De term ‘crowdsourcing’ werd voor het eerst gebruikt door Jeff Howe van Wired.com. Crowdsourcing is het handig gebruik maken van de massa als kennisbron. Bekende voorbeelden van crowdsourcing zijn de bekende encyclopedie Wikipedia, restaurant vergelijkingssite Iens.nl (70.000 restaurant bezoekers recenseren 17.000 restaurants), TomTom mapshare (TomTom gebruikers helpen mee met het up-to-date houden van kaarten).
Ook bedrijven als Procter & Gamble maken dankbaar gebruik van crowdsourcing. Het geeft ze de mogelijkheid om problemen voor te leggen aan grote groepen mensen tegen een fractie van de traditionele R&D kosten.
De kosten zijn niet de voornaamste reden om te doen aan crowdsourcing. De kracht zit hem in de kennis. De ‘crowd’ heeft heel veel verschillende invalshoeken, waardoor er zeer vernieuwende ideeën ontstaan. Inmiddels heeft P&G 137 producten op de markt gebracht die op deze manier bedacht zijn. Voorbeelden hiervan zijn: Olay Regenerist en Swiffer Duster.
Niet altijd
Crowdsourcing werkt niet altijd. Zo heeft Dell aan zijn klanten gevraagd wat zij graag wilden. Op nummer 1 stond een Dell computer met Ubuntu. Dell bracht de computer met het besturingssysteem uit, maar deze werd maar weinig verkocht.
Je crowd tevreden houden
De crowd doet alles vrijwillig. Ze vinden het leuk om mee te denken en om ideeën in te sturen of hun mening ergens over te geven. Het is belangrijk om te zorgen dat de crowd het leuk blijft vinden om je te helpen. Wikipedia heeft daar begin dit jaar problemen mee gehad. Ze hadden extra regels opgesteld om een wildgroei van artikelen te voorkomen. Een groot deel van de crowd kon zich niet vinden in de regels waardoor er in het eerste kwartaal 49.000 mensen ophielden om Wikipedia up-to-date te houden. Dit is 10 keer meer dan er indezelfde periode vorig jaar er mee ophielden.
Bronnen:
http://www.innovatieforganiseren.nl/innovatie-en-management/crowdsourcing/
http://nieuwemedia.web-log.nl/nieuwemedia/2007/10/het_nut_van_cro.html
http://webgrrl.nl/200812/crowdsourcing/
http://www.frankwatching.com/archive/2008/05/26/crowdsourcing-the-cream-gets-on-top/
http://www.dutchcowboys.nl/online/18387
http://www.microengagement.com/index.php?view=article&catid=25%3Aexamples&id=55%3Aproctor-and-gamble&option=com_content&Itemid=34



3 reacties
Cas 7 december 2009 om 16:01
Wow, waar is de vermelding van Surowiekci’s ‘Wisdom of the Crowds’?? Hij kan ook verklaren waarom crowdsourcing in het geval van Dell niet werkt.
Beetje een summier artikel; er is zoveel crowdsourcing potentieel!!!
Cas 7 december 2009 om 16:15
Hmm… weinig opbouwende die vorige post van mij.
THE FLIRT MODEL
The FLIRT model (http://www.slideshare.net/brayrie/flirt-overview) of crowdsourcing was developed by Viitamäki (2007). The purpose of the model is to give an overview of the most important aspects of crowdsourcing. The outer layer of the model comprises the main components, namely: Focus, Language, Incentives, Rules and Tools. The inner layers comprises more specific aspects that make up the main components.
The model describes that before implementating crowdsourcing one should consider the strategic focus and objectives of the project, after which tactical considerations such as language, incentives and rules are considered. Finally, the project is made operational with certain tools and technology.
The model is aimed at anyone considering crowdsourcing. It can be used as a guideline of aspects that have to be considered. During the course of our project we used to FLIRT model to analyze current initiatives, and to guide us with the development of our own initiative.
LIMITATIONS OF CROWDSOURCING
According to Surowiekci (2005) a crowd needs to satisfy four elements (see below) it can definitely be argued that our small poll does not satisfy these elements. So we cannot speak of a smart crowd. Therefore the results of this poll should viewed as a way to illustrate the possible functioning of crowdsourcing, not as a definite definition.
Divisity of Opinion
Each person should have private information even if it’s just an eccentric interpretation of the known facts.
Independence
People’s opinions aren’t determined by the opinions of those around them.
Decentralization
People are able to specialize and draw on local knowledge.
Aggregation
Some mechanism exists for turning private judgments into a collective decision.
CROWDSOURCING EXAMPLES
Innocentive
Founded back to 2001, InnoCentive built an initiative web community for open innovation, to collaborate and to deliver breakthrough solutions for innovative R&D-driven organizations. There are two important roles designed in this community: Seekers and Solvers.
