Crowdsourcing Moves Beyond Open Innovation

Point: Crowdsourcing is maturing beyond its amateur-content and open innovation origins toward core business processes.


In the beginning, companies used crowdsourcing as part of their open innovation efforts to get new ideas from lead users, customers, and the world at large.  But now, entrepreneurial companies such as Trada and CloudCrowd are moving beyond one-off design efforts and contests (e.g., the Netflix Prize) to encompass routine everyday business processes.  As CloudCrowd CEO Alex Edelstein sees it, “similar to the way Henry Ford’s early assembly lines created a new, more efficient way to complete work, we’ve designed an online process that delivers accurate finished work for even complex projects at a significant savings.”  Let’s look at these two examples.

First, Trada Inc., which recently emerged from stealth mode private beta, offers crowds of pay-per-click experts who create paid-search marketing campaigns.  Each vetted crowd member generates his/her own keywords, ad copy, and deep links to attract prospective pay-per-click customers to the client site.  The result is a much broader span of keywords with less chance of overpaying for over-used common keywords.  By giving access to the long tail of keywords, Trada executes campaigns at lower cost and with greater success than do traditional agencies with in-house employees.

Second, CloudCrowd has 18,000 registered workers who participate in its Labor-as-a-Service business. CloudCrowd’s project managers begin by breaking down a complex task into hundreds or thousands of smaller tasks. These tasks are then passed on to Cloudcrowd’s registered workers. CloudCrowd speeds delivery time and lowers costs in a wide range of BPO (Business Process Outsourcing) applications.  Tasks that CloudCrowd has deconstructed include tracking the University of Southern California’s “lost alumni” and a wide range of web content creation tasks.

Both Trada and CloudCrowd eschew the winner-take-all model of contest-oriented crowdsourcing projects.  Instead, they offer well-defined incremental pay for incremental results.  In the case of Trada, an expert gets paid for each click-through of the ad that the expert created (Trada also offers pay-per-sale crowdsourced campaigns).  CloudCrowd gives its workers a pre-agreed payment for each unit of work they successfully complete.  In CloudCrowd’s case, a worker’s “success” is measured using a system of escalating peer reviews that are also crowdsourced.

Both Trada and CloudCrowd create carefully-cultivated crowds — more like reliable workforces than mobs of transient volunteers of dubious quality.  Trada uses online testing and verified identities to ensure that its experts are really experts.  CloudCrowd assesses each worker’s percentage of correctly-completed tasks to compute a Credibility Rating.  Highly-rated workers gain access to higher-level, higher-paying tasks.


  • Evaluate which business processes might benefit from on-going outside expertise or labor
  • Create clear tasks and clear rewards
  • Create processes to vet or rate prospective crowd members on expertise or quality
  • Use the crowd to monitor the crowd


Personal interviews with Niel Robertson, CEO, Trada

and Cloudcrowd (via email)

9 Comments »Innovation, open innovation, Social Media, Software tool

9 Responses to “Crowdsourcing Moves Beyond Open Innovation”

  1. Calvin May 25th 2010 at 01:59 pm 1

    It is interesting that CloudCrowd and Trada turn away from contest yet 99Designs continues to use contests. Because 99Designs has been so successful, perhaps the other companies just decided they could not be successful by copying 99Designs.

    I think it would be interesting to compare these companies to oDesk and elance. Is the difference that CloudCrowd and Trada create the small jobs rather than the clients?

  2. Andrea Meyer May 25th 2010 at 02:34 pm 2

    Interesting question, Calvin, thanks!

    To answer your question, these companies are really using crowdsourcing is a very different way and for different applications. First, 99designs is in a different space and market, so there’s no competition between them and Trada and CloudCrowd.

    Second, the biggest difference is that 99designs is trying to find the one best design. With, the second-best design, 3rd best, 4th best, etc. doesn’t provide much incremental value even if they are good designs. In contrast, Trada and CloudCrowd create crowd tasks such that the second-best crowd member, 3rd best, 4th best, etc. still create significant value for the client and get paid for their effort. Consider the case of Trada: Trada is not trying to find the one best keyword for a paid search campaign. Rather, Trada is trying to find several hundred good keywords. Thus all of Trada’s different search experts add value to the client’s campaign by creating lots of different keywords. Both Trada and 99designs leverage the parallel search powers of crowds but Trada seeks to find as many good keywords as possible whereas 99designs seeks to find the one best design possible.

    Some criticize crowdsourcing contests by calling them exploitsourcing (see ) because so many members put in so much labor but only or a few members of the crowd reap the rewards. That criticism seems a little harsh to me, because contestants understand that design contests are just that — contests that reward the best of the best and pay zero to the rest. In contrast, companies like Trada and CloudCrowd show that crowdsourcing can be done differently, such that the labor of most members of the crowd can create real value for the client and lead to payments for most of the members.

  3. Rick Ross May 25th 2010 at 11:09 pm 3

    It’s exciting to see that this concept, with such tremendous potential for businesses, is beginning to mature.

    I found the article and your response to Calvin very valuable and informative. Thanks Andrea!

  4. Andrea Meyer May 26th 2010 at 08:37 am 4

    Thanks, Rick, for your kind words!

  5. Nick Boucart May 27th 2010 at 07:10 am 5

    A service like Cloudcrowd reminds me of Amazon’s Mechanical Turk ( I wonder how it relates or differs… I have the feeling we will see more of those “microtask” platforms, like or and probably a lot of others.

    Services like Trada seem to take this one step further, to create a real niche ecosystem around a certain task (in the case of Trada online ad campaigns) where experience and creativity is of higher value. In doing so, you can leverage the experience of others.

    Interesting article, thanks!

  6. Andrea Meyer May 27th 2010 at 07:54 am 6

    Excellent comment, Nick! I agree that Mechanical Turk is like CloudCrowd in doing crowd-completed microtasks. If there’s a difference between the two, it’s that CloudCrowd does all the task decomposition, workflow, task dispatch, layered quality control, etc. for the client. Mechanical Turk is more of an API and crowd host that assumes the client has already created a well-formed set of microtasks. And, as you correctly noticed, Trada provides a much more focused application and a vetted crowd of experts (in Trada’s case, PPC experts) . All the models have their uses, and I agree with you that will see more of this in the future. Thanks!

  7. What We're Reading This Week | Trada Blog: Paid Search Marketing, Online Advertising and Small Biz May 28th 2010 at 10:24 am 7

    […] If you didn’t see it, our CEO Niel Robertson (@nielr1) posted a guest post on the WordStream blog on Lifetime Value vs. CPA. The always incredible Andrea Meyer (@andreameyer), innovation expert extraordinaire, wrote about Trada in her post Crowdsourcing Moves Beyond Open Innovation. […]

  8. Crowdsourcing and Open Innovation and Knowledge-Oh My! « Iknow Iknow Iknow Jun 3rd 2010 at 12:18 pm 8

    […] June 3, 2010 · Leave a Comment Crowdsourcing Moves Beyond Open Innovation | Working Knowledge ®. […]

  9. Joe Raimondo Jun 4th 2010 at 06:16 am 9

    I posted a link to this at my new new blog, My comment was basically that there seems to be a sudden and dramatic swell in interest in this area. We have big corporate clients using these crowdsoruce tools as the “last mile” solution for data authentication after running things through automated means. Very interesting.