Since the launch of Innocentive back in 2001, the application of open innovation has spread out from its scientific routes to encompass a growing number of fields. The success of those forays have been somewhat mixed however, so I thought it might be useful to explore in more detail the kind of areas that open innovation is well suited to.
A good place to start is the work of Ivan Dale Steiner. He distinguished between what he saw the two main types of workplace tasks:
- Disjunctive tasks
- Conjunctive tasks
The distinction is a crucial one when it comes to crowdsourcing. Disjunctive tasks can be commonly regarded as those for which only one member of the group needs to succeed for the group as a whole to succeed. So many of the challenges posted onto Innocentive for instance could be regarded as disjunctive tasks.
Conjunctive tasks on the other hand require a good contribution from every member of the group for the group to succeed. So building a bridge for instance is a conjunctive task as every member of the team has to contribute.
So how does this matter for crowdsourcing? Well, in disjunctive tasks, the diversity of opinion that is so often a strength of open innovation can come to the fore because multiple approaches can be taken to solving the problem, with that one successful idea then benefiting the whole community.
Conjunctive tasks however often aren’t so conducive to open innovation because it requires people working collectively on a task, and this is very hard to achieve, especially when the various participants in the project are so distributed.
Indeed, it could be argued that the principle challenge of any organisation with distributed resources is to convert conjunctive tasks into disjunctive ones. This is something that the open source movement have done very well by turning complex software development into a collection of modular and disjunctive parts.
It is only when this occurs that the vast diversity of the crowd participating in the project can bring their benefits to the project. So, if you’re looking to make a success of crowdsourcing innovation, make sure your task is disjunctive.
I couldn't disagree more! The so called wisdom of the crowd is the result of synergy within the crowd. That is: the result of interaction within the crowd. I think that traditional forms of collaboration don't work well in a crowd. The ones that requires centralized control of the process. The success of for example Q&A tools such as Quora or StackOverflow shows that social collaboration is perfectly possible in a crowdsourcing process. Those tools are very suited to apply to an open innovation process.
See also: http://blog.mobbr.com/2014/01/crowdsourcing-in-20…
That runs counter to the rules established by James Surowiecki. He was quite clear that for crowds to be wise, they each have to be working independently of other members of the crowd. This was likewise established by people such as Scott Page in his book The Difference.
The only way open innovation has worked for larger style projects is where the tasks themselves have been turned into a series of smaller conjunctive ones. The open source movement is a good example. Each module can be worked on by lots of people, but by isolating those efforts to those modules, centralised co-ordination can make it work together.
I guess it depends on your definition of collaboration and of independent. Take for instance the traffic system. Millions of people go on the road individually each day in a mass collaboration effort. Yet, no central control. It's called stigmergy or platform communication.
That's kind of it though isn't it? If I'm driving somewhere, I don't need other drivers to help me get there. I need them to not crash in to me, but that's not quite the same 🙂 Each motorist on the road is acting independently of the others, and there isn't a shared goal amongst the motorists in the sense that they're all trying to get to the same location.
Let me begin with stating that this kind of discussion is very useful for the field of crowdsourcing and social collaboration. Because they are concepts that mean very different things to different people.
The crux here I guess is the difference between working independently and working isolated. The drivers in a traffic system do provide signals to eachother, so in that sense they collaborate. For people on a platform to collaborate their individual goals need not to be the shared goals.
This actually is what synergy is all about: the whole has properties or functions that cannot be attributed to one of its individual parts. Same for 'the wisdom of the crowd'. Its wisdom cannot be attributed to one of its individuals.
These emergent properties or functions are the result of the fact that the individuals on a platform form a whole. If the individuals on a platform are isolated and centrally managed, the question is to what extent they form a whole. In other words, how likely it is for them to create synergy?
For example, suppose you organize an open innovation challenge and 100 individuals send their answer. If you just pick the best answer, then you've not created synergy. You've just mined the crowd for the best answer. If you let them comment and rate eachothers answer, you will get more robust results, as the contributors in the crowd can build upon eachothers expertise.
Suppose it can work either way can't it? Threadless for instance allow, and no doubt benefit, from having people vote on the submissions of the crowd. Indeed, that's their entire business model.
Innocentive by contrast never do this, relying instead upon the sponsor to judge the best answer. The X Prize (and similar challenges) tend to rely more on the first entrant to achieve a certain result, and whilst there are instances of entrants sharing progress (NetFlix for instance), often they are working independently of each other.
Even when feedback is given however, that feedback isn't a requirement of a successful outcome for the competition. The task can just as easily be done by someone working without feedback from others.
Conjunctive tasks, by contrast, cannot be done without input from other people. This is why projects such as the Hyperloop, which are hoping to use crowdsourcing, have broken down the big task into much smaller modules, so that each module can be done without requiring the kind of management tasks associated with large and complex projects. If they'd put the design and construction of the entire Hyperloop project out to the crowd without breaking it down, then I don't think it would work well at all.
(quite agree though, enjoying the discussion here – very useful and interesting – thank you for participating)