Photo by Chris Blakely on Flickr.
This week at the Dish conference in Rotterdam I gave a presentation about all the do’s and don’ts, tips and tricks, lessons and hands-on advice about crowdsourcing from my experience at the Museum of National History. Well… that’s quite a lot to talk about. All in all I came up with some 25-30 little notes, which the audience of my presentation – in a little participatory trick – had to label as do’s or don’ts.
Here’s the full list, now all as do’s, with some additional ideas that didn’t fit in the presentation. Use it to your benefit and please add your thoughts when you feel I’ve missed some.
- Ask your potential participants a clear question or a clear task. A clear question is never ambiguous, unless you’re looking for (and only looking for) different ways to look at its ambiguity.
- Run a couple of real-life test sessions with your question. Even if it’s an online project, ask people in the street your question and see how they respond. Change the question all the time. Once people only respond with the answers you’re looking for, you’ve found your question.
- Ask a question that is meaningful to people. Questions that might be labelled emotional or highly personal are good. Not everybody will answer them, but the answers you’ll get will be so much more valuable.
- Pinpoint very specific groups of people you’d like to reach with your project. Design to meet their demands and answer to their needs. Preferably, involve this target group in the design of your project.
- That said: don’t exclude anyone from participating if they really want to.
- Be extremely clear about your limits to what people can contribute, and keep these as limited as possible. Racism, hate, advertising and unlawful things are usually enough to exclude.
- Accept all other contributions, regardless of they way in which you perceive their quality. Every time a person took the trouble to contribute to your project, this contribution is valuable (you can use peer reviewing to maintain overall high quality).
- Do not put limits on the length of a contribution, unless these limits are an explainable part of the project. Also, don’t limit the number of contributions per visitor, but design your system in a way that it gives all contributions equal importance.
- Quick win: even unwanted contributions can tell you something, so when moderating, don’t delete these, but make them invisible. Some of the best things I discovered about how to design good projects I learned from stuff that happened by accident.
- Never ever fake contributions, not even the first 10-20 to get the project going. Even though you think you might be a great ghostwriter/actor/impersonator, many will see through it immediately. Use your own name if you want to contribute.
- That said: there’s nothing wrong with asking your inner circle to be the first to participate. Encourage colleagues to join in and share, tell your friends. All crowdsourcing is incrowdsourcing in a way, so you better have this work to your advantage.
- Furthermore, don’t use VIPs to get your project underway. Next to the life’s story of Richard Branson or the photography of Anton Corbijn, my contribution will be so small I better not even contribute. In my experience VIPs that are not raised by the community serving the project will have a negative impact on your project.
- If you want discussion around contributions, specifically ask for discussion or design your system in such a way that new contributions show up near related ones. In my experience, people prefer to present their thoughts as a new contribution, not as a reaction to another contribution.
- Design a straightforward process to contributing, preferably as simple as possible and integrated in the normal things people do in your institution.
- Think about the different steps of the process (making contact, getting people interested, engaging them and making them enthusiastic) and make sure all these are well designed and work towards your goal.
- Furthermore, to your contributors, divide the project in clearly distinguishable phases. For instance, first everybody can tag words, then people can review earlier contributions to select the best and then final description is uploaded to your collections database.
- Crowdsourcing is not only about participation. It’s just as important to reach people with your project (PR, marketing, etc.), to continuously improve and redesign your project (project management) and to evaluate and report on the project. Focus on these elements as well.
- Design for different types of participation. Not everybody is a creator (deciphering words, telling stories), some people prefer commenting or rating or collecting. Make sure your project caters to these different needs and gives a place to everybody.
- Make participation almost invisible, for instance by making it part of the normal stuff people do in your institution or by counting and measuring things they very easily do (like taking guidebooks or making photos).
- Look for ways to merge participation in the digital and physical world. In my experience the best crowdsourcing projects seamlessly integrate online and offline, focusing on the objective rather than the choice of medium.
- Create a safe environment for people to contribute. Make it somewhat private, but also special. For instance, in my experience voting and selecting works best when it is kept individually and small, whereas creating (once done) deserves a bit of an audience.
- Don’t focus on beautiful websites and wonderful installations too much, focus on highly functional design. Some of the best crowdsourcing projects I’ve seen where made on a shoestring budget with stuff that was lying around.
- Celebrate contributions with contributors. All contributions are special, and everyone who dared to contribute is a hero, so openly celebrate contributions, and:
- Give credit to the contributors. Unless the idea behind your project is different, make sure you overdo the amount of credit you give contributors. Keep their names next to the photos forever, and use their contributions in future publications of which you send them a free copy (with credit line), invite them to openings and special tours.
- Because, always, always give feedback on the results and process. Keep people informed about everything. What happened with their contribution? What is going on with the project? And not just after 6 months, but all the time. Keep them in the loop.
- Have fun and make your crowdsourcing project fun. Share optimistic stories with the people that participate, focus on the small successes, share the unique things that happen, and,
- Allow for participants to have fun. Even if you’re mostly looking for serious contributions, the contribution that is a bit off but makes you smile is always one of the best ones you will get. (Sometimes this smile is a tear.)
- If you think the joy of participating and your continuous feedback isn’t enough, only give away relevant prizes. iPads are cool, but hardly ever relevant.
- Involve the people on the floor in your galleries in the project. Quite often, a human voice explaining the project or making people enthusiastic is the best tool to get people to participate.
- Continuously evaluate your crowdsourcing project and make sure you have budget to make changes once you’ve started. As a rule of the thumb I use that in participatory projects, only 1/3rd of resources should be spend on the launch version, and 2/3rds should be saved for improvements and iterations.
These 30 things might not all be applicable to your future project, and probably there’re countless examples of projects done differently that were successful anyway. And that’s cool, because, well, as number 31 I should probably add, “You know your audience best, so work with them in designing a project rather than with expensive consultants.” (You can ask a consultant to help you with that.)
What have I missed? Please add your ideas to the comments and I’ll add the best ones to this post.