Want to design the next great web app? Upgrade your product, but can't decide what to add or change? Add a new feature to your product, but can't decide how to implement it? Forget focus groups. Forget endless meetings and brainstorming sessions. Throw an ultra-rapid-design party, and do it in a single day. This approach exploits the wisdom-of-crowds through a process of enforced idea diversity and voting, so no consensus, committe, or even agreement is needed. And it's way more fun.
The Product Design Dinner Party takes 9 people, a pile of diverse "inputs", and has each of the 9 people voting on--and pitching--one another's ideas to continuously reconfigured groups of 3 people, letting the best ideas rise to the top. The process is a little complicated, but it's derived/modified from an existing rapid-prototyping design I'll talk about later in the post.
The basic idea looks like this, although there are a million ways to modify it:
1) Pick 9 people, ideally from different parts of your company and including some customers. (If you don't have a company yet, pick 9 friends--preferably those who don't know each other well)
2) Buy/borrow/find at least 20 "input materials" including books, magazines, a short film, graphic novels, etc. (a list of possibilities is a little lower in this post)
3) Assign (randomly) at least 2 "inputs" to each person. Do NOT let them choose (it's important they not be allowed to gravitate toward things they're already comfortable with)
4) Give the group 30 minutes to generate 4 ideas (if it's a feature/upgrade party, then 4 different features or feature sets... if it's a feature implementation party, then 4 different ways to implement the already-decided feature, etc.) These 4 ideas don't have to come directly from their input materials, although participants should be highly encouraged to describe at least one new thing they learned that inspired their idea.
5) Round One begins: split into 3 groups of 3 people (see chart below). Each person gets no more than 10 minutes to "pitch" four ideas to the other two in their group. There are 12 total ideas for this group, so allow about 30 minutes. Record (anonymously) the selections of each person, which represent a "vote" for the ideas.
6) At the end of Round One, each person must select their two favorite ideas from each of the other two members of their group. So if Group One had Fred, Mary, and Sue... then Fred must select his two favorite ideas from the four that Mary pitched, and his two favorites that Sue pitched.
7) Round Two begins: reconfigure the groups so that each person is now with different people (see chart below). Instead of pitching their own four ideas, each person pitches the four ideas they chose from their previous group members. Again, they have about 10 minutes to pitch the four ideas. Remember, the point is that each person is no longer pitching their own ideas!
8) At the end of Round Two, each person must again select their two favorite ideas from each of the other two members of this new group. Record (anonymously) the selections of each person, which represent a "vote" for the ideas.
9) Round Three begins: reconfigure the groups again. Each person in the group now pitches the four ideas (two from each of the two members of their most recent group) they chose in the previous (Round Two) round.
10) At this point, each person has pitched a total of 12 ideas:
* Round One: pitch your own four ideas
* Round Two: pitch four ideas from your Round One group to your new Round Two group -- two ideas from each of your previous group's other members.
* Round Three: pitch four ideas from your Round Two group to your new Round Three group, as before.
11) At the end of Round Three, again each person selects their top two favorite ideas from the ones pitched by the other two members. Record these as a vote.
12) You should now have a total of 108 votes. Choose the top 9 vote-getters (you'll have to be creative about tie-breaking... you could choose more than 9, for example).
13) Give each person a copy of the 9 ideas, and send them back for another round of "inputs." Again, assign each person different materials from the ones they used at the beginning.
14) Give the participants 30 minutes to use their inputs and flesh out a single idea from the nine. Their one idea can be a modified version of one of the nine, based on their "research." Their one idea could be a mashup of two or more of the nine ideas. It cannot, however, be something completely new. Participants should be prepared to explain how something they got from their inputs helped in some way (not an absolute requirement).
15) Now it's up to you what to do with the ideas. You might choose just one, or take all 9 "winners" with their pitches back to another person or group, etc.
Group Configurations (just an idea to get you started):
While it might be hard to believe a process like this could lead to any useful ideas, it's actually derived from a well-desiged, heavily-field-tested rapid-prototyping/development process from one of the leading training consultants on the planet, Thiagi. Granted, that doesn't mean my modifications haven't completely messed it up, but the main goals and benefits of doing it this way are:
1) Time constraints
Constraint-fueled creativity is something we've talked about earlier, so I won't discuss it here.
Build something cool in 24 hours
Creativity on Speed
How to make something amazing right now
and a little in Don't wait for the muse
2) Forced lack of attachment
By having to pitch someone else's ideas instead of your own (after Round One), it keeps people from getting stuck/married/attached to their own idea.
3) Random, outside-your-domain inputs
By having to use pre-selected (and pre-assigned) materials from outside your domain, participants have a better chance for a diversity-driven inspiration.
The whole thing is based on the assumption that you have all the knowledge you need -- the wisdom within your own company and your customers... you just need a way to tap into it that doesn't dilute the idea (as design-by-consensus would do) or prevent innovation (as design-by-listening-to-customers would do).
Ideas for "input materials"
Books on a wide range of topics outside your domain including architecture, astronomy, pop culture, filmmaking, comic books, wedding planning, education, children's book, romance-novel-writing, crafts magazine, travel book, sports, history, environment, etc.
If it's a software product, you might assign people to look at a variety of pre-chosen sites or web apps that are way outside your domain.
I've used this in the training world -- as a tool for learners to help come up with what they ought to be learning, but I've never used it in the way I've described here. I'm looking forward to trying it...
(And yes, I took a little artistic license with the photo at the top--pizza and coke might be better than alcohol. Then again...)
I'd love to hear ideas for modifying this, or from anyone who's done anything like this!