’t work. No million-unit titles, just respectable sales in a new market segment that got crowded really fast. Our eight CD-ROMs and our website won all kinds of awards but didn’t make us profitable in time to satisfy our investors. But I think there may be some other measures of success, and I think there are things to be learned from our experience. Let me tell you a story.
When Purple Moon got its plug pulled - only six days after we had shipped our eighth CD-ROM title - we determined that we would need to close our website. Even though the hosting bills were paid through the end of the month, the intellectual property was in legal limbo, and more seriously, with no operating company, there would be no one to watch the safety alerts and keep our promise to girls to make the site a safe place.
So, with much sadness and beating of breast, we scrambled to create the semblance of a graceful ending in the midst of our corporate catastrophe. Kristee Rosendahl, lead designer of the site and director of our online business, worked with her team to put together a "farewell" screen on which our characters told the girls it was time to say goodbye. There was a cartoon of the web team created in Rockett's Adventure Maker, and they let me post a message to the girls about how things don’t always go the way we want, but we have to learn from our mistakes and carry on with honor and keep a positive attitude, etc. etc., stiff upper lip.
I was bombarded with email sent to my personal address, which someone would have had to search for on the web to find, from girls and parents dismayed by the closing of the company and the site. I found myself explaining bank-rupt-cy to 10-year-olds, trying to help girls get their hands on merchandise they needed to complete their collections, and struggling to comfort heartbroken girls who lost their online friends when the site closed down. Most of them had followed the practice of not giving out personal information to people you meet online, so the only connection they had with Purple Moon postcard pals was on the site itself. All of us were devastated that there hadn’t been time to warn girls so they could figure out another way to connect with friends they had made.
Somehow Nancy Deyo, Purple Moon’s CEO, finagled the funds to re-open the site temporarily on March 1st to give girls a chance to re-connect with their friends. That's when we noticed the miracle. During the nine days that the site was "down," it had acquired 274 new registered users. It turns out that the program was still running behind the farewell screen. Girls who had bookmarks inside the site could get in, and most of the girls who were using it didn’t know that there was anything unusual going on. The only way that new users could have registered would have been that their friends helped them sign up from the inside, where we had cleverly installed a button labeled, "Want to let a friend sign on? Click here." Once the site re-opened, the number of new registered users quickly regained its pre-shut-down level of about 400 girls per day. The interesting thing here is that this happened without a penny in marketing expenditures. It’s like, the dead rise up and walk.
When we first launched the site, we let kids send post cards to characters and we responded to them in the characters’ voices. I personally thought that this would be a very cool way to employ moms in cheneille bathrobes stuck at home with a bunch of kids - helping the web fulfill the promise of providing actual employment – but I was advised that this plan would be too difficult to recruit and administer. So folks in the studio started answering the postcards, which were rapidly becoming an avalanche. People chose the characters they felt most resonant with. Our story director Pamela Dell answered rocker Ruben Rosale’s email; the male receptionist donned a rhinestone bracelet and plunged enthusiastically into channeling fashion-queen character Nicole. Thousands of heartfelt, personal postcards poured in for the characters — asking for advice, giving advice, befriending some characters and chastising others for being mean. With six or seven folks working on it fulltime, we were falling farther behind every day. After several weeks we were so swamped that we had to ask girls to back off. The principal of our fictitious Whistling Pines Junior High, Mrs. Herrera, explained to our girls that the characters weren’t getting their school work done because they were spending too much time on email and asked everybody to help by sending fewer post cards. Eventually we moved to a bulletin-board style of communication because we didn’t have the budget to respond personally to each postcard, but the numbers of girls who visited and registered continued to grow.
Now, this is a kind of success and there are some ways to measure it. Purple-Moon.com was one of the three largest kids’ sites on the web in terms of traffic. There were about 290,000 registered users as of March 1st and we were averaging about 12 million page views per month. Over the life of the site, girls sent each other and the characters about 11 million postcards. The average registered user logged in more than once a day and spent about 35 minutes on the site. Girls collected and traded over 4 million treasures — collectable software objects that were scattered around the site. We launched a shopping area in the fall of 1998 that quickly garnered a better than average conversion rate. But the dollars weren’t there in either web or CD-ROM revenues to make the kind of investment that would have made e-commerce a profit center in the short term.
