Wikipedia: Example of Customers as Collaborators (Complete with Unintended Consequences)

In my upcoming book, Outside Innovation, I’m using as one organizing principle the different roles that customers can (and should) play in shaping your business and its products. I’ve identified lots of different roles. Now the challenge is to describe and differentiate them.

Engaging customers as collaborators is one of the most advanced, and the most challenging, business propositions. Open source software development is a prototypical example of customers as collaborators. Customer collaboration is when customers band together to co-create something new and complex. They work together in an open, transparent environment to co-create new products, solutions, and works of art. Each customer contributes his or her own intellectual property and/or builds on a contribution made by a colleague.

Wikipedia is a good example of the power and pitfalls of the “customer as collaborator” role. Launched in January 2001, and based on the open source model of collaborative development, Wikipedia is the world’s fastest growing, most current, and largest encyclopedia. Wikipedia provides a great example of how complex projects like writing software, solving really difficult problems, and creating and classifying knowledge are increasingly becoming customer-collaborative projects.

Here’s the official description: “Wikipedia is currently the world's fastest-growing, most current, and largest encyclopedia, with 2.5 million articles under active development in over 120 languages. It is created entirely by volunteers who contribute, update, and revise articles in a collaborative process. The content¬¬––text, images and sounds—that is contributed to Wikipedia is licensed under the GNU Free Documentation Licence (GFDL), which lets users copy and modify each other's work based on a principle known as ‘copyleft.’ The entire database is freely downloadable.”

Anyone can author or edit content to contribute to Wikipedia. And thousands do. For example, 28,258 people contributed at least five times each (Active Wikipedians) in October 2005.  In the same month, there were 4,573 “Very Active” Wikipedians (people who contributed at least 100 times that month). There’s a close-knit cadre of about 1,000 volunteer authors and editors who monitor the newly-submitted edits and articles. Articles are categorized by the community as Good or Featured articles. In order to qualify as Featured articles, they need to be nominated, meet specific criteria, and gain approval by a consensus of reviewers.

The world is definitely a better place post-Wikipedia. No matter what topic you need to know about, “Google it” and you’ll probably get a Wikipedia entry with a great definition and overview, and a good starting point for locating other resources. What’s really great about Wikipedia is that it’s free to all. It’s available around the globe in many languages. It’s more current than any printed source. Most importantly, it closes the knowledge gap.

But there is a dark side. Watching the Wikipedia Foundation deal with the dark side provides a good source of lessons learned for any company that would like to unleash the power of customer collaboration to provide tangible results that would be too costly and time-consuming to create by any other means.

You Can’t Put the Feathers Back in the Pillow

In his editorial to USA Today on November 29, John Seigenthaler, Sr. told this story: “When I was a child, my mother lectured me on the evils of "gossip." She held a feather pillow and said, ‘If I tear this open, the feathers will fly to the four winds, and I could never get them back in the pillow. That's how it is when you spread mean things about people.’ John went on to say, “For me, that pillow is a metaphor for Wikipedia.”

John was slandered by someone who maliciously edited his Wikipedia biography, falsely accusing him of being a suspect in heinous crimes. The libelous material remained on Wikipedia for four months, during which time it was syndicated by other heavily-trafficked information sources, including Reference.com and Answers.com. Once he was alerted to the slander, John and his lawyer spent four months trying to locate the slanderer and to seek redress. Wikipedia couldn’t identify the culprit. All of its postings were anonymous. Wikipedia could (and did) remove the offending material, but it couldn’t pinpoint its source. Because Wikipedia is classified as an information service, not as a publisher, it’s exempt under U.S. laws from being libel for the damaging misinformation that was propagated from its site.

Another interesting part of the Seigenthaler saga is that according to an article in Editor and Publisher, the culprit did confess and apologize, but only after he was tracked down by Daniel Brandt of San Antonio. Daniel Brandt has been very vocal in his criticism of the Wikipedia Foundation on his Wikipedia-watch.org Web site and has been labeled as a kook by members of the Wikipedia community.  Daniel asserts that Wikipedia has been structured so that it actually encourages slander, pranks, and misinformation, and promotes libel. It was nice of him to help resolve the Seigenthaler issue. Here’s a prescient editorial that Brandt wrote calling attention to many of the issues now under discussion.

