Your Brainstorming Sessions Suck? Four Drivers of Success and Where the Web Helps


photo credit: Freshview

ReadWriteWeb’s Bernard Lunn penned a piece recently, How Can Web Tech Help Enterprises with Innovation Management? The post argues that Innovation 3.0 will include a large dollop of external idea generation via social media. And companies are hungry for it:

There is no more important an issue on the agenda of top management than driving innovation.

Bernard’s point about the hunger for innovation is right. It’s always been an imperative, but the ease with which companies can come from anywhere to disrupt markets has raised the importance of new ideas.

While his post focuses on externally sourced ideas, I’d like to talk about what’s happening inside organizations. Specifically, I’m talking brainstorming. Those “have-the-potential-to-be-fun” meetings with your colleagues to work on the next generation of your companies products and services. But they’re not always fun are they?

Turns out there are some new services that can improve the way companies’ employees generate top-notch ideas. Before hitting those, there’s an intriguing study by some INSEAD and University of Pennsylvania researchers that sheds light on where brainstorming can be improved.

The Four Drivers of the Best Ideas

In Idea Generation and the Quality of the Best Idea, the academics examine the different processes that  lead to the best ideas (not just good ones) being generated and selected in brainstorming.

Their hypotheses are interesting. Here are the four drivers for getting the best ideas from brainstorming:

  1. The sheer volume of ideas generated
  2. Average quality of all ideas generated by brainstorming
  3. The amount of variance in the quality of generated ideas
  4. Ability to evaluate what the best ideas are

These are covered below, including which styles of brainstorming are better. Before that though, here’s a little more on how these academics arrived at their conclusions.

First, they studied existing literature. A must for any researcher. They then created their own field study, using Wharton students in brainstorming experiments. Now whether that fairly models company brainstorming…let’s see what they found, eh?

Two styles of brainstorming were analyzed:


Hybrid: Individuals went through their own personal brainstorming exercise, then met as a team and generating more ideas.

Team: All idea generated occurred in a group setting.

Turns out, it makes a difference which brainstorming style is used. This is discussed below.

Volume of Ideas Generated

This one ranks up there with motherhood and apple pie. The more ideas you generate, the higher probability you have of generating outstanding ideas.

On this metric, the Hybrid approach is much more productive. When individuals sit down and come up with ideas by themselves, they produce more ideas than what a group typically generates. The primary difference in quantity of ideas generated is termed “production blocking”. Production blocking is the inability to generate ideas when others in the team are speaking.

Important to note here is that there is a time restriction in this metric. As in, “# ideas per hour” type of a metric. The more time given to a brainstorming session, the less difference there is in quantity of ideas generated between the two brainstorming styles.

Average Quality of Ideas

This one is little less intuitive. The Hybrid process generates ideas of higher average quality than does the team process. I can’t say what I expected, but hearing this was a bit surprising.

Seemingly, the ability of the group to refine an idea generated during brainstorming would ultimately raise the overall quality. But it seems that individuals have pretty good internal regulators. I’d guess we actively suppress the worst of the ideas, or those that we’re not so sure about.

Thus we raise the overall quality. But in doing so, do individuals snuff out potentially high value ideas?

Variance in the Quality of Ideas

This is the one that will probably surprise you. To get the highest ranked, the best ideas, you want a higher variance in the quality of ideas generated. That means more really crappy ideas, along with some truly inspired ideas. In terms of a statistical distribution, think of it as more ideas extending to the extreme left and right tails of a population quality.

Turns out, Team has the edge here over the Hybrid approach. As the study authors say, “we believe there is more potential for both breakdown and collaborative success in teams then in individual idea generation.”

What an interesting statement! On the downside, the variance comes from poor group dynamics inside that brainstorming conference room. Have you ever been in that situation? I have. A bunch of elephants coming together into a room, with existing political connections, and the result is a really bad session with few ideas of middlin’ quality. Because of these qualities, some idea gains currency among the group, and discussing that idea becomes the theme of the meeting.

And it feels like you just wasted an hour or two. Frustrating.

But Team brainstorming also has its high points. When the team comes together without agenda, and brings a serendipitous variety of viewpoints. People feed off one another, and imperfections in one idea are overcome with different thinking from someone else’s idea. These brainstorming sessions are gold, and incredibly valuable when they happen.

One thing to take away from this. When considering brainstorming in your workplace, have an honest assessment about your company’s culture. Can people really come in and have an idea jam? Or will things inevitably get mired in the same old agendas and relationships to reduce brainstorming effectiveness?

Evaluation of Ideas

This is the final step, and it’s where a lot of brainstorming sessions fall short. How good is the team in evaluating the quality of the ideas generated? According to the research of the academics, the Team approach is less effective in evaluating idea quality than the Hybrid approach.

To ascertain the “true” quality of ideas in the Wharton student experiments, the researchers had an independent panel of people rate the ideas that came out of the brainstorming sessions. They then compared these independent ratings to the self-evaluated ratings of the different teams.

This quote from the paper actually made me laugh a little:

We find that the ranks obtained in the Team process have no correlation with the panel ratings whereas for the Hybrid process they exhibit a significant positive correlation.

No correlation for how the Team approach evaluated the ideas. My fellow workers of the world, does that ring a little true to you? The researchers ascribe this finding as supporting “the theory that in a team, ownership of ideas, social pressures, team dynamics and interaction of different personalities limit objectivity and the ability to discern quality.”

In the Hybrid process, there were actually two idea rating events: individuals rated their own self-generated ideas, and the team evaluated its ideas. The researchers found that the individuals rating their own ideas was the primary reason for the correlation of self-evaluated ratings’ high correlation with the panel ratings.

Turns out we’re pretty good at discerning idea quality when we’re free from the group setting. Even for our own ideas.

Implication for Using the Web to Improve Innovation

The academics findings lead them to this conclusion:

These results suggest that it would be best to employ team processes in the idea generation stage and then use an independent individual evaluation process.

I’m going to disagree somewhat with the first part of that statement. I’d say a mix of individual and team processes is best. Inside an organization, there will be plenty of times where you as an individual will have an idea. Some of those individually-generated ideas will be top-notch. And there will be times where a Team approach will be employeec. See what ideas come from those sessions. As I said before, just be mindful of your existing company culture in term of the quality of ideas that come from these sessions.

The second part of the reseachers’ conclusion is where Enterprise 2.0 comes in. Once ideas are generated – whether individually or in a Team process – they need to go through an evaluation process. Google employs a form of this with its internal prediction markets.

There are a couple companies out there who are working in part of the Enterprise 2.0 space:

Spigit: Spigit provides InnovationSpigit, which has a prediction market orientation to evaluating ideas.

BrightIdea: Brightidea provides Pipeline, which has a project management orientation to idea management.

There may other companies out there as well. The point is to recognize that we employees are imperfect.  Increasing the visibility and accessibility of ideas and independent evaluations is a great way to bring structure and a diversity of opinions to bear on ideas. Remember, companies are hungry for innovation.


See this post on FriendFeed:


About Hutch Carpenter
Chief Scientist Revolution Credit

3 Responses to Your Brainstorming Sessions Suck? Four Drivers of Success and Where the Web Helps

  1. Thanks for the tips, I will use them in the next meeting!!!

  2. Pingback: HyperBored » Blog Archive » Link Distractions 02/05/2009

  3. Josh says:

    What if the president of your company is a micromanager and brainstorming meetings end up being not about what the team is brainstorming but more about what the president wants?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: