Brainstorming Electronically

Electronic brainstorming (EBS) allows working groups to generate an abundance of ideas anonymously. Our experience with electronic brainstorming in several settings has shown it to be useful for large and small groups, for a variety of topics, for groups that meet face to face and for those that are dispersed, whether throughout a building or around the world. This evidence suggests that electronic brainstorming is a better way to generate ideas than both traditional brainstorming and nominal groups (individuals working alone).

Traditional brainstorming has been used for several decades. Whether formal or informal, the process is the same: think of as many ideas as you can and say them out loud; leave the evaluation of the ideas until later; build on and combine others’ ideas; be as imaginative as possible — the wilder the ideas the better. Sometimes brainstorming works just like that. When it does, ideas flow freely from an interplay that may never have occurred if the group hadn’t brainstormed together. Unfortunately, what often happens is quite different.

Two problems have plagued traditional brainstorming: production blocking and evaluation apprehension. Production blocking happens when we have an idea, but someone else is talking. When it’s our turn, we’ve forgotten the idea, we think our idea is redundant or not that good, or, particularly if the group is large or dominated by talkative people, we lose interest and don’t say what we think. Fewer good ideas are generated than might have been without production blocking.

Evaluation apprehension refers to our anxiety about what others will think of us if we say what we think. We censor our ideas. Instructions to be creative and novel aside, we don’t want to say things that may get us labeled as odd. This can be a particularly strong inhibition when our ideas may be construed as critical of current practice or when the group includes our boss and other people who may affect our fate. Under these conditions, we don’t produce as many new (and potentially useful) ideas. We keep them to ourselves, thereby defeating the purpose of brainstorming. As a result of these problems, traditional brainstorming hasn’t worked nearly as well as it’s supposed to.1

In this paper, we describe the use of networked computers to help groups generate, disseminate, evaluate, and act on ideas. The method is, in essence, a technological update of traditional brainstorming that eliminates many of the problems that have plagued traditional face-to-face brainstorming.

Traditional Brainstorming

It has been clear for at least twenty years that brainstorming is not as productive as its early advocates claimed.2 Dozens of studies have compared the number of unique ideas produced by groups of four or more individuals with the productivity of the same number of individuals working alone (so-called “nominal groups”). These studies have included managers, advertising personnel, scientists, and students. In every case, the brainstorming groups produced fewer ideas than the nominal groups, and in no case did the brainstorming groups produce better or more creative ideas.

Despite these results, the popularity of brainstorming remains high. This is attributable to brainstorming’s face validity: many of us have had the experience of thinking of an idea right after someone else expresses one. We then attribute our idea to synergy: we think that bringing people together creates new energy and ideas. The popularity of brainstorming also seems attributable to the satisfying social nature of the interaction and to the fact that we rarely experience the main alternative (nominal groups) in which individuals write down their ideas and toss them into a pile. Because brainstorming seems worthwhile, it is likely that brainstorming in some form, including the nominal group technique, will persist as a way of generating ideas.3 What EBS does is retain the synergy while reducing the main drawbacks of traditional brainstorming.

Electronic Brainstorming

Jim Westland, regional vice president of North American Life and Casualty (NALC), decided that fresh ideas were needed about ways to cut costs and improve efficiency at the northeast regional head office.4 Pressures were mounting to reduce costs, attract new business, and maintain service. He decided to bring his senior management team and selected lower-level employees to a one-day, electronic brainstorming session at a local university’s computer-supported decision center. The team consisted of seven senior managers from various functional areas, two clerical staff, and one sales representative. As Jim explained, the purpose of the session was to generate fresh cost-cutting ideas for the firm so that at the end of the day they would have four or five good ideas that could be implemented.

The development and use of electronic brainstorming software is a recent innovation in group decision support systems. We now have several reliable programs that (1) enable groups to generate, disseminate, and evaluate ideas and (2) reduce the problems of having to wait your turn, worry about how your ideas will be heard, and function in a group that is dominated by a small number of individuals.

Programs such as GroupSystems, TeamFocus, Vision-Quest, and Software-Assisted Meeting Management all provide the same basic functions. All are available commercially and are compatible with most hardware. We will use GroupSystems to describe the generic features of these programs because it is the software with which we have the most experience. Table 1 summarizes the electronic brainstorming process.

