Entries in crowdsourcing (14)

Sunday
07Mar2010

Most terrifying application of crowdsourcing ever

The term alone is mildly disturbing - "human flesh search" - and its application - vigilatism translated from the online to real-life worlds is terrifying. Simulatenously, it may be one of the more impressive examples of crowdsourcing around. Human flesh searches in China have tracked down individuals accused of wrongs ranging from stomping a kitty to death to instigating the suicide of a young woman based solely on images and videos uploaded to the internet. Intensively active chat room users put together the pieces to identify the wanted individuals' identities and place their personal details online for the community to enact its retribution.

The terror arises from nearly exclusively anonymous chat room participants who spend vast time (60% of users spend over 3 hours/day) investigating supposed crimes committed by people they don't know based upon accusations made by fellow community members who, thanks to psyeudonimity, they also don't know. I have expressed my optimism that the American way of using the internet increasingly based on building a verifiable online identify will encourage more creative expression. Now I'm also relieved that it is unconducive to vigilante mob behavior.

Monday
01Mar2010

Book Review: You are Not a Gadget by Jaron Lanier

The ascendant tribe is composed of the folks from the open culture/Creative Commons world, the Linux community, folks associated with the artificial intelligence approach to computer science, the web 2.0 people, the anticontext file sharers and remashers and a variety of others. Their capital is Silicon Valley...their favorite blogs include Boing Boing, TechCrunch, and Slashdot, and their embassy in the old country is Wired.

Thus Jaron Lanier describes the "cybernetic totalists" or "digital maoists" whose rising influence Lanier fears is leading us down a path of online culture where appreciation for humanity is displaced by blind trust in technology. In You are Not a Gadget Lanier laments recent trends in the online world - belief in the wisdom of crowds, reliance on algorithms for recommendations rather than people,  mashups and other piecemeal appropriation of others' content, templated web 2.0 designs - and argues that this failure to appreciate individual expression in the web world may have grave consequences for creativity and culture.

Lanier presents a variety of compelling examples of the web 2.0 platforms that online users now rely upon as well as basic structural elements designed decades ago that help drive to the characteristics that he fears most: fragmentation, anonymity, abstraction, plagiarism, and uniformity. He is most skeptical of many technologists' excitement over tapping into collective wisdom or consciousness: "Emphasizing the crowd means de-emphasizing the individual humans in the design of society, and when you ask people not to be people, they revert to bad mob-like behaviors."

He also strikes hard against the claim that "information wants to be free," focusing on Google's project to digitize the world's books. He fears that if the user interface of this project in the cloud encourages cutting-and-pasting and mashing up, then the result will be like the Bible: an anonymous single book. Such a scenario seems overly hysterical, but I am glad to see a strong stand for the value of content and the people who produce it. I agree wholeheartedly that information does not deserve to be free. But in arguing for a humanistic approach to technology, Lanier laments that "authorship - the very idea of the individual point of view - is not a priority of the new technology." This is hard to defend - the concept of shaping your online identity - one that is not psyeudonymed - and contributing via it online has never been stronger. People use their full names as their Twitter handles, they use their real names as they blog, they create Facebook profiles that connect them publicly with other real people. The movement to me seems to be in the opposite direction than Lanier describes - towards honesty and transparency rather than away from it. Sure, anonymous trolling still exists, but in mainstream online culture, that behavior is nearly universally rejected.

I was surprised that Lanier omitted crowdsourcing from his list of complaints. Given his concern for the "artistic middle class" that he views as assaulted by a file-sharing and mashup culture, I would suspect that he would be sympathetic to the arguments of the NoSpec! movement that lobby for online designers. These anti-crowdsourcers hold that the work of designers is constantly devalued by sites like 99designs that use contest formats rather than an interactive format between client and designer. Such interactions produce two outcomes that Lanier loathes: 1) stale and predictable web 2.0-ish logos and websites, and 2) only a small chance of compensation for the artists who produce them.

The book is a manifesto for humanistic creativity and quality online. He is right that much of the content on the web is repackaged, mashedup junk and he makes some suggestions about how people who contribute to the content of the web can improve it: designing a website that says something about who you are that isn't templated, creating an online video that took 100x longer to make than to watch, and writing a blog post on something you've been thinking about for weeks.