Seigenthaler wasn’t the only prominent person libeled in Wikipedia.  Norway’s Prime Minister, Jens Stoltenburg, discovered in November 2005 that his Wikipedia biography had been inappropriately updated and contained a number of libelous statements. By the way, Europe’s libel laws are not as charitable to Internet service providers. And, once print editions of Wikipedia are produced, libel will definitely become a legal concern.

In another high-profile example, in December 2005, MTV star Adam Curry admitted to having edited the entry on Podcasting to give himself more credit than others for inventing the popular genre.

How is Wikipedia Dealing with These “Unintended” Consequences?

The Wikipedia core community of volunteer, dedicated contributors and editors has been operating under the assumption that their communally-created and refined knowledgebase will be self correcting. Errors will get caught and reported by the general public who reference the material. Misapprehensions or misstatements of fact will be challenged and addressed.  What the founders and the community apparently didn’t think (enough) about is the possibility of slander, character assassination, and pranks with unintended consequences.

How Wikipedia Is Changing Its Modus Operandi

Experiment: Article Authors Are No Longer Anonymous. In response to the Seigenthaler flap, Wikipedia has entered an “experimental” period of disallowing anonymous people to create articles on the English version of Wikipedia. Only registered (and presumably traceable) users will now apparently be allowed to create and post articles. In its press release of December 5, 2005, the Wikipedia Foundation wrote: “Founder Jimmy Wales… experimentally removed the ability of unregistered users to create new articles in Wikipedia. Unregistered users will still be able to fix spelling mistakes and add to existing articles but are required to register a user account before creating new pages. Wales said: ‘This will reduce the work load on the volunteer editors controlling contributions to the project.’”

This is a good first step, but it clearly doesn’t prevent the kind of slander to which John Seigenthaler and Jens Stoltenburg fell victim. In both those cases, it appears that anonymous editors made changes to their respective biographies. A better approach would be to require all contributors—both article creators and editors—to be registered and traceable users. I suspect that the Wikipedia community will eventually decide that anonymity for editors as a public good is outweighed by the need for accountability and redress.

Semi-Protection Policy Approved. There are two types of vandalism that have plagued Wikipedia. The first type is an occasional malicious change made to an article by an individual (which may or may not get detected and corrected).  The second type is a pattern of rapidly-recurring changes to a particular article. The most notable of these is the incessant vandalism of the George W. Bush article.

To deal with this kind of “piling on,” the Wikipedia board has approved a change (which as of this writing has not yet been implemented in software), to wit: “Semi-protection of a page prevents the newest X% of registered users and all unregistered users from editing that page.” If a page is being repeatedly vandalized, it can be protected by request. As the policy explains, this is not a pre-emptive measure against vandalism, but it’s a response to vandalism.

Published Version Is Reviewed. The Wikipedia Foundation is now discussing the possibility of having two versions of the site: a stable, published version and an editable version. The articles that appear in the stable published version would have been through a community review process (not necessarily an expert peer review process). Essentially, the author of any “featured article” submission can request “creative feedback” (also labeled peer review) from the community. The purpose of the stable site is to improve the quality of and trust in Wikipedia information. Presumably if you are reading an article on the stable site, you’ll be seeing neutral information that has been checked (at least by members of the community) for obvious vandalism, editorial slant, and proper citations.

I think this is an excellent solution. When you’re searching for information, you could check both sources. The stable source would be the most reliable. The editable source might have more up-to-date information on a topic, but it might also contain information that is suspect. Caveat Lector.

Dealing with Vandalism and Slander Is Different from Improving Information Accuracy and Authority

The Wikipedia Foundation appears to have taken the first baby steps in confronting the biggest problem with Wikipedia: vandalism and slander. Now, it will be interesting to see how well the community deals with the second challenge: improving the accuracy and authoritativeness of its articles.

Should Wikipedia Be the Last Word or the First Word? Wikipedians want their collaborative brain child to be viewed as a free alternative to the Encyclopedia Britannica and other published works that many people in the world may not be able to afford. That means that its information needs to be trustworthy. The “we’re as accurate as the other guys” argument seems to have been bolstered by the special report published in the journal Nature, comparing the accuracy of scientific articles published in the Encyclopedia Britannica and those in Wikipedia, and essentially declaring a tie. “Only eight serious errors, such as misinterpretations of important concepts, were detected in the pairs of articles reviewed, four from each encyclopedia. But reviewers also found many factual errors, omissions or misleading statements: 162 and 123 in Wikipedia and Britannica, respectively.”