So what can we conclude from these facts? First, making money on the internet is haaaaaard. Gee, I bet you already knew that. But I think that our story gives some good examples of how to create and sustain a community. For the success of a brand or a set of characters, an active fan culture may not be sufficient but it sure is necessary. Early on, we learned a lot about fan communities from Professor Henry Jenkins, founder of the Comparative Media Studies Department at M.I.T. and author of several books on fandom, including Textual Poachers and Science Fiction Fans: Watching Star Trek and Doctor Who.
From our work with Henry and our own research with girls, we identified some key ingredients in the fan-culture starter kit. First and foremost, fans need to be able to "appropriate cultural material to construct personal meaning." This is culture-theory-speak for being able to fool around with characters and their universe and create your own stuff with them. Star Trek slash video, fiction, and ‘zines are probably the best examples of this kind of activity. Paramount has been clever enough to co-opt some of that fan energy and turn it into profit through "official" novels and fanzines. The fact that Trek fan culture survives despite Paramount’s vigorous prosecution of "poachers" is due to the extraordinary level of brand loyalty, and also to the tenacity and cunning of Star Trekfans, who will stop at nothing short of anonymous servers in the outer solar system to strut their stuff. In general, however, the reflexive desire of businesses to protect their "intellectual property" by stomping on unauthorized use of characters and narrative worlds is actually a pretty bad idea, since it serves to discourage the central activity of fandom.
In terms of both CD-ROM and the web, you can encourage and nurture a fan community by providing ways for users to create and publish their own content based on your world. The ability to create your own version of Doom is a great example. Purple Moon published Rockett's Adventure Maker to let kids create new characters in the Rockett art style and then put them into our world and combine them with our characters in a comic strip format. On the web, we published questionnaires and used the results in the design of new CD-ROM products. We ran writing contests based on story starters and published the winners on the site; we published user submissions to the Whistling Pines newspaper and user-authored poetry and prose; and we helped users create and publish their own personal pages. Through the postcard system, girls arrange swap meets for Purple Moon treasures. They formed clubs within the site based on treasures, zodiac signs, sports interests, animals, geographic location, and favorite characters. They also spun off independent fan sites.
One of the limits to Purple Moon’s growth as a small, independent company was the difficulty of getting our characters exposure beyond interactive media, something I’m sure that many of you have struggled with. That big TV deal never quite closed. The characters will appear in a new series of Scholastic books starting next fall – better late than never. The point is that fan behavior kicks in only when characters are sufficiently familiar and deemed to be personally relevant. Purple Moon's fandom was small in comparison with Star Trek's, but it was large and extremely active compared to other interactive properties with similar reach. I think we can attribute that to two things: the affordances offered by the website, and the appeal of the characters and their world.
This seems to be a good place for me to segue into research and how it relates to design, because it was our research with girls that taught us how to design our characters and what kind of world to put them in, as well as our products and our site. You may or may not be interested in our research findings, but I think that the research process we followed could work for any new audience that you might want to explore.
Since I started the gender & technology project at Interval Research in 1992, my principal collaborator has been Cheskin Research, headed by Christopher Ireland and her husband Davis Masten. Cheskin does work for huge consumer brands like Microsoft, Pepsi, and Frito-Lay as well as for smaller companies like Interval and Purple Moon. Christopher has developed some incredibly effective techniques for finding out about people in a potential audience. In fact, she is giving a talk in this track later this afternoon entitled "What's Love Got to Do with It? Exploring Consumers'’ Emotional Attachments to Products" that will give you a taste of her talents.
When I say "research," I’m not talking about what we typically call "market research." Market research as it is usually practiced is problematic for a couple of reasons. Asking people to choose their favorites from among all the things that already exist doesn’t necessarily support innovation; it maps the territory but may not help you plot a new trajectory. On the other hand, most people are not very good at inventing new objects of desire. If you had asked someone in 1957 what new thing they would like to play with, chances are they would not have asked for a plastic hoop that they could rotate around their hips. Somebody had to invent the hula hoop. Well then, you say, the point is to find good, frisky inventors. Who needs a bunch of pencil-heads in white coats asking questions? Innovation is for the creatives. But no one in their right mind would suggest that research is a replacement for creativity. Research does you no good if you don’t empower creative people to invent novel, cool things. Doing research and paying attention to your findings can simply better the odds of success by illuminating the space of possibility and focusing creative energies.
I learned while working with Interval and Cheskin that it’s not enough to understand people statistically; you need to find out as much as you can about what their lives are really like. Some of that is quantitative information, like how much TV they watch or how often they play videogames. Some of it is qualitative, like what bothers them or how they define themselves or how they represent themselves to other people. Some of what you need to know may be scientific - for example, how good are people at pattern-matching, and how do their vision and reflexes change with aging? Some things are better gotten at through personal conversations and stories - what are your favorite things to do? Or, what magical powers would you like to have? Questions like these aren’t particularly useful if the answers are multiple-choice - you can't learn much from "none of the above."
Here’s what I think are four handy tricks for doing good research that can inform design.
The fundamental research for Purple Moon combined several techniques. We defined our goal very broadly. Initially, our question was, "Why aren’t there any successful computer games for girls?" In 1992, this was basically the state of affairs, and attractive though it might be, a sexist conspiracy theory was an inadequate explanation. That led us to ask, "Is there any inherent reason why girls wouldn’t play computer games?" and "What kind of games would attract girls?" In order to answer those questions we needed to understand fundamental characteristics of how girls and boys play. So we finally arrived at an appropriately broad research question: "How is play influenced by age and gender?"
We started with a massive literature search to find out what was already known about gender differences that might be relevant to play styles and preferences. Then we followed up on the most promising findings with interviews of the scientific and academic experts who produced them, along with "play experts" in other fields like sports and toys. We complemented the expert interviews with focus groups consisting of adults who actually spend time with children at play, giving us a chance to corroborate the expert findings from the perspective of people "on the ground." These findings gave us a pretty good idea of the key topics we needed to probe in interviews with children.
Cheskin advised us that focus groups with children this age would be problematic because of the tendency of groups of kids to be strongly influenced by other kids’ responses. They recommended that we do what is called dyad interviews - you invite a child to be interviewed and ask him or her to bring a friend. The two friends still influence each other and often express the same opinions, but they also tend to keep each other honest. They call each other on inconsistencies, which provides a check on the fairly common habit of kids (especially younger ones) to try to please by saying what they think the researcher would like to hear. We talked to kids in eight different U.S. cities to make sure we would capture any regional differences and paid special attention to racial, cultural, and socioeconomic diversity.
Over the course of about a year and a half, we talked to about 1100 kids. The interviews typically lasted a little over an hour. We videotaped each interview and periodically summarized findings and created tapes of video selects. We bracketed this qualitative research with quantitative results - questionnaires relating to key points — from about 10,000 kids. We also did focus groups with about 500 parents to be sure we weren’t going to do something that would send them through the roof.
After this period of gathering information, we represented our findings in a variety of ways. Cheskin created a taxonomy of play types and a map of play preferences as they vary with age and gender. They also provided a model of segmentation and trend movement in teen culture that we were able to modify for girls. The new model became the basis for character types and social dynamics in our narrative universe. I created a diagrammatic representation of how core girls’ concerns translated across their personal and social lives, which we used to identify themes and activities for the products. We translated our basic findings into a set of design principles that we would refer to every time we designed a new product or feature. One of the values of having design principles is that when you violate one of them, you have to do so consciously and with a pretty good reason.
