Kevin Cheng  

Brainstorming vs. Idea Seeding: The New OK/Cancel Process

October 21st, 2005 by Kevin Cheng :: see related comic

It’s been over a year since we’ve written any features and one in particular strikes me as terribly out dated. [The Making of OK/Cancel][1] was an article that detailed both our writing and art process for a given strip. If you’ve never read it, I highly recommend you do so first because it’s somewhat entertaining but more because it’s actually relevant to what I’m about to discuss.

A lot of times, brainstorming is seen as a positive addition to the design process. In the early stages, it’s great for generating ideas of what would make a good product or feature. Even when you do know what you’re building, designing alone can lead to stagnant ideas or getting stuck in loops that might be pretty easy to break out of with a touch of outside perspective. So one would think that for creating something like a comic, a group brainstorm would be a great way to kick off the writing process. Such was not the case.

##Our New Process
Our [old process][1] of writing a comic every Thursday was rather time intensive - usually spanning 2-4 hours of IM conversations debating the topic we should cover and how best to extract humor from a given topic. Recognizing the huge inefficiency inherent in the process, Tom adopted a creative process he sometimes [employs in music][2] to our comic writing.

This process was dubbed switch sessions and involves two collaborators blocking out a set amount of time for the project. In our case, each “session” comprised of 3 half hour sessions. That’s 90 minutes within which we were completely dedicated to the creation of quality comics.

In the first session, each of us takes a spreadsheet template we created and enters ideas for 4 comics. The template is set up to facilitate quickly entering standard 3 panel comics, each panel on the template has room for caption, setting, dialogue and special effects. This session is “closed” in that no communication occurs between us.

Script Template

In the second session, we swap and refine or complete the other’s comic scripts. This session is also closed. Finally, in the third session, we get our original comics back and open the communication channels. This last half hour usually involves signing off on comics that are already sufficiently polished, clarifying what the objective was for some more ambiguous comics and more refinement for those the more difficult scripts. Often, we end up discarding one of the comic ideas altogether when we can’t massage it to a sufficiently humorous state.

##Seeding vs. Storming
Altogether, we end up with about 7 quality comics in the span of 90 minutes. This normally would have taken us 14-28 hours to accomplish under the old system. A marked improvement to say the least. What’s surprising is that the quality of our comics, at least in our opinions, has been at least on par if not better than the comics written using the old system.

While the name of our process is “switch sessions”, I would call the more general approach to creative collaboration __idea seeding__. Instead of starting from a blank slate and working together to come up with a final product, we document some idea seeds. These have numerous advantages. Firstly, as the name suggests, a seed helps to give some direction to the discussion. Even if the initial idea for the comic is bad or incomplete, it may establish enough of a seed to spark ideas from others.

Unlike group brainstorms, there is also certain pressure within the sessions to produce quality ideas when you know that your respected peers are going to be looking at them. Initially when Tom suggested this process to me, I thought the target of 4 comics in half an hour was less than realistic. As it turns out, I’ve come up with some of my best ideas under this artificial deadline and even sometimes have an overflow of ideas that I reserve for future sessions.

I’ve often wondered what it might be like to apply this process to other more formal design processes such as wireframe generation or even visual design directions. We do use switch sessions to do initial sketches of the written scripts occasionally but this is less practised especially as both Tom and I have gotten faster at our respective art tasks.

Would something like, “come up with three interaction solutions for this problem in half an hour and I will do the same and then we swap,” work better than getting in a room with a team to discuss how best to solve a specific problem?

Maybe we should brainstorm some ideas to improve idea generation …

[1]:http://www.ok-cancel.com/archives/feature/2004/01/the-making-of-okcancel.html “The Making of OK/Cancel”
[2]:http://www.switchsessions.org/ “Switch Sessions”

12 Responses to “Brainstorming vs. Idea Seeding: The New OK/Cancel Process”
ct wrote:

Yo, Kurioso. Love the description of your process as well as the documentation that illustrates it.

But I must say this idea seeding thing is very much like what I would consider a proper brainstorm.

The problem with brainstorming is not that it’s deficient; it’s just that it’s done deficiently.

What most people consider “brainstorming” (imagine my index and middle fingers on both raised hands crooking here) is usually just a poorly-done debate session with personal agendas and egos going to battle.

What I understand to be proper brainstorming is just this: Idea generation, with the emphasis on generation. There are no stops to consider feasibility, no “buts”, no “Nos”, no pre-judging, no voting, no arguments. Like the improv classes in which I saw brainstorming in its most beautiful and powerful manifestation, it’s a bunch of “Yes … and”s, all building on top of each other, crazy as the last one might sound. The only goal should be to come up with X ideas in Y time, in response to a clearly-defined challenge or set of challenges. All the voting, editing and decision-making comes later.