At the same time, Wikipedians appear to be contemptuous of people who actually rely on it as if it were an authoritative source. They are critical of reporters who take them to task for not being trustworthy, suggesting that reporters, in particular, should know a thing or two about source checking. In the summary of its “Researching with Wikipedia” article, the Wikipedia Press Kit suggests “In some cases, it's better to use Wikipedia as the first step in the research process, rather than the last step.” In fact, the BBC’s Bill Thompson was praised by one Wikipedian for this quote: “I use the Wikipedia a lot. It is a good starting point for serious research, but I would never accept something that I read there without checking.” Nice to see at least someone in the media knows how to use Wikipedia! :)

The Most Current Word and the First Word. Over time, as the Wikipedia Foundation tightens up its quality control and peer review processes, the published/standard versions will become a more trusted source. But my prediction is that Wikipedia will never become the most trusted source of information—the last word—on anything. What it will do is fill a hugely important gap in every field of endeavor. Wikipedia will become the first place you go to find out the latest information on a topic.

Comments

[Patty Seybold]

Dave,
What a great idea! You should send it to Jimmy Wales--Wikipedia's founder...I'm not sure you should limit it to 3 students per teacher, or one article per student. There might be some folks who really excel in becoming experts on a topic and crystallizing it for the rest of us.

I agree that having a signed Wikipedia article that is held in high regard would be a great credential for anyone to use.. whether in applying to college or on a resume...

Patty

[Dave Lance]

I propose that all high school teachers worldwide be empowered by their boards of trustees to offer an opportunity for three of their best and brightest students to earn an extra credit "A."

The offer would be limited to three extra credit "A's" per school year, per teacher.

The teacher would select the three students, and the subject matter for each project. The opportunity would unfold in two phases:

(1.) Each participating student would study and learn the research standards applied to articles published in the Encyclopedia Brittanica. (Or perhaps the Columbia, or another, more suitable publication.) The student would study and learn the research principles and techniques, and demonstrate a firm grasp of them by passing a standardized test.

(2.) The student would then apply these principles to produce a Wikipedia article. Strictly following the aforementioned standards, the student would perform and document the research. They would produce a complete and accurate set of citations and source references. They would explain how and where the information was gathered. They would then write a series of drafts which the teacher would review and edit. Finally, the student would publish a Wikipedia article based on their research. The final article would include all of the research documentation, citations, and a complete description of their research techniques. Both teacher and student would affix their names to the finished article.

Everyone invoved could take pride in the articles that bear their names. Their accomplishment could be integrated into resumes, college applications, etc. National and International contests could be established to select the best articles for each school year. Wikipedia would certainly win, as would the students, teachers and their school districts.

It would be a win-win all the way around, and it would - over time - add a level of credibility to Wikipedia that the current system simply does not warrant.

[Patty Seybold]

Bardo,
I LOVE your suggestion about having a single version of Wikipedia containing both the approved version and the comments/suggestions in a single place.

That sounds like the right solution.

Patty

[Bardo N. Nelgen]

Patty,

thanks for these acknowledging words. :-)

Just some small add-ons to the two-versions concept:

My reservstions towards the two-version Wikipedia originally resulted from interaction issues:

How would you use such thing ?

Look at the "trusted" version first, to get the facts right and then turn to the "open" version for more resources ? That would mean having to compare two almost identical articles literally phrase by phrase (the way changes in a wiki use to work...). I just would not consider this being too attractive to the user.

So as a "lightweight" version of the trusted-authors-system, my proposal would be, having articles inside ONE version of Wikipedia, which contain "edit-protected" areas, that have been counter-checked (by whomever of the trusted editors) right inside each otherwise freely editable article. Maybe including some way to propose corrections to the peserved parts.

This would preserve the open nature of the wiki-concept, while helping to ensure information-reliability.

However I agree with you, that currently there is a lack of transparency, when it comes to WHO approves certain content.