Because this session is not technically about girls and girls’ games, I won’t dwell on our research results. But I do want to highlight a couple of findings that dovetail with my earlier comments about establishing a fan community. The desire to create and share narratives propels fans to come together. It turns out that girls tend to prefer narrative play and are attracted to narrative complexity more than boys in the 7-to-12 age range. The vast majority of creators of Star Trek slash video and fiction are female, and it’s a good bet that this fascination with narrative construction continues into womanhood. That doesn’t mean that boys don’t like stories or that girls can’t get engaged in things that aren’t stories — both of those statements are clearly untrue. It simply means that there is a tendency - a preference - that can help us design things that are likely to be attractive to girls.
Another set of interesting findings has to do with personal relevance. Boys may be embarrassed by materials that are too close to their real lives; girls demand them. Again, these are tendencies and preferences, not hard and fast rules. Boys tend to be attracted to superheroes and exotic action scenarios. Girls are drawn to characters who are more like them - even their superheroes. So, for example, Sailor Moon talks about her mother and Sabrina the Teenage Witch goes to school and has regular friends. We heard a lot of demand from girls for stories and characters that related to their actual lives and the issues they face – problems with parents and siblings, the desire to be a little older and more sophisticated, the tension that often arises between being popular and doing the right thing.
There are findings about how girls collect and arrange things and how they represent themselves that are hard to summarize without giving you a look at the data. We used a method that I found invaluable – asking kids to take pictures of their lives and the things they value. To give you a bit of the flavor of such qualitative research, I have put together a short video montage of girls’ photographs along with their voices from some of the interviews.
[videotape]
How can some of the techniques I’ve described be used to explore new audiences? The first step is to identify the group you want to reach. At Purple Moon, we knew that we wanted to reach girls, but there are huge differences between age ranges. We had to understand these differences and figure out where our energies would best be applied. We settled on 8-to-12 year-olds for two reasons. We could see that 3-to-7 year-olds had more offerings to choose from in the "child entertainment" category. We also knew that girls were at greatest risk of moving away from technology at about sixth grade, so if we were going to change their minds about that, we would need to intervene before it happened.
How do you identify a new audience? There are several simple questions you might ask.
Who isn't playing games? - what groups under-represented in the game-playing community? Demographically clear examples are groups like adult women and seniors. Another slice is people who do not yet have computer or videogame hardware. Yet another segment might be people who feel that playing games would be a waste of time, for themselves or for their children.
Who isn't playing much? - what groups are playing only one or two games, instead of lots? Do they have something in common with people who only go to one or two movies? Are they simply brand-loyal to a single company, like parents who will only buy their kids Disney software? What are they doing instead of playing games, and how might we offer some of the same sorts of experiences? - for example, web-surfing seems actually to be displacing some television viewing in many people’s lives. Is it also displacing computer games? What are some of the aspects of web-surfing experience that could be captured in games?
Another way to identify a new audience is to identify a new kind of product. Where are the empty spaces in the landscape of possibility? What sorts of entertainment experiences do people have with other media that we haven’t yet captured in ours? Certain forms of socialization and conversation come to mind; for example, when I was working at Atari in the early 80s, I was totally bummed that my favorite game, Star Raiders, didn't have a "negotiate" button. Another example is material culture - that is, all the fun things people do with things. Purple Moon’s treasures are an example of what is coming to be called "i-ware" - software objects that enable some of the activities associated with material culture.
Another elusive sort of segment is all those people whose expectations and aren’t being met by what’s out there. What are people’s criteria for "fun" and "ease of use"? What objects, products, or experiences do they hold up as ideal? Christopher Ireland’s talk this afternoon will give you some great examples of products people love and how we might extrapolate from them. What designs come to mind when you target people who like adventure travel, or Coca-Cola chotchke, or the new Volkswagen bug? See Christopher’s talk to find out.
Well, I want to leave time for questions so I need to stop talking, as much as I would love to wax eloquent about the seven deadly warning signs of a flailing startup. I hope you will note that I have obeyed my GDC handlers and have refrained from discussing social or humanistic values throughout this talk. Any enlightened beings out there who might want to know what I think about that are invited to visit my website, tauzero.com. Thank you for being a great audience.