Using computers to facilitate brainstorming sessions has been (occassionally) talked about in academic and management fields for the last two decades, but I have never heard of it actually implemented in a real-life scenario. Back in 2008, I blogged about the potential and problems of electronic brainstorming, but I haven't encountered the topic subsequently.
Until last month, when I was invited to participate in an electronic brainstorming session for my faculty's IT department. Granted, this project was in an academic setting and was part of a research study, otherwise it was a real organizational project. The goal was to help provide ideas and prioritization for future IT projects.
Description of Electronic Brainstorming Session
The software used was SAP's Streamwork, which from a user perspective was easy to use and worked flawlessly. It probably wouldn't be difficult, however, to build such a system or make use of existing online tools to do this as the functionality is fairly simple.
The first part of the electronic brainstorming session involved using a web-based interface to generate ideas. Participants were instructed to add as many ideas as they wanted to. The software was a standard web form that allowed one to enter free-form text.
The second part of the session occurred a couple of weeks later. Using the same software, we were then asked to prioritize everyone's items. Duplicates were manually filtered out by the project coordinator, but I believe no other contributions were censored (certainly none of mine were altered). All items appeared as unordered group with no indication of the contributor. There were approximately 40 items which we then individually dragged and dropped each item and placed from top to bottom to reflect our priority order.
Pros & Cons
Volume of idea generation
In real-life scenarios there are time limits and group dynamics that make it difficult for everyone to offer their ideas. One of the benefits of electronic brainstorming is that participation can be opened up to a much larger number of participants than is feasibly possible in real-life environments. The project I participated in seemed to have a small number of people participating (i.e. under 20), but this seems a deliberate intention to use a Delphi Decision-Making process, which uses selective recruitment. If there is no management requirement to participate or incentive offered - it is possible that only the usual suspects of keeners and malcontents will volunteer. Increasing volume is not necessarily a great idea, as when it came time to prioritize I found the 40 items I had to examine was my limit of what I could keep track of and handle. Nonetheless, I have been in real-life brainstorming sessions where after a few initial ideas, it was painfully difficult to get contributions. So by opening up participation to a larger and more geographically dispersed group (if applicable) and allowing people to participate when and where they choose, it is likely that a greater number of contributions will result.
Duplication of ideas
In real-life group brainstorming sessions, participants will have seen and heard others' contributions, so there is little concern of duplication of ideas, as there is with electronic brainstorming. This can be a serious limitation as I witnessed with the Canadian government's Digital Economy Consultation last year. The government asked citizens to generate and vote on ideas. But there was so much repetition and overlap - and no means to filter or merge - that it undermined the ability to read all the items let alone vote meaningfully on specific ideas. (Not that I believe the government had a sincere desire to harvest these ideas generated anyway.) I'm not aware of a good way to automatically purge or merge duplicates. In the case of the project I participated in, someone manually had to do this. For large-scale projects this could be time-consuming, but the benefits to users are substantial.
Value of anonymity
Anonymity is a central concept of electronic brainstorming (although not a system requirement). I believe it is essential to allow free participation. Without anonymity people may not want to offer ideas for fear of being judged or challenging organizational sacred cows. Thus real-life sessions tend to be dominated by a few people, and in my experience generate ideas that are fairly safe and banal. When it comes to prioritzing ideas, anonymity is even more important as it is difficult to give a low priority to ideas from one's supervisors or befriended co-workers.
In contrast electronic brainstorming, there is no fear of real or perceived judgment or reprisals. I felt like I could say whatever I wanted to free of repercussions and could also offer ideas that I would otherwise have felt were too trivial or personally relevant. Anonymity has its downsides in that it does allow people to offer offensive ideas or unhelpful, possibly insane, rants (as per trolls and flaming on open online forums). But in a workplace setting, this hopefully won't be an issue. All the 40 ideas I encountered were valid and coherent.
Quality of ideas
Although the contributions were legit, there were some ideas that needed clarification or persuasion. There was no way to follow up with an idea originator for more information, thus the ideas must stand on their own - and some didn't. As anonymity was promised, I'm not sure if even the project facilitators were able to request a contributor to refine a contribution. It would be possible for a system to allow a facilitator to receive all contributions unidentified and then submit clarification requests via the system and consequently not identify the person. In real-life sessions, it is easy to ask someone to explain or defend their idea. If the goal is only idea generation, then this step might not be important. But when it comes time to prioritize ideas, it is essential that participants understand what they are voting on. Perhaps a step could be built into the process allowing participants or facilitators to ask for clarification prior to the prioritization.
My biggest complaint with the electronic brainstorming session I participated also applies to real-life scenarios and that is that there is no mechanism to indicate which items are to be followed up on. An ideal step of any brainstorming project would be for every item to have an official response from management on their plans on how to address it or why they are unable to. Such transparency would show participants that their input is truly valued and useful to the organization.
I feel electronic brainstorming was a better method for me. In my experience, real-life meetings tend to be dominated by a handful of people or a lot of time is wasted with group dynamics (e.g. turn-taking, social niceties, idle chit-chat). This approach not only afforded me more opportunity for me to share my ideas, it was also more time efficient.