Generating Ideas

Arriving at the decision center at 9:00 A.M., Jim Westland and his team took their seats around a U-shaped table facing a large projection screen (see Figure 1 for a decision-center layout). The room reminded Jim of the “war rooms” he had seen in movies and on TV. Each place around the table had a sunken computer screen and keyboard.

The day began with an electronic brainstorming session. The meeting facilitator, an employee of the decision center whose primary task was to run the software, entered the question that would focus the group’s brainstorming efforts. The question was: How can we cut costs at NALC? Jim reminded the participants that they should not feel constrained. They needed a lot of ideas that could then be culled to a few good ones. For now, the team members should suspend judgment about which ideas were good ones.

The process started tentatively, with people typing in an idea and hitting a function key to enter that idea. Someone mentioned that she saw an idea on her screen that she didn’t enter. The facilitator explained that this is the essence of the electronic brainstorming process: people will see a random selection of their own and other peoples’ ideas on their screens that will help them think of other ideas. Suddenly everyone was typing in ideas. There were chuckles when funny ideas appeared on some screens and exclamations of “That’s a good idea!” and “Who thought of that one?”

In twenty minutes, the team produced more than eighty-five cost-cutting ideas.

Electronic brainstorming is a simple process. Individuals enter ideas on the topic at will. Regardless of whether the group is face to face or dispersed around the world, interacting simultaneously or over a period of days, members simply type an idea and hit the designated function key. This automatically disseminates the idea to all members of the group. Ideas are also accessed by hitting the same key, which gives members a random selection of ideas generated by the group. Members can continue to hit the function key to sample the full supply of ideas as well as to scroll up and down through the set. No one can tell whose ideas are whose because no identifying marks are attached to the ideas. Figure 2 shows an individual electronic brainstorming screen from the NALC session.

A few years ago, the vice president of financial services for Royal-LePage, a large Canadian real estate organization, brought in a team of regional finance directors to use the electronic brainstorming facilities at Queen’s University’s Executive Decision Centre. The team’s purpose was to generate ideas about the opportunities and problems that would result when the federal government imposed a new tax on goods and services. This is the kind of topic to which EBS is particularly suited; there are a large number of potentially useful ideas, only some of which are likely to occur to any one individual. Each team member began to enter ideas into the system. In less than half an hour, the seven-member team had generated forty-six ideas that were considered worth evaluating and refining. Parallel entry and anonymity are the main advantages that electronic brainstorming has over traditional brainstorming (see Table 2).5 Parallel entry means that individuals can immediately start producing ideas without the impatient attempts at interrupting that occur in traditional brainstorming groups. The sound of a hot electronic brainstorming group at work is a constant din of tapping keys interspersed with talking, exclamations, and occasional laughter.

The value of anonymity is most evident in groups that closely monitor each other’s ideas as a way of knowing what ideas are currently in favor. This seems particularly true of closed groups such as the staff of penal institutions. When groups from these organizations have used electronic brainstorming, they have sometimes initially shown frustration with not knowing whose ideas are whose. Once the participants become used to the anonymity, they begin to produce without fear of putting forward ideas that may be disfavored.

In our research, electronic brainstorming groups have consistently been more productive than traditional brainstorming groups that operate under the same structured rules. Across five studies involving more than eight hundred people, productivity advantages have ranged from 25 percent to 50 percent for four-person groups and to nearly 200 percent for twelve-person groups.6 And it is not only the number of ideas that is greater. Because so many more ideas are generated, the chance of producing more high-quality ideas is also greater. Our research has shown that the number of high-quality ideas generated by electronic brainstorming groups is higher than in traditional brainstorming groups.7 The results of a recent study in Canada and the United States are shown in Figure 3.8 The advantages of parallel entry, anonymity, and, no doubt, novelty serve to make electronic brainstorming a much more productive way to brainstorm, particularly among larger groups.

Some of our participants have found EBS to be a “colder” process than traditional brainstorming. Participants can talk if they want to but most choose not to because they are concentrating on entering ideas and reviewing ideas on their screens. Electronic brainstorming is more task oriented and lacks the richness of the oral medium.9 Nevertheless, the exchange and interplay of ideas makes the process a pleasant, satisfying, and sometimes even entertaining experience for most. Groups from many sites in North America, Europe, Australia, and New Zealand have consistently evaluated electronic brainstorming as more satisfying than traditional brainstorming. It isn’t so much the novelty of the electronic brainstorming process that people find exciting as it is the enhanced ability to express innovative and off-the-wall ideas without having to wait or worry about being criticized. Under the cover of anonymity, people will produce some surprising ideas.

Electronic brainstorming offers the additional advantage of not requiring face-to-face or real-time interaction. The group members do not have to be in the same place or even the same time zone to interact. As long as group members have computer-linked workstations and the requisite software, they can enter their ideas at any time. Over a period of days or weeks, they can repeatedly access the brainstorming program, enter new ideas, and sample ideas that have been or are being generated by others. This means that group members can brainstorm anonymously at their convenience with no concern about not getting enough air time or fitting into other people’s schedules.

Another advantage of electronic brainstorming is that group size is not a limiting factor. Traditional brainstorming groups usually are limited to fewer than a dozen people because it is difficult for all group members to get ideas out when the group is much larger. People are also more reluctant to express novel or different ideas in larger groups because there is a greater possibility that someone will criticize or be offended. With electronic brainstorming, group size is not a problem. As shown in Figure 3, larger group size is an advantage with EBS, not a disadvantage. Groups of up to fifty people have successfully used the electronic brainstorming process (the typical size of an electronic brainstorming group is twelve to fifteen people). This means that more people with varying perspectives and backgrounds can be brought into the brainstorming process, thereby increasing the chance of good ideas being generated.

Some technical expertise is necessary to set up the computer hardware and load the brainstorming software. In addition, some organizations that use electronic brainstorming on a regular basis, such as IBM, Texaco, and the IRS, have found that the use of a trained group facilitator has enhanced the productivity of EBS groups. The experienced facilitator combines knowledge of other creativity techniques with the capabilities of electronic brainstorming to produce as many good ideas as possible. The training of group members to use these programs usually takes less than five minutes and consists of explaining the rules of EBS and the location of the one function key that runs the system. As long as people can accept the idea of being anonymously creative, they can work well as an electronic brainstorming group.

Editing Ideas

“Now comes the hard part,” Jim thought. “How are we going to sort and evaluate all these ideas?” At the facilitator’s suggestion, the team decided to use the “idea organizer” software tool to put the ideas into categories so they would be easier to evaluate. Each person scanned the ideas and created categories. These were shown on the large screen at the front. As the team went through the list of ideas, removing duplicates, it became clear that there were some common, well-defined categories and that most participants agreed about how the ideas fit into them.

The process took quite a while (about an hour-and-a-half), but all felt that it helped them better understand the categories and the remaining ideas. They also began thinking about the merits of each idea.

The fun part of EBS is producing the ideas. Groups can produce dozens, even hundreds, of ideas. Inevitably some of them are redundant. The grueling part is editing out the redundant ones so that the group can proceed to the real work of evaluating the unique ideas. Most of the group brainstorming packages have a related program that provides a method for editing and evaluating the ideas. These programs allow for dispersed access and the same anonymity found in the idea-generation programs.

An example is the “idea organizer” program in Group-Systems. It has a simple sort feature that organizes ideas by key words identified by group members. Group members can then quickly combine or delete ideas that are similar to others. Another tool enables each group member to create his or her own categories for the ideas that have been generated. By scanning all the ideas generated, each group member can assign each idea to any number of categories. This process makes it easier for the group to identify similar ideas and combine or discard those that are redundant. This is the tool used by the NALC team in their session. Of course, one person’s redundant idea is another person’s subtle, but important, variant on a theme. We choose to err on the side of inclusion. If the participants can’t come to consensus about an idea, then it should be carried to the next stage.

Evaluating Ideas

The outcome of the “idea organizer” process was a list of forty-nine ideas divided into seven categories. Jim decided that it was time to evaluate the ideas. The regional claims manager suggested that the group evaluate the ideas one category at a time, and the group agreed. A group life manager suggested that within each category, each idea should be rated as to (1) the feasibility of implementing the idea and (2) the benefits to the organization of implementing the idea. Her suggestion was adopted, and the “alternative evaluation” tool was chosen.