But the point that struck me the hardest and that I have been thinking about since I finished the book two weeks ago was around the web's influence on music. This challenge of Lanier's has inspired me to undertake conversations with every fellow music-lover I know:

Popular music created in the industrialized world in the decade from the late 1990s to the late 2000s doesn't have a distinct style - that is, one that would provide an identity for the young people who grew up with it. The process of the reinvention of life through music appears to have stopped.

Lanier himself is a musician and makes this accusation soberly. As evidence, he points to his own "experiment" to play varied songs to members of the "Facebook generation" and ask them to identify the decade that it was created. His observation is that people can do very well for the decades 1940s-1980s (clearly gansta rap didn't exist in the '50s and Depeche Mode is a definitively '80s sound), but that even true fans have a hard time telling if an indie rock or dance track is from the 1990s or 2000s.

In response, I began furiously attempting to brainstorm examples to prove him wrong.  I have failed.  Even the bands who I admired in the 2000s for their fresh sound - Arcade Fire and The National for instance - could plausibly be bands that emerged in the 1990s. But later conversations challenged me to view the absence of a single distinct musical style to be an asset rather than a weakness. Online music has opened up access to entirely new worlds of music to its listeners and if that diversity leads to an inability to clearly identify an artist's sound as "characteristic of the late 2000s" then so be it. I blame my terrible taste in music in the 1990s partially on the limits of the taste of my immediate networks and partially on the limits of traditional curators of content at the time. For their weaknesses (which I wrote about recently), Pandora and last.fm still serve as an amazing discovery engine. I am into more diverse types of music than ever before - I am better for it and so are the artists whose music I purchase daily. Yes, I agree that the mashups he critiques are a total bore, but I think that trend has passed. Musicians are drawing upon a variety influences just as they always have but that doesn't make them "retro".

Though I don't agree with some of his conclusions and I could do without some of the grand philosophizing asides of You are Not a Gadget, Lanier's challenges regarding current online culture are provocative and important. There are real weaknesses of online culture worth confronting (especially when we see them penetrating beyond the web as covered in today's NYT article on plagiarism as the new mashup-like "technique" in fiction writing). But what I see now in the rising online communities are people interacting as individuals and striving to engage thoughtfully in content and these trends promise a brighter online future.

Saturday
26Dec2009

Do crowdsourced Q&A sites deliver any value?

The theory behind free Q&A sites is appealing: ask a question to the masses and someone out there should have the answer. Yet, I'm beginning to think that free advice really isn't so great. Here's why:

Yahoo Answers. Probably the oldest and most known of these sites, the quality of answers on this site is appalling. Maybe worse is the quality of the questions. Perusing the questions can be mildly amusing for a while, e.g., Q: "Can your baby get pregnant if you have sex while pregnant?" A: "The baby can get pregnant only if it's a female. If you suspect that your baby is pregnant, try not to have sex again. You run the risk of getting your baby's baby pregnant and that can lead to complications like an infinite loop." But actually, it's overall more just depressing. I can't imagine the site making a comeback; the people contributing seem to be so useless and the domain of answers is so vast, that smart, new participants can't possibly feel any karma by helping others as the probability that they will be helped themselves is so low.

Hunch. I've written about its weaknesses before, but in a nutshell, the challenge is that the value of the site is entirely driven by the participation of the community. Since Hunch allows you to ask questions about anything, my experience is that across the board, the Q&A engine is weak.

Aardvark. I participate in this community and answer a lot of questions (which I like to think add value), but I must admit that I have never had a question of my own satisfactorily answered on this site. Recent example: Q: "What is a good website to discover apartments for rent in Washington, DC (anything but Craigslist please)" A1: "Craigslist" A2: "Craigslist". Their model makes a lot of sense though: identify what you areas you actually know something about (e.g., cooking blogs, the south of France, decoupage) and questions will be filtered before they arrive in your inbox. Both the questioner and I are better off by not waiting on me to give an answer on something say, sports-related.

Mahalo. It's like a slightly more intellectual Yahoo Answers. They try to encourage the karma factor by awarding "points" for participation, and allowing questioners to give "tips" for good answers, but these points only buy you karate-inspired levels. To attain a black belt, for example, you need 13,000 points. Almost all activity, however is in ONE point increments (with the notable exceptions of 1) initially joining and 2) embedding the Mahalo widget on your website, each which is a one time 50 point add), so this is a ridiculous amount of dedication. I'd rather go after a real karate belt.

