Crowdsourcing has been judged to be one of the most impactful, but least used
digital hacks by organizations by Gartner. Why is that? Well, it’s an emerging discipline and there’s a lot to learn about how to do it well – how to communicate, how to moderate discussion, how to provide intrinsic value, and more. And one of the questions that we receive every so often is how to correct for voting bias. Sometimes new ideas don’t get as much support as ideas that have been in the community for a long time, sometimes popular ideas are more a reflection of the popularity of the idea author than the merits of the idea, and other complexities. So here are some suggestions that you can use in your IdeaScale community to avoid voting bias in crowdsourcing.
1. Don’t use voting. Use one of IdeaScale’s other evaluation methods: five star-ratings, pairwise comparison, etc. You can invite everyone to participate in those reviews or you can restrict those reviews to particular groups. Voting isn’t the only way to gather crowd feedback.
2. Use a separate stage for voting. This doesn’t always correct against popularity, but definitely corrects for recency bias and can sometimes correct for popularity contests since idea submission and promotion are separate tasks. First launch a stage where ideas are just submitted (voting turned off) and then move all those ideas into a time-limited stage for voting (as in voting can only take place between this date and this date) so people have a limited time period for getting in there to support their friends’ ideas.
3. Use the fund stage. You can give everyone, specific groups, or individuals a budget of tokens (aka votes). You can also set goals for how many tokens an idea has to receive in order to be considered. You can even automate the software so that when an idea reaches that goal it automatically moves to the next stage for review or team-building, etc.