Each person rated each idea in the first category on feasibility and benefits (using a ten-point scale). Then the software collected the ratings and displayed them for the group on the large screen. It was clear that the group thought some ideas were less feasible and had fewer benefits than others. The participants decided that two ideas were very good ones and should be implemented immediately. They went on to rate ideas in the rest of the categories (with a break for lunch) and selected six more very good ideas.

One software feature that can be used to evaluate a set of ideas is the “rank order voting” feature. It enables each participant to individually and anonymously prioritize the list of ideas. The computer combines these individual rankings into a group ranking that all members can see.

In New Zealand, the Decision Support Centre at the University of Auckland is being used by industry sector groups to generate ideas and plans that will enable their industries to become more competitive in global markets. These groups are composed of twenty-five business leaders, government planners, and politicians. Twelve of these one-day sessions, called the “Advantage New Zealand Meetings,” have been held so far. Meetings begin with participants entering ideas on what they see as the strengths and weaknesses of their industry. The voting tool is used to assist the group in reaching consensus on which of these ideas are the most important. Next, ideas are generated on initiatives that could capitalize on the industry’s strengths. The voting tool is again used to help the group select the five best initiatives, which are then developed into action plans. Follow-up on these sessions has shown that most of the action plans developed in these meetings are now in the process of being implemented or have already been implemented.

The voting tool allows any number of votes to be taken on a list of ideas. For example, a group might initially rank order a list of seven ideas. The resulting list shows that the bottom three ideas have little support from the group. These last three ideas can be deleted from the list, and another rank order vote can be taken on the remaining four ideas. This method can be used to generate consensus about the relative priority of ideas. Because group members enter their rankings without pressure, fear of criticism, or knowledge of who generated and supports which idea, the evaluation process can take place under conditions of equalized power. Members can immediately see how their rankings compare to the group’s and whether they are contributing to group consensus.

The anonymous evaluation of ideas is an important feature of the EBS technology. We have all had the experience of seeing ideas adopted because they were advocated by powerful individuals. Once this happens, a groupthink process can produce unanimity on what some (privately) believe to be a poor idea.10 EBS has the ability to undercut this process. Because no one knows whose ideas are whose, doubts can be expressed about ideas that may or may not originate with senior people, and enthusiasm can emerge for ideas that come from less powerful people.

All the EBS procedures are designed to equalize power within the group. But sometimes it may be worthwhile to have some people’s opinions carry more weight. On some topics, there are group members who are acknowledged to have better judgment than others; for some groups, there are members who are more crucial to the implementation process. EBS software makes it possible to provide differential weighting to members of the group at the evaluation stage. When this is done —and we acknowledge our misgivings about violating the power equalization feature of EBS — it should be made explicit that not everyone is equal in the group and why.

Another software tool is the “alternative idea evaluation” tool, which allows each group member to rate the ideas on a number of criteria. The assistant deputy minister of one of the Canadian government’s largest ministries wanted his senior staff to generate ideas and develop an action plan for promoting quality work and fostering excellence within the department. The eight-member team used electronic brainstorming to generate ideas, which were edited down to twelve good ones. They used the alternative idea evaluation tool to analyze them based on three criteria: (1) feasibility, (2) cost to the department, and (3) time needed to implement. Each team member entered ratings, and the system presented the group with the ranked list of ideas. The list was discussed at length, and three ideas were dropped. A consensus then emerged that, due to budget cutbacks, the cost criteria should be given greater weight. This resulted in a reordered list of ideas. Six months after this meeting, seven of the nine ideas had been implemented.

Implementing Ideas

Now the group had to flesh out the eight ideas and assign responsibilities for implementing them. The meeting facilitator suggested that the “topic commenter” tool be used. Jim agreed, and each idea became the title of one of the eight windows created in the topic commenter tool. For each idea window, team members entered their comments on how the idea should be implemented and any problems or concerns that might arise during implementation. The system gathered and displayed this information, and the group considered the ideas one at a time. To indicate responsibility, team members entered their names beside one or two of the ideas or beside certain implementation steps, along with completion dates.