What would I like to see? A movement towards niche, small community Q&A sites. By definition, all crowdsourced sites depend on their community for their power and in turn, communities are most engaged when they are focused. I like the attitude of Aardvark, but it is still challenged by how diverse it seeks to be; I am getting discouraged by the number of questions I am asked that I am forced to pass on because I have nothing of value to add. Etsy forums is a good example of a dedicated community question site, but it's not quite at the level of a true decision engine.

Saturday
31Oct2009

Victors & Spoils – the First Ad Agency based on Crowdsourcing Principles

Thursday marked the launch of Victors & Spoils, the world’s first creative advertisement agency built on crowdsourcing principles.

The company is based on the idea that bringing together the leadership and management of an ad agency together with the diverse creativity of crowdsourcing can create a new business model that can change the industry.

The company is still figuring some things out.  The logo and branding for Victors and Spoils will be the first project to be outsourced.  The compensation for designers also seems to be a work in process.  However, the company’s founders have made it clear that they believe that all participants should be rewarded in some form for being a part of the creative process.  As they state in their blog, “members will not only be rewarded for the solutions they develop (both individually and as a group) but also for participating in the community itself.”

In a response to the controversy surrounding spec and to shed more light on compensation, Victors & Spoils explains in another post:

Now if you’re thinking that “crowdsourcing” has gotten a bit of a bad rap lately from the “No-Spec!” movement, you’re right. It has. And we believe those naysayers make some really great points. So another thing to know about our model is that each V&S project will always yield more than one winner (Victors), will always provide more than one way to “win,” and will always have some of the largest rewards in crowdsourcing (Spoils) attached to it.  And perhaps the most important thing to know about our model is that all of these ways to contribute and win will build each creative’s V&S Reputation Score (or better name TBD).”

It doesn’t seem that Victors & Spoils plans to set up their own crowdsourcing platform for the time, but instead leverage existing ones, such as CrowdSPRING, 99 Designs, and Genius Rocket. 

As the company continues to develop its operational process it will be interesting to see what develops.  Will it make more use of social networking principles to build the community that its founders envision? Will decision making be limited to the company managers?  Will it truly become a place where talent is developed, so that even if participants don’t win a prize they win a better chance at a job somewhere else in the long run? I believe that this new company has many opportunities to innovate a new model that truly uses the masses to create a better product.  A part of how successful it will be in doing so depends on how far it is able to limit the “sweepstakes” model.  But there is also room for adopting new and innovative tools and processes.  As the first of its kind, I look forward to following its development.

Tuesday
29Sep2009

Kickstarter is the epitome of awesomeness

22 hours to go to contribute to the development of "Put This on: A Web Video Pilot About Dressing Like a Grownup," a project crowdsourcing its financing on Kickstarter.  The project is already 264% funded and it's easy to see why.

Donations can be as small as $2 (with the reward of a warm and fuzzy feeling) to $50 (to receive funding credit in the video) to $250 (earning a personal style consultation with the project's creator), and beyond. No one has taken up the $250 offer, although it is creative rewards like that that make this site so fun to browse.

Contribute to an artist's dream project + concretely share in the experience. So cool.

Friday
25Sep2009

Taking the prize model to government: innovation, taxpayer savings, greening

Following the model of the Netflix Prize (winners officially announced this past week), the Department of Energy has its own L Prize: $10 million and consideration for future federal purchasing agreements in exchange for inventing a bulb that produces the same amount of light and color of a 60 watt bulb while using only 10 watts of power.  For the government procurement business, known for waste, contractor relationships, and stringent statements of work that hamper creativity, this type of open call for innovation by industry at a total bargain to the government is a welcome approach.

Like Netflix, DoE is implementing a goal-oriented program. Netflix told its participants that it wanted a 10% improvement in its recommendation engine; it did not say how that target should be reached. Similarly, DoE has laid out clear metrics on what the bulbs should be able to achieve, noting the ultimate goal that replacing 60 watt incandescent bulbs in the United States with their LED equivalents as described by the Department would cut carbon emissions by 5.6 metric tons annually; it did not provide the interim benchmarks and deliverables standard in government contracts.

Which gets to a major take-away: By freeing itself from traditional government contracting procedures, DoE has been able to incentivize massive research and development at private companies at a great cost savings to the government. Phillips has already submitted the first entry that is currently undergoing testing in the Department, while other companies are said to also be developing their own alternatives. DoE is also shaking up the lighting industry that has seen very little innovation since the 19th century when the incandescent lightbulb was first invented, by targeting the ubiquitous bulb that has nearly 50% market share in this country. An industry that wasn’t really taking energy efficiency seriously is now encouraged to pay attention. It’s the best type of government intervention: nudging through incentivizing rather than new standards that just pass additional costs onto consumers.

