10 Commandments of Effective Crowdsourcing

By: Ehsan Ehsani

For the companies which have embraced the crowdsourcing mindset in their business processes, the motive is more than just outsourcing. It’s about better collaboration, better innovation outcomes and ultimately superior value. But like many other new business models, some fail and some succeed in accomplishing this mission.

The central idea behind this piece is to compile a practical list of tips which can help executives make their crowdsourcing initiatives work and get the most out of them: Some of these ideas are new; but to a surprising degree the most important ones are the old things, that we probably tend to forget as we are bombarded by countless applications and ideas in the crowdsourcing space.

These tips are based on the result of a multi-year study and discussions with some of my colleagues in management consulting industry, academic friends and executives who have had a positive or negative experience with crowdsourcing. They also cover observations of more than 50 regular winner-takes-all challenges as well as online multi-stage competitions which I refer to as “innovation tournaments”. I hope this will become a starting point for our online community to share and discuss their views and enrich our mutual understanding of the emerging best practices for running effective crowdsourcing initiatives.

So without further ado, here comes the 10 commandments:

1. Make sure the nature of your problem is crowd-friendly:

Companies face many challenges and these challenges can certainly be crowdsourced. But that doesn’t mean the crowd can solve all those problems on time. There is an inherent uncertainty associated with crowdsourcing in the sense that companies don’t know if they are going to get an answer by the end of their campaign.

Therefore, it’s better for mission critical and extremely difficult issues to work with a smaller group or a  “controlled crowd” or alternatively have a plan B in place in case the crowd can’t come up with a solution on time.

2. Define your problem at the right level:

Companies can decompose an issue into smaller and more abstract problems to a level digestible by the crowd.

At its raw form, crowd might not be able to solve certain problems. Think about the optimization of a sub-system software for a satellite which might look scary for many participants in a crowdsourcing community. However, companies can decompose an issue into smaller and more abstract problems to a level digestible by the crowd. In case of the satellite example, this can be a math problem which in essence addresses the optimization challenge.

3. Make sure high performers are definitely involved:

The number of solvers participating in a crowdsourcing effort matters: It is certainly a good thing to generate a high number of ideas, but the quality of the ideas coming out of the challenge is even more important. These quality ideas typically come from high performers and experienced experts. Since in many settings, the high quality solutions are the only ones which are being picked up by the company, it is essential to involve high performers to increase the likelihood of high quality idea generation.

4. Link your problem to real value:

Often time companies launch crowdsourcing campaigns just to generate ideas. The result if a long list of ideas which are not used at all. This is a complain I have heard from many executives – They don’t like crowdsourcing campaigns to be done just for the sake of doing it.  Ideally the problem at the heart of the campaign should be aligned with issues facing the company. These problem typically come from non-innovation executives and if the focus is on such issues, the solutions will be more likely to be implemented by them.

5. Assign the right monetary prize:

The participants in a crowdsourcing initiatives should be motivated and a monetary prize helps with achieving such a goal. However, if the prize too low, the crowd might not be motivated enough to participate. Too high and people might be scared to participate as they might think they don’t have a chance: They might think the risk and reward ratio is not favorable and since the effort required is too high, the challenge is not worth participating.

Based on previous experience, it seems that mid-term initiatives with a range of $30,000 to $50,000 yield a good result. Additionally, bundling such prize figures with a quality bonus makes sure people focus more on the quality of the outcome.

6. Pick the right duration for your initiative:

A challenge period of 4-8 weeks typically seems to yield better results.

Ideally we want the participants in a community to work and propose solutions throughout the challenge period. If the duration of the challenge is too long, participants typically postpone the work and contribute in the last minute in an ad hoc way. On the other hand, if the duration is too short, there won’t be enough time to participate. A challenge period of 4-8 weeks typically seems to yield better results.

7. Look where the knowledge is:

The crowd of the website InnoCentive is very different from the crowd of Kaggle. Which one is better? It depends on the problem you have in hand and what you want to get out of your initiative. It’s always good to do a bit of reflection or research to see where the knowledge resides and focus the initiative on the right community.

8. Maybe we should look inside:

The common definition of crowdsourcing is related to work done with an external crowd. In large global companies, there is another crowd as well: The employees working across the globe. They can be a source of tremendous knowledge for addressing certain issues. Extending the concept of crowdsourcing to internal operations, a company can sometimes benefit from making the open call internally. Two great example of this are Danone, the French dairy company and its global knowledge sharing initiative and IBM Jam which orchestrates the insights of tech company’s employees across all continents.

9. Make it part of the day-to-day activities:

Running a successful crowdsourcing campaign is only part of the equation. The knowledge of the crowd should be absorbed by employees to make an impact. The success in adoption depends on whether people have to use it to get the job done or the platforms and initiatives are disconnected from day-to-day activities. Unless, crowdsourcing is an integral part of day to day activity, it won’t have an optimal effect. Running such initiatives regularly (making them a habit) or making it part of the workflow in business processes might improve the integration.

10. Measure the value and experiment with improvements:

What can’t be measured, can’t be improved. Techniques like Balanced Scorecard, common in strategy field, can be used to link the impact of such initiatives to bottom line. This way, the organization can learn what works and what doesn’t and can improve through a learning process.

By Ehsan Ehsani

Photo credit: Miss C.J. / IWoman / CC BY-NC-ND