Seekers are corporations in large or small size, or non-profit organizations and government entities. Some examples of current Seekers are: Procter & Gamble, Boeing, DuPont, Eli Lilly and Company, GlobalGiving and The Rockefeller Foundation. The Seekers reveal their R&D tasks, so-called “Challenges”, to the community platform, so-called “Marketplace”. The Challenges types cover wild ranges: Business and Entrepreneurship, Chemistry, Computer Science and IT, Engineering and Design, Food Science, Life Sciences, Mathematics and Statistics, Physical Sciences, Requests for Partners/Suppliers.
When a task is fulfilled, Seeker gives award to the Solver of that particular task. Awards range from $5,000 to $1 million based on the complexity of the problem. Up to Feb. 2009, out of 814 Challenges, 391 have been awarded, for prizes typically in the $10,000 to $25,000 range.
The other important role in this community is Solvers. More than 170,000+ engineers, scientists, inventors, business people, and research organizations in more than 175 countries are registered. Thus, we may agree with what InnoCentive claims: “the InnoCentive Solver community is one of the worlds most diverse, socially committed and financially rewarded network of creative minds.”
For all 814 Challenges ever posted on InnoCentive, the Solver community submitted more than 12,500 solutions. Based on the statistics from InnoCentive, the average success rate is 40%. Despite the chance of solving a problem is not very promising, yet the winning Solvers already have received more than $3.9 million to date (Innocentive, 2009). Jill Panetta, InnoCentive’s chief scientific officer, made a proudly remark on the Solvers solutions, “it is the percentage more than would have been solved using a traditional, in-house approach”.
Threadless
Founded by a technology company called Skinny Corp., Threadless’ main goal is to profit from selling and distributing the T-shirts. In order to eliminate the need for a design department of its own, Threadless crowdsources its T-shirt print designs. On the Threadless website, they have built a community of registered account users which exceeds the 700,000 members. There are two key functions for this community: (1) an ongoing open call for T-shirt design submission, and (2) evaluating for the submitted designs. All members in this community can play both roles as design submitter and design critic.
A design submitter interacts with the community in two options, learning suggestions or getting final evaluation from critics. A designer first can go to the critics blog, where members give comments on the design. Later when determining whether to publish the design for official selection, he/she needs to download a ready template from Threadless, and follows certain rules as size of design, max. number of colours in design, and etc. Seven days after submitting, the designer will get a final evaluation from the other members in the community. This evaluation is used as a gauge by Threadless to decide what gets made into a shirt or print. If the submitter’s idea is selected for print, he/she will receive some awards: $2,000 in cash, $500 Threadless Gift Certificate (can be redeemed for $200 cash) and $500 in cash for every reprint. If the design wins a “Bestee” in the Threadless Awards, a further amount up to $22,500 will be granted (Threadless, 2009).
A submitter of one design can still be a critic for other people’s designs. A critic gives comments freely on the critic blog. Since a critic is very possibly a designer him/herself, there are often practical designing skills suggested on the blog. Acting to a official submitting for design selection, the critics have 7 days to evaluate designs by giving score from 0 to 5. Later on, among the most popular designs as scored, Threadless selects some as mass production patterns.
Yahoo!Answers
Launched by the internet service provider Yahoo! In year 2005, Yahoo! Answers is a community-driven knowledge platform which allows users to both submit questions to be answered and answer questions asked by other users. The question categories are very diverse, almost cover from A to Z, such as subjects in daily life (food, health, pets), educational (science, business, society, culture), or recreational (arts, restaurants, gardening, sports) and etc. It also has a very large member base, consisting of more than 21 million unique users in the U.S. and 90 million worldwide. Based on regional differences, Yahoo! Answers has regional servers targeting local services of 26 countries, providing available languages as English, Chinese, French, German, Indonesia, Italian, Japanese, Korean, Portuguese, Spanish, Thai, and Vietnamese.
Comparing to its massive global framework, the set up of individual local site is pretty simple. Every user can ask, answer, discover or vote other people’s questions easily and almost totally freely. Everyone who answers a question can get basic honour points. When a best answer is selected, some extra honour points will be granted. Besides, in the Taiwan and Hong Kong’s region, where the titles are valued by the societies, every member is titled as “freshman, expert, master, and etc.”, following by level 1, 2, etc.
Yahoo! Answers basically let the community itself control the content quality. A group of people, so called “Moderation Community”, are chosen by the webmaster to act as delegates, enforcing the community rules on the webmaster’s behalf. These moderators are given special privileges to delete or edit others’ contributions and/or exclude people based on their e-mail address or IP address.
Source: van den Brand, J.L., Hung, Y.T., Janssen, D., Kramer, J., Schalkx, C., (2009) Crowdsourcing, towards a definition. Working paper for Seminar Information Management.
Maarten van Strien 14 december 2009 om 14:45
Dell blijft haar best doen. Ik voel me niet helemaal aangetrokken tot het meedenken in het designproces. Misschien omdat ik ooit tevreden Dell-klant ben geweest en dus geen reden heb tot aandringen tot verbetering? Zou dit dan betekenen dat tevreden klanten dit niet laten merken en crowdsourcing wordt gevoed door criticasters?
Jouw reactie