Of course, the prize model has its drawbacks. Just as The Ensemble in the Netflix contest found out upon submitting a winning product 10 minutes after BelKor’s Programmatic Chaos, there is no reward for second place. Three years and thousands of hours of work in this case lead to $0. It’s similar to the critique made of many creative design crowdsourcing sites – participating in prize competitions is inherently risky. Maybe there is a bit of glory in making it to spots 2 or 3, but in most cases, that is not worth the effort expended. When small designers are the ones slaving away in this high risk environment, it’s hard not to feel that something isn’t quite right. However, when huge industry behemoths that have resisted innovation for years are competing for lucrative government contracts, the drawback of this model – a first place winner-take-all outcome – seems to be greatly outweighed by the benefits – fostered innovation, taxpayer savings, and greening.

Saturday
19Sep2009

Unilever Testing the Value of Crowdsourcing

As the heated debate surrounding crowdsourcing in design and marketing rages on, another big name has decided to turn to the crowds for some help in branding.  Unilever, one of the world’s largest fast-moving consumer product companies, recently decided to offer a $10,000 reward on Idea Bounty for the winner of a competition to advertise Peperami, the company’s popular meat snack for children. 

Unilever owns more than 400 brands, including those pictured to the right.

Idea Bounty is a new crowdsourcing platform for designers and despite being less than a year old, it has attracted big names such as Redbull, BMW, and World Wildlife Fund.

Unilever claims this is not a publicity stunt.  In fact  the company believes that crowdsourcing marketing ideas might be a sustainable strategy they would be willing to consider for their other brands as well. The Peperami competition seems to be in some ways their pilot to test the value of crowdsourcing. Unilever’s Matt Burgess, (Managing Director of the division that owns Peperami) indicates that the company sees potential in creative platforms: "We want to get the creative back from 'good' to 'outstanding' again. The best way to increase our chances was to increase the amount of creatives exposed to this brief. This is the overriding driver."

Putting aside the debate about sites such as Idea Bounty or 99 Designs, it is interesting to see Unilever step into this space in such a strong manner.  Unilever is a company that is willing to take risks in the name of innovating ways to stay ahead of its completion.  One example is its Shakti program in India. By partnering with self-help groups it enables rural, poor entrepreneurial women to take on microcredit and purchase consumer goods from small retailers that they can then sell door-to-door and earn income. After several years, some pilots, and alterations, the program broke even, scaled rapidly, and increased revenues by cutting costs, reaching new clientele, and leveraging (while also supporting) local microenterprises to better market its products.   Could Unilver’s decision to crowdsource mark another such attempt to stay ahead of the competition? And if such an attempt is a success, what could it mean for sites like 99 Designs and Idea Bounty?

The truth is that companies like Unilever are not the typical “clients” one might expect at crowdsourcing sites.  Their entry into this field may mark a transformation for creative platforms, (from platforms generally catering to the small guys to ones that have real value for companies earning annual revenues in the billions) and it will be interesting to see the implications of such a transformation.

Saturday
12Sep2009

Crowdsourced product development at Quirky.com

I expressed my skepticism last week over the effectiveness of crowdsourcing creative work and most of the commenters agreed.  I suggested that crowdsourcing at its best should be collaborative, not winner-take-all.

A real collective creativity platform may now exist: Quirky.com is a site to bring "community developed" products to the marketplace.  Would-be inventors submit their ideas to the site, the crowd of Quirky users votes on the best idea to go into development, and then the site community makes decisions on design before the product moves out for manufacturing. The original inventor earns 12 cent for every dollar the product brings in and other contributors also share in the future value of the product.

At first glance, the concept is a big odd. As an inventor, you PAY $99 to share your idea with little chance of it turning into something big and the assurance that, if it does, you've just given up a big share of its future profit. Yet, I still love this site. It's probably not the place to bring your sure-fire brilliant idea, but the site is full of funky, oddball, just-might-be-big proposals for small manufactured items. If your idea is terrible, the site community will let you know so and save you the hassle of moving forward with it at greater expense. If your idea has promise, you'll get objective feedback from a community of interest that may greatly improve the concept. And best of all, Quirky has agreements with manufacturers to actually move a concept into production.