The meeting ended around 4:30. Jim believed his team had eight good ideas that could be implemented. He knew the follow-up would be crucial but felt the team was committed and motivated to make the changes happen. He was impressed by how focused the session had been and how positive the team members felt about their contributions. He has already scheduled another EBS session in the center on marketing strategy.

Electronic brainstorming cannot do the implementing. But it can help by ensuring that many more ideas are thought of, expressed, and evaluated. This can encourage participation, build consensus, and make the implementation phase run better.

One tool that many electronic brainstorming programs have is a “plan-of-action” tool. This tool allows the group to generate action steps for implementing the best ideas. In GroupSystems, the topic commenter tool creates a “window” for each idea (see Figure 4). Group members enter their action steps for each idea in the respective idea windows. Any window can be opened by any group member at any time for the entry of action steps. Everyone can see all the action steps entered for any idea. Once the process of action step collection has been completed, the group can proceed, one window at a time, to refine a plan of action for each idea.

The group leaves the session with a set of high-quality ideas, the action steps to implement them, and clearly identified responsibilities. Group members feel that their ideas have been heard and that the best ideas have been identified and prepared for implementation. The process has been satisfying and focused for most participants in EBS sessions. Table 3 summarizes the advantages of electronic brainstorming.

Exploring the Possibilities

The use of the computer to assist the brainstorming process opens up a number of possibilities to make the creative process more effective. A study done at the University of Arizona showed that the productivity of groups could be raised or lowered when the computer technology was used to enter anonymous supportive or critical comments about ideas.11 Productivity rose when anonymous, critical comments were entered. Productivity was not as high when anonymous, supportive comments were entered because group members did not feel as motivated to generate more and better ideas. At the University of Auckland in New Zealand, brainstorming software is being designed to automatically enter supportive or critical comments to increase productivity.

In one medium-sized service organization, electronic brainstorming has been used to facilitate discussions of critical issues facing the organization and its clients. The organization thought of itself as open to ideas from its professional staff. However, we noticed in a meeting that several senior professionals referred to difficulties felt mostly by newer, lower-status members and then discounted them. Electronic brainstorming sessions were held with junior members in which they identified several very serious, but previously undisclosed, problems regarding professional practices. A summary was circulated, and subsequent EBS sessions provided the opportunity for a significant portion of the more senior professionals and some clients to acknowledge these problems and identify others. These sessions in turn prompted most staff to participate in further sessions, out of which came a task force, additional anonymous data gathering, and a set of consensual ideas for action. These have helped focus the organization, providing a stronger sense of community and increasing satisfaction among both staff and clients. The controversial and largely undisclosed nature of the issues (e.g., unprofessional conduct, inconsistent practices), the status differences within the organization (e.g., long-term supervisors, contract members), and the espoused values of openness made the use of electronic brainstorming ideal for surfacing issues in an anonymous and thoughtful way.

Another possibility is to integrate individual brainstorming software with group brainstorming software. Individual brainstorming software allows the user to define the topic area in a variety of ways, which then enables the software to generate idea phrases on that topic. “Idea fisher” is a personal creativity program that, when given key words and phrases describing the topic, attempts to generate words and ideas that will stimulate the user. We are working on integrating this sophisticated personal creativity software into group brainstorming software. The individual software will be like another “person” generating ideas. This mix of human- and machine-generated ideas may further increase the productivity of group electronic brainstorming.

Another approach that is being explored is to have the software provide automatic feedback to individual team members. Team members can see, on their own screens, how the number of ideas they have generated compares to the average number generated by the group. The intent is to stimulate higher productivity from those team members who tend to free ride.

The selection and evaluation of a large number of ideas can be a difficult process. A number of software developers are currently working on expert systems that will reduce this effort. One approach is to have the software automatically generate categories for the ideas based on keywords. The ideas could then be automatically inserted into the categories and duplicates removed.

One of the disadvantages of traditional, oral brainstorming is that the ideas generated in one session often go unused in future sessions. Even though ideas may have been recorded on sheets of paper or flipcharts, they are not usually in a form that can be used easily in the future. In EBS, the ideas are saved on computer disk and can be loaded easily for future brainstorming sessions. These past ideas can be thought of as “seeding” ideas for the next session.