Maybe they could borrow some ideas from 'professional' publishing workflow software, which - in my opinion - has introduced some quite successful approaches towards community-based editing during the recent years.

Bardo

[Patty Seybold]

Bardo,
What a wonderfully thoughtful commentary. You're absolutely correct that the issues that have gotten so much press with Wikipedia lately are not new, and probably were predictable.

The pseudonymous registration sounds like a great idea. If people were trusted and vouched for, they probably would behave.

I'm not convinced that the two version idea doesn't work. If there's still a completely fluid version that is being constantly updated, we wouldn't lose the currency.

I have colleagues in the world of scholarly publications, where referees decide which articles are published and which are not. How do you become a referee? You write articles and get them published, then you apply to be a referee. Granted, that's too much of a hurdle for Wikipedia, but the idea
of earning the right to approve articles for publication is a good one. It's akin to the practice in a number of open source development communities, in which the most respected contributors are the ones who vet others' contributions before they're "published."

Today, there's no real visibility into the process of WHO approves Wikipedia entries, although there are very clear stages of quality control that submissions should go thru. It seems to me that by adding pseudonymous registration to submit, enabling anyone to contribute and/or revise, but adding a mechanism for people who are authorities in each subject matter area to be included in the final approval process for the "published" versions, Wikipedia could improve the quality of the collaborative output without sacrificing the level playing field for submissions.

Thanks again for your thoughtful contribution!

Patty

[Bardo N. Nelgen]

Non-Collaborative Customers:
Fighting Predictable Consequences

The current awareness of quality loss of Wikipedia articles as much buzz - as it may generate at the moment - is not exactly a brandnew phenomenon, which has never happened before. Much more this has occured to DMOZ, Yahoo! and loads of others before and it can be observed at Google within every second.

Trust. It's about who you trust to reliable and trustworthy.The degree of availlable trust, mainly depends on two factors.

Scale. With added scale (number of units people, machines,... - involved) trust decreases.
Control. If you want to keep trust at a certain level, you commonly can't avoid some sort of control.

Scale

But why is trust going down at all ?

People You Know
If you are only a couple of people and know each other for some time, this is no big issue. You can quite safely make your personal assumptions who's trustworthy by talking to her or him and maybe to some people you both know. Since this is only possible if dealing with a quite small number of people.
This is the situation at the beginning of almost any web project: you invite people you know and who may even know each other to try out your latest product or service.

People Who Know You
As the word is spreading about your idea, more and more people may enter, that you don't know and who you do not know in person either. But they have heard about you and your project from their friends or colleagues and therefore trust in their judgement towards product quality and reliability. These people also share a lot of the mindset of your first-grade testers so someone practising sabotage intentionally from within this group will be very unlikely either. Most of them will at least attempt to play by the rules (in Wikipedia terms: only publish facts they really know about and in case of doubt - can prove).
This mindset-model may reach to fairly large scales as seen at many famous community projects on the web.

People No One Really Knows
However, as this thing gets bigger and (important!) gets influence on a certain market, people no one knows they may just know your thing from the press coverage start joining. Most of them will actually like your idea and do their best to help it spread and grow. Though they may not be as self-critical related to the quality of what they contribute, since, hey, no one knows them and they won't be held reliable for anything. (An issue which also relates to most types of anonymous customer survey, by the way.)
And some of these folks may even have nothing at all in common with the ones recruited by the original spread of the idea. They follow their own goals, may even be competing with you. So they will traditionally start to figure out your weaknesses or how they could rope you in for their very own purposes.
Unfortunately on a worldwide scale as on the web it's very hard to tell this last sort of people from all the others.

Control

Since we are living in a big bad world ;-) trust is an important issue to almost anything you do. But as you know, trust is decreasing with added scale - so you are deploying control-mechanisms to get you through your day on a somewhat assured basis.

With People You Know…
…this is easy. If someone you know or have at least been introduced to causes you harm or tells you lies, you can just confront him with his or her behavior right the next day. Also your trust towards the one who may have introduced this person to you may be affected. This procedure works for most everyday challenges on whom to trust. (Greetings to all Social Networking sites, at this point !)

