All Your Ideas ± Lessons from a Crowdsourcing Experiment & Game
For this year¶s Craigslist Foundation Boot Camp, we expanded on past efforts to include the community in what Boot Camp should include. We decided to invite our community ± past attendees, newsletter recipients, twitter followers ± anyone and everyone who cared to participate, to share what THEY wanted Boot Camp to be this year. Thank you to the hundreds of you who participated in this experiment. We received tons of great ideas, learned a lot in the process, and heard from many of you who want to know how we did it
and what it was like. (a word cloud of ideas for Boot Camp) What We Did The Craigslist Foundation team held a brainstorming session around topics for Boot Camp ± what worked best in the past, emerging topics and presenters that we¶d seen or heard or heard of, and community challenges to address. We then took our ideas and pared them to phrases that complete the sentence ³boot campers will experience, participate in or learn«´ The second half of this sentence (ie, ³advanced social media skills´ or ³a discussion about building blocks for building communities´) was then updloaded into All Our Ideas, a free and fantastic crowdsourcing game
led by Matthew Salganik from the Dept. of Sociology at Princeton University. What All Our Ideas Is About
All Our Ideas describes themselves as: ³a research project to develop a new form of social data collection that combines the best features of quantitative and qualitative methods. Using the power of the web, we are creating a data collection tool that has the scale, speed, and quantification of a survey while still allowing for new information to ³bubble up´ from respondents as happens in interviews, participant observation, and focus groups.´ I first heard of All Our Ideas from my colleague, Arthur Coddington. We would refer to it as ³the kitten war thing,´ because that was the first application ± and a memorable one! ± we had seen of this type of game. You have to check kittenwar.com out for yourself, but be prepared to disappear down a productivity-sucking hole into a land of cute kitten photos. Kittenwar and All Our Ideas function on the same basic principals ± you get to vote for two comparable things, side by side, and the tool aggregates the votes over time. All Our Ideas is an open source platform. (Kittenwar is not, and is an unrelated commercial site. But it serves as a wickedly cute example!) With All Our Ideas, your participants can vote for one of two ideas over and over again. They can also add their own ideas (each is tweetably short ± 140 characters) to the voting mix or indicate they can¶t decide.
Here¶s an example from the Washington Post of the scores from an All Our Ideas poll of the most wanted Christmas gifts of 2010: All Our Ideas explains that a gift, kitten, idea, etc. is given a score that ³is the estimated chance that it will win against a randomly chosen idea. For example, a score of 100 means the idea is predicted to win every time and a score of 0 means the idea is predicted to lose every time.´ Setting up your own game on All Our Ideas is easy, and their blog has great tips for success. What We Learned We had our ³game´ live for a little more than a month. It¶s actually still active now, but we stated that we¶d use the data that was provided until February 15. Here are ten observations about our results, and some ideas about what works: 1. What our community wants to learn and experience matches what we aim to provide at Boot Camp. 2. Strong themes emerged. Related topics bubble up to the top ± sometimes four or five related topics were all popular. Top themes for what people want out of Boot Camp include workshops and sessions on fundraising, social media, collaboration, working with government and across sectors, storytelling 3. Hundreds of people participated. More than 3,000 votes were cast by more than 400 people 4. Lots of ideas came from the community. We ³seeded´ the game with 98 Boot Camp ideas. 55 additional ideas were contributed by the community. 5. Our crowd is savvy. Of the top-ranked ideas (I counted all that ranked 60% or higher in the ³top,´ for my purposes), almost half
were participant submitted. AND they were relevant, timely, and appropriate. 6. Trust the tool. Some participants added ideas that were almost identical to ones we had seeded. This worried me, but it was actually really useful, because several related ideas all wound up being ranked similarly, and strengthened a single idea into a theme. 7. It¶s easy and fun to use. I didn¶t do very well in statistics in college. But this was easy to analyze. The data almost sang to me ² ³storytelling, storytelling, storytelling ideas!´ and ³collaboration, facilitation, conflict resolution, collaboration tools!´ 8. Don¶t worry about how many votes come from one or many people. Many people didn¶t vote more than once. I wonder if our instructions or the tools¶ directions need refining to encourage repeat voting. On the flip side, some people voted many, many times. I briefly worried they¶d skew the results, but I would check every few days, and no one gamed the game, it appears. 9. Promote broadly. We asked friends and followers to retweet, share and help promote this. That was necessary to get the votes we did. 10. Share as you go. As themes emerged and participants submitted ideas, I tweeted what I was seeing. This generated even more interest and retweets. We will definitely continue crowdsourcing event content, and I loved using All Our Ideas. In the near future, I¶ll post about the final session results, as well as the steps I took to turn the crowdsourced data into actionable offline content for Boot Camp.