Finally, there are over thirty manual techniques that have been developed to increase the productivity of the brainstorming process.12 An example is the “Superman/ Superwoman” technique (Imagine you are Superman/Superwoman. What ideas do you have?). These techniques could be stored within the brainstorming software as “creativity templates” along with instructions, specific screens, and brainstorming questions. The most appropriate template could be loaded and used, depending on the brainstorming topic and the team composition. All of these ideas and techniques are being developed so that the basic process can be extended and improved.

The Down Side

All techniques have drawbacks, and EBS is no exception. We will describe seven, which are summarized in Table 4. Perhaps the most worrisome is that EBS may be oversold. It’s always tempting to think that a new technology will solve old problems. In the case of EBS, it is clear that people have to have ideas in order to take advantage of the technology. The technology can’t substitute for them. Once the ideas are generated, judgment is involved in editing and evaluating them, and the effective use of the system depends on trust. The software also can’t help much with actually implementing the ideas. Once an idea has been selected and an action plan developed, it remains the responsibility of team members to do all the things necessary to implement the ideas, including building support, marshaling resources, and taking action. Thus the superiority of EBS is dependent on its users.

A second problem is that EBS requires people to use the keyboard. People who are unaccustomed to typing, such as some senior managers, can get frustrated. But although practitioners of hunt and peck are likely to produce fewer ideas than they would if they talked, most group members we have worked with don’t find their typing speed to be a significant problem. This is probably a transitional problem as keyboard skills become more widespread and voice recognition technology and pen-based computing become more sophisticated and accessible.

Third, the anonymity is a drawback for some. Those used to having their ideas accepted because of who they are can suffer in the evaluation stages. As a result, high-status individuals may have a more difficult time getting their ideas accepted than they would using traditional brainstorming. These individuals experience a loss of control in EBS groups that would not happen in traditional brainstorming groups. This disadvantage is, of course, the mirror image of the advantage enjoyed by lower-status members.

A fourth drawback is that, although most find the process more satisfying than verbal brainstorming, some miss the social aspects of the traditional process and the richness of spoken language. There are no prohibitions against talking, but most group members don’t talk in order to concentrate on idea generation. This, for some, is a disadvantage because it reduces the opportunity for personal interaction. One partial response to this absence is that people can comment on the ideas they see on the screen, laugh, and otherwise react to the ideas that are generated. As no two people are looking at the same set of ideas, it also introduces some useful uncertainty about what is going on. If talking works for a group, there’s little reason to discourage it.

Another drawback is that too many ideas may be generated, resulting in idea overload. This is particularly true for the idea pools generated by large groups. Although people don’t find this to be a problem when they’re generating ideas, some people become frustrated by the effort required to edit, select, and evaluate so many ideas. This problem should be reduced as the software tools for editing and evaluating ideas become more useful and sophisticated. Nevertheless, it is the part of the process in which groups most often bog down, and this will probably continue to be true.

Sixth, an EBS team requires networked computer hardware and software. This technology is still relatively expensive, especially for small firms, although the cost is rapidly shrinking. Several organizations such as IBM and Texaco (as well as many universities) have EBS centers that can be used by outside groups. As the price comes down, more organizations will establish their own facilities.

Finally, some problems lend themselves better to brainstorming, both electronic and traditional. Brainstorming seems most suited to problems where people have many ideas but accessing them is a problem. Topics such as how to cut costs, how to attract new business, and what problems people are experiencing in doing their jobs are the types of topics for which EBS is designed. Choosing a new president lends itself less well because the task requires close scrutiny of an already identified set of choices.

Thus EBS does have drawbacks, but they are less serious than those that have plagued oral brainstorming, some are being reduced by current research, and all are manageable.

Conclusion

Traditional face-to-face oral brainstorming is widely used to generate ideas. Despite its drawbacks, it will continue to be used. We also expect to see electronic brainstorming gradually grow in use. This will happen because it is new and will spread, as with all fads. Its enduring value is that it helps groups generate, edit, evaluate, and plan to implement their ideas on topics where it’s difficult to get people to share their ideas openly.