With People Who Know You…
…this starts getting difficult. If the number (see 'scale' for details) is small enough, there may be enough people sharing your mindset, that may actively take in account somebody they may o may not know, if they realize him or her doing something that doesn't goe with the community's intend. But of course the number of such encouraged interventions decreases with increasing anonymity and over time (since usually no one is rewarded for the courage of his convictions).

So When It Comes To People No One Really Knows…
…you have a serious problem.

Therefore you may want to:

LEAVE EVERYTHING RUNNING AS IT GOES
Easy and cheap, but in most cases not an option if you want to keep up a functioning community.This way Wikipedia would probably end up as a library of contemporary spam an issue that Google is currently experiencing with their Blogger service.

ESTABLISH TOTAL CONTROL
A temptation for many politicians: Watch, check and verify anything and anybody before they get anywhere. Unfortunately this would not only be at the cost of any free society or community but simply unaffordable in most cases. Also experience from totalitarian countries tell us, that people will always find a way to get around the system. So if we want to keep Wikipedia running, this isn't an option either.

ESTABLISH PARTIAL CONTROL
Executing traditional law is a good example for this. You tell the people what it is about, occasionally remind them (like putting up traffic signs) and let the police do occasional checks (even though their actual 'hits' may be on the wrong ones) to prevent all kinds of people from ignoring the rules.Though this may be a proven procedure, 'occasional' correctness-checks may (though suitable on many purposs) just not be enough for a fact-library like Wikipedia.

BUILD A PARADISE FOR THE BELIEVERS AND THE FRIGHTENED
Building a high fence around your housing estate, putting up weaponed guards at the entrance, keeping you save inside and bad world outside, to many seems to be a considerable approach. Unfortunately these checkpoint will keep anything outside you don't alredy own: new impressions, cultural diversity, word of mouth, creative happenings for short: If you want to get new experiences or information, you will have to get outside.
This is why Wikipedia's dual-version proposal to me seems more like stiring up the patient during the visiting hours, rather than curing the disease. It indeed may currently deliver a bunch of more trustworthy results, but at the same time, it will apply the brakes to wikipedia's most powerful advantages; as there are topicality, completeness and being a good and up-to-date - starting point, which will loose its attractiveness when getting 'frozen'.

So what then ? Well, how about just making the world somewhat smaller ?
Since all of the issues mentioned are not new to the world, but occur anywhen trusted information is dealt with, there have been innovative approaches e.g. from the trade world.
Imagine yourself getting into some foreign shop to purchase something. If you pay the owner chash he will be likely to accept, since he can verify your payment imediately. But what happens if you don't have enough money with you ? You will not want to grant the shop owner general access to your bank account… Of course, you can always promise to pay for your shopping somewhen later, but how will the owner know if he can trust you ? He can't.

Until you draw out your credit card.

He will take your card, use a qualified electronic procedure to verify it and eventually accept you for credit. Even if he doesn't know (and mabe not even trusts) you at all. But he trusts your credit card company.

The little electronic device, that checks your credit card for validity, makes the difference: It shortens your separation in trust from the shop owner to - for this particular purpose (buying at a shop) - acceptable two degrees.

Let's apply this to Wikipedia, now.

The 'being free' aspect of Wikipedia is very important to its success.
But proven facts and freedom to say what you think do not always go together well, since a fact as itself rarely cares about the opinion you might or might not have on it.
For the quality of Wikipedia articles it is essential, that someone e. g. from a less liberal country cannot be tracked down for writing in his knowledge there. On the other hand, as we have seen, being to anonymous can end up to be dangerous for the entire project.

For the short term the actions that are already out for testing at Wikipedia currently may indeed suffice.

For the long term I'd propose a pseudonymous registration for Wikipedia and quite a bunch of other occasions where respecting the user's privacy is crucial.
Basically this works just like your credit card, only that it does not tell about solvency but about trustworthyness (actually a bit like the karma-points you get from the other users on certain web-forums). The issuer of these certificates will verify you in person and then guarantee on behalf of your trustworthyness towards others, just as the credit card company does for your payments, but without telleing the person opposite who you are in real life. It's your verifyable, digital ID if you like it that way.