It's a fun way to be an armchair inventor and submit an idea that you've been thinking about for a while --whether it's a "sexy thigh holster carry", "thimble mouse", or "curvy groovy wrist rest" -- and see if it goes anywhere. Surely not a mechanism to profoundly change the world, Quirky is nonetheless a fun collaborative creative community and I like the market approach to rewarding thoughtful participation.

Tuesday
01Sep2009

The other side of crowdsourcing

I write about the power of crowdsourcing a lot, from its applications in venture capital to the Netflix prize to funding artistic ventures to micro-volunteerism to patent review. My own blog's header was created at 99 Designs.

Yet...the design community is not so thrilled. Especially with the launch of a campaign by advertising agency juggernaut Crispin Porter + Bugusky to crowdsource the logo of Bramma through the site Crowdspring, designers are railing against what they see as the devaluation of their work.

They may have a point. CP+B will offer a paltry $1,000 prize to the winning design, while the remaining 753 submissions will receive nothing. For Bramma, it's a great deal (although it was certainly a dumb move to hire CP+B as a facilitator): 754 logo choices for a mere $1,000. On the other hand, what quality was lost by having no interaction between client and designer, especially with a short design brief with this gem: " We like the idea of representing a bull...We also like the idea of not representing a bull" ?

Sites like No!Spec argue that language like "crowdsource" hides the basic fact that 99 Designs, Crowdspring and Genius Rocket are essentially design prize sites for speculative work. While I'm not crazy about their tactics, I agree that the term crowdsourcing may be misused. While design sites tout the "community" aspects of their sites, the truth is that they play a winner-take-all game where work is purposefully not collaborative. The power of crowdsourcing comes from when many people working together can achieve something big, like when hundreds of people spend a few minutes each tagging museum photos through the Extraordinaries or when 50 people each contributing $50 can collectively fund an artist's dream project through Kickstarter.

Design sites are fundamentally different. There is no cumulative effort among designers to produce the best result. Rather, each works independently, maybe learning gradually from designs that have been thrown out or voted up, but too much learning and adoption of good ideas is strictly disallowed. These are prize sites, and while I think they still have value for small projects (such as my blog header -- I would have never engaged a professional design firm) and for budding designers (they are a mechanism to build a portfolio), the idea of crowdsourcing design should not be oversold.

Crowdsourcing should be about collaboration to solved problems and make decisions. Even the $1 million Netflix contest saw great sharing and discussions in their forum. Even if a single best idea needs to rise to the top in the end, it should do so because it has been honed by the crowd over time, not because one selector was able to simply take his choice from several hundred disparate submissions.

[Update: Based upon the many comments and conversations that I've had in relation to this post, I should probably clarify my position. The CP+B example is indeed an outlier; most of the entities that choose to use design sites are small (like mine) and simply not in the ad agency market. I would also highlight that designers enter this sites voluntarily of course, so they clearly see some value in participating, making the No!Spec attacks seem a bit sensationalist. The main objectives of this piece were to: 1) discuss some of the reasonable critiques made of the disruptive market power of design sites, since they are often described universally favorably; and 2) question what it really means to crowdsource -- is discovering diverse points of view sufficient or is some greater good of collaboration implied? I would argue for the latter.]

Friday
14Aug2009

Taking crowdsourcing to the next level at Trampoline

Finally, a successful application of crowdsourcing as an alternative to venture capital investing. Trampoline Systems, a UK-based social analytics business is relying on "crowdfunding" for its series B investing -- soliciting smaller investments of a minimum of £10,000 each to collectively total £1 M.

Trampoline has navigated the difficult waters of regulation to enable this creative approach. We've seen crowdsourcing in action for project financing before at Kickstarter, but there backers are donors, rather than investors, without any share in the future value of the projects that they support. Just as with social lending, it is far easier from a legal perspective to make Kiva work (where loans are interest-free donations) than Prosper (where loans are interest-bearing). With Trampoline's model, nearly anyone can become a mini-VC of sort, funding projects at small amounts and sharing in the returns.

The crowdfunding model lends itself to better investment diversification, although investors likely have little to say about how the company is run. Maybe that is for the best from the company's perspective; when funding is distributed across many investors, each investor has less power and in the end, the company may likely have more freedom of action.

Trampoline isn't sharing the details yet of how they are making this work exactly, but lessons learned from this experiment may be invaluable for both startups looking for alternative financing options in a tough VC climate and for smaller time investors looking to get involved in the VC space.