Crowdsourcing is an interesting phenomena that works, but it is not obvious how does it work in details. This post is about crowdsourcing, when it works and when it doesn’t.
The Evidence: Crowdsourcing Works
We all benefited from successful crowdsourcing already, I’m sure WordPress backend is running on Linux servers; Linux is probably the most iconic product of the crowd. Linux allowed to cut hosting costs tremendously. Without Linux we would not have the Cloud. Free BSD was also suitable, but reality is that Cloud was enabled by Linux and LAMP, when you pay for the hardware and don’t pay for the software. Scaling became pretty cheap. Amazon, Google, dotCloud, Rackspace – all of them – have to say ‘thank you, the crowd’ – for building Linux. MS has to eat own dog food, run Azure on Windows without selling Windows licenses anymore… Otherwise they can’t compete with Linux-powered clouds.
The Evidence: Crowdsourcing Doesn’t Work
Try or recall of the usage of OpenOffice or other crowdsourced (and open sourced) apps and we will be on the same page. Many crowdsources apps suck. There is no clue how to polish them and bring to the better quality.
Why Linux rocks then? Because it is not a crowdsourcing. It is open source but it is controlled design and development. There is huge authority known as Linus Torvalds, who is a driver, designer, manager, inspirator, cop and God. Without such person there would be no Linux. Hence we see the dialectic shift from crowdsourcing to semi-crowdsourcing. As usually, the truth is somewhere in the middle, between single person design and total swarming by the crowd.
The Evidence: Crowdsourcing Works via Funding
Let’s take another product – already mentioned OpenOffice. Oracle has supported it. Financially. Without such support the quality would be even worse. Let’s take Redis. VMware (EMC2) supports it. Without VMware support Salvatore Sanfilippo would not probably continue it. VMware put him on staff and pays for continuation of the work on Redis. Let’s take Hadoop. There is a dedicated team at Yahoo! working on Hadoop, starting from Doug Cutting and his guys. Now Hadoop is maintained by Yahoo, Cloudera and others. BTW Doug Cutting works at Cloudera and he works on Hadoop.
The Shift: Open Source is No More the Crowdsourcing
Shift happened. Open source is not a crowdsourcing. Literally all modern companies produce open source which is not a crowdsource. Samples are Thrift and Cassandra from Facebook, Protocol Buffers from Google, Voldemort by LinkedIn, set of Scala-related projects by Twitter.
The open source is a channel of communication. Sharing of information is very important nowadays. Companies that share get indirect pay off. It is about new rules and relations in business, gamified business. It is good to see that outsourcing companies do the same! CloudMade shared Leaflet library for mobile mapping, the library is used by Flickr (Yahoo!) for World map and by dozen of other big names. Obviously the pay off is an advertisement and better karma for the CloudMade. ELEKS shared distributed storage for events called EventStore. It is a trend already and other should follow. Open source but not a crowdsource.
Where Funding Comes From?
At this point the reader should be comfortable with understanding that somebody has to pay so that the crowd produce something useful. Payments can go directly to the contributors, or indirectly. Sample of direct payments are maintenance and evolution of Hadoop or Redis. Sample of indirect payments is Wikipedia. Many people and organizations pay, but money does not go to those who write the pages.
This phenomena when multiple people (and organizations) pay is called crowdfunding. It is natural intersection between demand and supply to achieve something or create something. People from the village helped their representative to win some sport competition. People from same tribe gather money to solve somebodies problem. People from same religion gather money to do worship and rituals.
There is virtually unlimited number of people organization, virtual networks, by interests, by need, by whatever, for short, mid and long terms. But the fact is that people do collaborate and they are capable to fund the mutual goal. Everybody gives depending on capabilities (donations) or interest (share in future product). Wikipedia is a donation-based venture. Kickstarter and Quirky are interest-based ventures. Crowdfunding is what the crowdsourcing needs to work.
Big Commercial Funding vs. Crowdfunding
Big companies have adopted the variation of crowdsourcing known as crowddesign. BMW designs the cars by the crowd for years. BMW has success. Mitsubishi decided to follow the BMW, with Lancer Evolution X, and failed. Something went wrong, but what? The wisdom is hidden in pharmaceutical companies. They use crowddesign and crowdresearch for years and very successfully, e.g. InnoCentive company. The recipe is to keep core intellectual property behind the wall, while outsource some IP to the crowd. Gold mining company Goldcorp did an ultimate sharing of IP to the crowd and still succeeded. But more reliable (from business perspective) formula is to balance how much to share and how much to keep inside the company. Fundind (expenses) are partially on the interested company and partially on the crowd. Important aspect is a prize at the end. The crowdresearch, crowddesign is organized as a challenge, hence participants are motivated to win a grand prize at the end. Governments follow the same recipe. Right now we are participating in White House design challenge for better EMR, THE PATIENT RECORD. Great samples how it works could be found in Wikinomics and Macrowikinomics books.
On the other hand, small projects could be funded by the crowd. What if you want to make a movie for the Sundance festival? It is possible to gather money on Kickstarter. Here is evidence of 14 selected movies that were crowdfunded by Kickstarter. Quirky works differently, but it’s model also cool. Quirky triggers crowddesign and promise rewards for the winners and finalists. Anyhow the system is looped, there must be a feedback loop between those who produce and those who get paid.
What if all that stuff is running within your enterprise? Crowdresearch is a equivalent to Ideation. There could be same model with grand prize for the winner and bonuses for the finalists. It is called process gamification. The crowd has the power, it is underused in enterprises. There are still more orders than information sharing and swarming. Prototyping of new ideas could be done via crowdsourcing. Interested departments or managers could fund together based on donation or interest schemas, i.e. perform crowdfunding.
We can use same tools as used outside of the walled gardens of the enterprises to improve our enterprises. People can start something on their own, try to attract the others, build virtual teams and produce some end results. Based on the results, further plans are decided, either to fund, or crowdfund, or management-fund. Management can organize challenges for corporate initiatives and fund them partially, while rely on enthusiasm and gamified process for the crowd.
We have revealed the powerful modern mechanism for the businesses and the employees. Let’s establish it for the better operations and business results. At the end it is all about money, hence true way is to use this at your enterprise! This goes further than what is called Enterprise 2.0, it is more advanced than SLATES. It is much more about microeconomics rather than using some tools behind the firewalls.