Who may perform as the ID issuer ?
Probably not a state official or your bank or anyone else, who could have an interest in putting you under pressure.
For business purposes this might be a good job for your employer or the chamber of commerce. On other occations founding a specialized non-profit institution may be the action of choice.
This may also be encouraged by politcal action, as many countries of the European Union have already executed the directive of setting digital signatures equal in law to the traditional handwritten ones. This will make it quite easy to apply for a pseudonymous digital ID remotely even in other countries where it will be unlikely that someone gets non-permitted hands on your real identity.

O.k. as a comment this has now been quite lengthy, but I was just hoping to hit more of the relevant details by taking the big round-up.

[Patty Seybold]

Angus,
Thanks so much for calling this to my attention! I have been tracking Lego, of course, but missed this!

So good to know that Mindstorms is back and it's a great story...

Patty

[Angus McDonald]

Patty,

I wonder if you have seen the preview Wired article about LEGO Mindstorms NXT. It appears that LEGO went to their customer base to get some Mindstorm fans to help them design the next version of that product line, and even ended up incorporating hardware developed by those fans into their new product.

The article is here:
http://www.wired.com/news/culture/0,69946-0.html?tw=wn_tophead_1

This is mainly customer as collaborator, with a bit of customer as supplier thrown in. The guys involved did it for some free prototype LEGO bricks, and not much else!

[Patty Seybold]

Dana,
Thanks for the link to Tony Long's post on Wikipedia about "Your Right to be an Idiot." I don't think he has an axe to grind. We're really saying the same thing. Free speech is important, but.. there needs to be some caveat and some mechanism for fact-checking.. I'm hopeful that Wikipedia may get there with the two-version idea.. one fact-checked/published/static and the other dynamic and in process...

Ironic about the broken post/feedback form but hey, we all show our dirty laundry some times...

Patty

[Dana Richardson]

Hi Patty and brilliant readers: All these issues about the "Wiki" are so important, it really got me to wondering about this, and then I saw this article by Tony Long - of Wired -

http://www.wired.com/news/politics/0,1283,69903,00.html

Which really raises an interesting question - here is Tony Long, online for Wired, discussing the validity of a "Wiki" - would you wonder if he's asserting superiority of editorial review at Wired versus "Wiki"? Don't know if you will find this amusing or not, but when I went to the comment input form at Wired to ask Tony about this very thing...it errored out- off this URL:
http://www.wired.com/support/feedback.html

giving forth this error message -

"Fatal error: Call to a member function on a non-object in /usr/local/apache/htdocs/lenya-cache-dir/support/feedback.html on line 40"

Hope you won't think me daft, but I found that sort of ...amusing?

[Patty Seybold]

Ben,
Ahh.. so here's a really interesting point. Who is a customer? I define a customer as the person who USES or CONSUMES a product or service. (They may or may not pay for the privilege of doing so, as in a corporation, where the end-user may be using a resource provided by the corporation, or in the case of the client for a not-for-profit organization).

As I research many of the examples of customer-led innovaton, I find that there are often multiple customers. The folks whom Eric von Hippel refers to as LEAD USERS.. these would be the people who contribute to creating and editing Wikipedia. And the END-USERS, e.g. the readers/researchers who use Wikipedia as a first port of call in finding out about something.

Here's another, similar example. Who are the customers for a scientific journal? My colleagues in that field assure me that the authors -- the people who create the content for the journals -- are, in fact key customers. Getting published in a journal is very important to them. And journals compete with one another for the best authors. (Most authors do in fact pay for the privilege of being published.) The other key customers are the readers--many of whom are the colleagues for whom the scholarly article is of value, to learn from, build on, or rebut. The librarians who subscribe to the journal are the purchasers, but not the end-customers.

Marketplaces are also interesting when it comes to ferreting out who the customers are. EBay, for example, caters to its largest sellers as its most important customers. In an auction or marketplace, buyers and sellers are, in fact, equal players in co-creating the market. The customers--or end-users of the market place--are both buyers and sellers.

You are correct that many people wouldn't be contributing to Wikipedia if it were running ads or charging for its content. It has a business model. It's a not-for-profit.

Patty

[Ben Yates]

Good post. I do have one caveat, though: I'm not sure I'd consider wikipedia editors (or readers) customers -- the foundation is non-profit and the site is free. It's more of a community project than a business project, and it's questionable whether people would be as willing to contribute their time if the wikimedia foundation was (say) running ads.