An Experiment in Crowd-sourcing

futurelab default header

by: Matt Rhodes

I had my first real experience of using crowd-sourcing to solve a business problem last week. We’re working on the design and concept for an online community that will be public later in the summer. The community is going to be covering an exciting topic and one that should appeal to a broad range of the target audience. What’s even better is the client is enthusiastic about sponsoring a community about the issues, rather than it having to be overtly product-led.

  • As somebody looking for a name, I describe what I want named and pay $99. The project is then open for 48 hours.
  • As somebody suggesting names, I can just enter a name (including an explanation if desired) and it is then added to the short list. All users of the site earn points and can use these points to ‘invest’ in names they think might win.
  • When the 48 hours are complete, the site calculates the best name based on these investments and a bit of maths. The person who suggested the top three names earn money, as do the people who invested in them. In total $80 of these $99 are paid out (although users are only actually paid when they have earned more than $50, so I would imagine many of these rewards are never cashed in).

The process is simple and quite exciting. Within minutes of putting up our brief, we’d had a handful of suggestions, and in the full 48 hours 401 different names were nominated. Okay, so some of them weren’t quite right, and a few would suggest a website of a rather different nature, but we got more names than we might have got if we’d stayed in that room and brainstormed for an hour. And because the names came from over 300 different people, there was real variety there.

You can’t see on the site which are the more popular names, so any ‘investments’ you make are purely because you personally think the name is a good idea. So the winning names should be the ones that the majority of people who contributed to the project, independently, think are good names. So I was excited about the results.

When the results came I think we quickly decided we wouldn’t be going with the ‘winning’ name. It’s not a bad name, it just isn’t right for us. The second-placed name was one we couldn’t understand (was it a person’s name? was it a place?); and the third a Greek Goddess. So we probably won’t go with any of these. But of the 401 names that were suggested there are some really good ones and we’ll be suggesting a couple of these to our client.

So, what did I learn from my experience? First, namethis appears to have a large number of users from the US, so the names suggested often seem ones that might work better in that consumer market than in the UK. A second learning would be how important it is to get your brief right. This can, of course be difficult. In a real-world brainstorming scenario, you tend to refine and revise the brief as the ideas come in. You have a two-way dialogue that wasn’t possible on namethis. This might have helped, I could have said that some ideas were great but wouldn’t be quite right for us and then given reasons. This would have helped subsequent users to suggest or invest in different names.

But was it worth it? In terms of the quantity of names suggested I think so. There are some great ideas that I know we would never have come up with. And for $99 it was probably cheaper than the opportunity cost of four of us spending even half an hour brainstorming.