Which is what your idea seedings with Tom look like.

It’s exhilarating when it’s done right. I just wish more people knew how to do it. I wish there were more people who knew how to play.

Nathan wrote:

This seems to be a variant of the brainstorming approach where firstly people generate their own ideas (each person scribbling out on a notepad) and then we go around the room discussion/extending each others list of suggestions.

The new step (of having to extend someone else’s suggestion before the general discussion) is a good one, I think — it reduces discussion time (because you’re already engaged with each others’ ideas) and improves the quality of discussion.

I think I’d like to try this on a wiki…

Joe wrote:

> …consider “brainstorming” (imagine my index and middle fingers on both raised hands crooking here) …
That’s a good one. First air-quotes to represent written quotes in verbal communication, now written quotes to represent air-quotes…

Bob Salmon wrote:

I think many schoolchildren do session switching. Take a big strip of paper and write down a boy’s name. Fold the paper over to hide the name and pass it on, write down a girl’s name, fold and pass, write a place etc. Gets a bit silly.

Ash Donaldson wrote:

I like the idea of ’switching sessions’ as it builds upon what is known of idea generation - that is, that individuals are able to generate more ideas and can then refine or build upon each other’s. I use something similar (but have never given it a name) for greenfields projects and creative sessions.

Park-Gates’ PhD dissertation in 2001 was entitled “Effects of group interactive brainstorming on creativity” and is an interesting read on the effectiveness of brainstorming.
http://scholar.lib.vt.edu/theses/available/etd-08092001-104543/unrestricted/FinalCopyDissertation.pdf

The dissertation relays that, like many other techniques employed, brainstorming has long been known to be an inefficient way of generating creative ideas. It’s refreshing to see someone challenging such a technique.

Stephen wrote:

I like the fact that this technique produces ideas which aren’t designed by committee.

Instead, we’re looking at two simultaneous author/editor feedback loops:

A:design->B:feedback->A:design->B:feedback->Joint:finaliz
B:design->A:feedback->B:design->A:feedback->Joint:finalize

Original ideas, with twice the output in the same amount of time — sounds like a great process!

Tom Chi wrote:

To give credit where it’s due, the process was originally design so that Dave Peck and Tom Chi could write music:

www.switchsessions.org

It definitely does not sound like music by committee.

Kevin D wrote:

I have found that brainstorming, when done correctly has many benefits. While the inefficiency of brainstorming can be debated, the side benefits of team building and idea ownership can greatly enhance the chances of a particular idea to grow to fruitation. If a design team brainstorms an issue and brings it to a product team it has value, but get a couple of devs and test in there and suddenly prototypes fall out of the air and example code is written. I have seen this time and time again. Most of the time the passion of the people in the room makes the difference. I have also seen brainstorming go bad, and one negative node can really pull a session out of whack and create a stagnant idea arena much worse than your comic, which is just an agenda driven brainstorm session. Thats why I insist on external facilitation of brainstorm sessions.

Guy wrote:

The initial seeding session is an especially great aspect of this approach. I’ve found that often the best ideas come from trying to come up with what might seem like too many ideas in too little time. While you may end up with a bunch of duds, since they are only the seeds of ideas and you’ve got a bunch of them, you’re bound to have a few great ones. Brainstorming is supposed to facilitate this, but often the group aspect hinders an individual’s creative flow.

This does leave the question, how does idea switching work with larger groups of 5 or 6?

Vilius B wrote:

Hi,

this switching technique is sometimes is called “Brainwriting” (in contrast to brainstorming loud). It makes working environment more silent (=> more space for concentration, esp. if people are tired).

Nice article and comments!

Greets, Vilius Benetis
Board of European Students of Technology (BEST)
http://www.BEST.eu.org

Brent wrote:

I recently ran a company innovation retreat and started it with giving people 7 minutes to give a presentation on up to 3 ideas for new products. The presentations were in front of all attendees, and the ideas presented formulated the discussions for the next two days. People put a lot of effort into their ideas, and we ended up with a great variety of well-considered possibilities to kick-off the group “brainstorming.”

I wish I had read this before that meeting, I would have added a second stage with this switching concept before moving onto group discussions.

t wrote:

sounds similar to the surrealists’ technique of exquisite corpse — perhaps other of their approaches would be useful to you as well…


Leave a Reply


OK/Cancel is a comic strip collaboration co-written and co-illustrated by Kevin Cheng and Tom Chi. Our subject matter focuses on interfaces, good and bad and the people behind the industry of building interfaces - usability specialists, interaction designers, human-computer interaction (HCI) experts, industrial designers, etc. (Who Links Here) ?