Emerging Views

a blog for the 2.0 Council
Tags >> contest

Cisco Systems and IBM                 

have experimented successfully with crowdsourcing, with enough common results to suggest trends, and enough differences to spark further investigation.  The Cisco story (fall 2007) was the subject of a recent Harvard Business Review article, and the IBM experience (summer and fall 2006) was treated in the Sloan Management Review.   Both articles are recommended if your firm is considering a large scale crowdsourced innovation project.   

 

Do not underestimate the sheer complexity and large time demands on senior management, as both Cisco and IBM seem to have done.

 

The Cisco project was an external (open to public) innovation contest with $250,000. cash prize to the single winner.   The objective was "to find an idea that would spawn a new billion-dollar Cisco business."  Twenty-five hundred innovators participated,  putting forward 1,200 separate ideas.  

 

 The IBM Innovation Jam was somewhat more restricted in reach, open to employees, families, business partners, clients (67 firms), and university researchers.   Their objective was to speed the development of existing IBM technologies from a list of 25 clusters and create new businesses of substantial size.  It was not a case of a single winner, but Chairman Palmisano promised $100 million in funding for up to ten new business units.   No cash prize was mentioned, although normal IBM IPRs and rewards may have been in operation.  The total number of participants was said to be 150,000 and tens of thousands of postings were generated in a three day nonstop Jam.

 

Beyond the design and scope differences, the IBM Innovation Jam revealed a few behaviors that vary from the Cisco experience. 

 

IBM Jam was

 

  • not so collaborative: IBMers did not build upon ideas of others, to such an extent that the Second Round of the process has been dropped from their model.   Cisco on the other hand found 70% of the final 40 ideas did benefit from teams pooling ideas
  • Why the difference?   Good question.   Perhaps the cash prize came into the practical decision to combine forces with others who had valuable bits to add to your own.
  • largely confirming existing processes: most ideas had already been raised to IBM management, but there was still seen to be added benefit to IBM for the visibility, recombination, and refinement of ideas in the Jam round of examination

 

Similarities: both Cisco and IBM found

  • idea generation is “the easy part” ... the fun part
  • management process is “not cheap” ... it is ponderous, time-consuming to filter, guide, develop biz case, and assess ideas
  • hybrid human-machine processes are needed in filtering
  • voting is not a reliable filter, the most attention grabbing ideas ("promote votes") were not necessarily the most robust, practical, or business compatible; Comments were much more valuable, in part because the expertise of the commenter would become apparent
  • live, interactive meetings are required with contributors in final steps, either face-to-face (IBM) or Virtual Presence (Cisco)

 

Unlike small group and face-to-face brainstorming, there was little sense of excitement, the discovery-in-the-moment.  Rather, it was in both cases a drawn-out period of filtering, analysis, guidance and training loops in how ideas translate to business cases, and repeat refinement.  The process was described as ponderous, and somewhat painful in sheer time and complexity.

 

Neither Cisco nor IBM saw crowdsourcing as an exercise in workplace democracy.   Decision making on the value of ideas, and the winnowing process down to the winners, was firmly in the control of senior management.

Cost to download pdf of Cisco story in HBR (Reprint R0909C) is $6.95, same as for IBM Jam in MIT Sloan Mgmt Review (Reprint No. 50101).  Links are given above.

 

Does your company use crowdsourcing?  Just how far can the software take you, before you need human intervention?   Is it valuable enough, over the obvious huge time demands, that senior management would like to do it again?    Once every __ how many years?   Do you use monetary incentives?   share royalty rights?   Or just go with the reputation effect and "bragging rights?"    Leave a comment.