Crowdsourcing is very trendy these days and is touted as the answer to many of the ills of poor design and the need to reduce costs. In these cash strapped days any way to make innovation better-cheaper-faster is extremely desirable.

But crowdsourcing is just one of the many tools we have at our disposal, and each tool is suited to particular kinds of applications. To simply adopt an idea like this without considering its suitability to the problem domain or to the desired results can be risky.

To assist with critical thinking about crowdsourcing I have collected a few alternative viewpoints & list five reasons why it might not always be the best approach to adopt. Please note I do not agree with everything in the articles linked below — they are meant as a thought starter & to provide different perspectives on crowdsourcing (i.e. if you've got any issues with the articles please contact the author directly).

Since no single tool is the answer in all cases, here are a few times when crowdsourcing might not be the right solution:

1. When the crowd does not have sufficient understanding or knowledge

For crowdsourcing to work you need to find the right crowd.  If the technical or scientific knowledge required is rare then crowdsourcing might not be helpful unless you can find a crowd of people with the requisite foundational knowledge.

2. Where the problem is diffuse and complex

Crowdsourcing lends itself to solving clearly focused problems where there is little ambiguity or nuance — a great recent example of this was the DARPA balloon challenge.

For diffuse and complex problems it might be necessary to chunk up the challenge (if that is possible). And for problems that require painstaking layering of knowledge and information with long term focus it might not be commercially viable.

A good example of this is the discovery of longitude via crowdsourcing in the 18th century. It worked in the long run, but it took a really long time and was funded by the government. However, it might be argued that this kind of discovery would be much quicker today with computer power.

3. When you want to keep your plans secret

Clearly secrecy requires that only a few people know the secret. Thus crowdsourcing something that is meant to be a secret is probably a bad idea (unless you are executing a cunning hide in plain sight sort of plan).

4. Your problem needs to be compelling enough for contributors to care

Experience of Wikipedia indicates that people will contribute to things that are interesting to them. Thus if nobody cares about solving your problem then crowdsourcing might not be the answer.

To get an idea of how crowdsourcing works on an everyday basis there is a good discussion of how Wikipedia contributions happen by Henry Blodget in: Who The Hell Writes Wikipedia, Anyway?.

There is also a well known report by Forrester about Social Technographics that segments the participation of people within social networks. It shows that only a small proportion of people create or share content, a few active creators or editors, with the bulk of people lurking or not participating at all.

5. Crowdsourcing for complex problems requires dedicated resources

To undertake the kind of knowledge work required to solve complex problems contributors need uninterrupted time in the zone.

This is exemplified in some of the large open source software projects where companies pay people to work full time on open source projects for commercial advantage:

Many of the leaders of key projects (like Guido van Rossum, the inventor of Python, who works at Google (nasdaq: GOOG — news — people )) are paid by their employers to continue to lead their projects. Is there an open source community? Of course there is. But on the most prominent projects, the members of the community have jobs and are paid to work on open source because the software is so beneficial to their employers, even though it is not owned by them. True, there are hybrid models, and the smaller the project, the more likely it is unfunded. But when it becomes a big deal, open source becomes commercial.


Link to original post