William Beutler on Wikipedia

Posts Tagged ‘Conflict of Interest’

Future of Wikipedia, Wikimedia Foundation, Wikipedia in the news

How to Stop the Next Bell Pottinger

Tagged as , , , , , , , , ,
on December 12, 2011 at 10:57 pm

I’m somewhat late by now to one of the bigger Wikipedia-related stories to come along in recent months: the revelation of secretive Wikipedia edits by a London-based PR firm called Bell Pottinger. As reported by the BBC and The Independent and others, Bell Pottinger was caught airbrushing client entries, adding promotional material and removing critical information. Of course, the company’s own Wikipedia profile is now disproportionately about this incident, at least for the time being.

In a swift and thorough investigation, Wikipedia’s volunteers determined that Bell Pottinger employed at least ten accounts, and probably more, to edit more than 100 separate pages. These changes included adding “promotional/excessive language”, including “puffery” and in some cases “unambiguous advertising” by accounts with such innocuous-sounding names as “Biggleswiki”. (Ask not for whom the Bell Pottinger tolls, it tolls for Biggleswiki.)

In spite of myself, I was amused: why is it that supposedly smart, sophisticated PR professionals seem to think the best approach to Wikipedia is duplicity?

Problem is, I think that narrative may be driving the response a bit too much. While the coverage has been mostly responsible, noting that Bell Pottinger committed “possible breaches of conflict of interest guidelines”, it is easy to come away with the impression that any interaction with Wikipedia articles by interested parties is inherently illegitimate. Not unlike the widely-reported incidence of U.S. congressional staff edits to Wikipedia in 2006, or similar incidents uncovered with a tool called WikiScanner in 2007, it ends up stigmatizing editors who would make legitimate edits.

The BBC writes: “While anyone is free to edit the encyclopaedia, the site’s guidelines urge users to steer clear of topics in which they have a personal or business interest.” This is not true for personal interests, and while true for business interests, anyone who knows the site well also knows that it is not the full picture. At least the BBC also quoted Wikipedian David Gerard, noting the investigation would focus on whether the edits were carried out in “bad faith”. More Gerard: “We’re having a close look. What the team is going to do is look at Bell Pottinger’s clients and see what edits have been made.” It so happens these details actually do matter. And even Jimmy Wales, amid more forceful denunciations of the bad actors, told The Independent: “There are ethical PR companies out there.” Not that you ever hear about them.

♦     ♦     ♦

As some readers will know, I’ve long been interested in the topic of COI (“Conflict of interest”) editing at Wikipedia. I don’t spend a great deal of time dwelling on the topic here, but indeed it has been a professional focus as well. Over the past few years I have developed best practices for clients, mostly large companies and organizations with existing articles, to facilitate the improvement of those Wikipedia articles in a constructive manner, following Wikipedia’s rules. As noted on the About page of this blog: “My goal has been and will always be to improve such articles while working within consensus.” I’ve carried many of these on my back—these projects are not difficult to find—and helped clients engage under their own name as well. I’m proud of all these, not least because so many find it so surprising.

It shouldn’t be this way. Earlier this year, I teamed up with creative agency JESS3 and marketing automation firm Eloqua to produce a “white hat” guide for marketers and business professionals titled “The Grande Guide to Wikipedia”—a how-to for constructive interaction with the Wikipedia community. The feedback was positive, but I heard more from Wikipedians than from marketing professionals. I have no doubt that furtive, undisclosed edits are common at most firms, not because they seek to do harm (like Bell Pottinger), but because editing transparently seems like too much trouble.

Another reason, and I want to be careful here, is because statements by Jimmy Wales have created the impression that anyone who works for a marketing firm is unwelcome. This goes back to the business involving Gregory Kohs and the MyWikiBiz controversy, where Wales’ “shoot on sight” comments remained effectively the only quote on the matter for a long time. Kohs, openly hostile to Wikipedia and vocal about his intent to subvert Wikipedia was, for a long time, the only model. No doubt this unfortunate turn of history kept well-meaning COI editors in the shadows.

But I’m not alone in thinking that this needs to change. Recently, a social media marketer named David King wrote a very good blog post titled “Why Wikipedia Needs Marketers”, which included this astute observation:

The volume of [Wikipedia] content is growing, but the active contributors to maintain, update and police those articles is shrinking. As this trend continues, vandalism, bias, outdated information and blatant factual errors will run even more rampant.

Marketers are the most motivated to maintain Wikis on subjects important to them and invest the time in providing quality, well-verified content. We can fill this gap if we can learn to support Wikipedia’ s encyclopedic goals and follow the rules.

The response to his post was, perhaps surprisingly, very positive—with encouraging replies in the comments from respected editors including Lori Phillips, FT2 and Wikimedia Foundation reader relations head Philippe Beaudette. King was subsequently invited to expand on the theme at The Wikipedia Signpost, where he continued:

COI contributors introduce bias, but I’m also concerned of the bias without them. Some of our most knowledgeable and motivated contributors are COIs. Does that mean we open the doors wide? Absolutely not. COIs are like political lobbyists. We’re needed but our participation needs to be a delicate and well regulated one. But through teamwork, education, awareness, process, a better ecosystem we could change the tides.

I half-agree with this. I think the analogy of lobbyists is incorrect; “COI editors” should self-regulate their own contributions, as Wikipedia’s Conflict of interest guideline itself says: “Where advancing outside interests is more important to an editor than advancing the aims of Wikipedia, that editor stands in a conflict of interest.” Conflict of interest is not fait accompli; a conscientious editor can and should acknowledge the potential for conflict of interest, and take steps to mitigate that. This should include seeking consensus for making edits outside of what the COI guideline describes as patently “non-controversial edits”.

But he’s right that such edits should also be well-regulated, although they are not now. In practice, following the advice of the Paid editing essay and seeking consensus at the Conflict of interest/Noticeboard (COI/N) or at various WikiProjects can present significant delays, another non-trivial obstacle for marketing and PR professionals who might then choose to just edit without providing adequate disclosure.

♦     ♦     ♦

David King is also right that there needs to be a better ecosystem, both to support and to regulate such editing activity. But such a system is unlikely to happen on its own. The answer may lie in an accommodation not unlike the one that accepts the role of ethical PR professionals on Wikipedia. To wit: although the spirit of Wikipedia is for it to be volunteer-edited, there are cases where COI editors, whether paid representatives or smart employees, can help address problem areas with certain articles. Likewise, the Wikimedia Foundation plays no role in setting editorial policy, but it can and should play a role in facilitating responsible COI activity.

There are good, active editors at COI/N who frequently catch bad actors (and infrequently help good ones) but unless their ranks are expanded significantly, they would have a difficult time handling the volume, were marketers to wise up and learn to follow Wikipedia’s rules. Why not help them out?

I suggest that a model already exists: through outreach efforts described in the Wikimedia Foundation’s Strategic Plan (PDF) and embodied in the Wikimedia Ambassador Program, resources could be put toward meeting PR professionals halfway. I don’t think the Foundation needs to seek more such editors, in part because they are already here. But it can provide a safe harbor for assistance requests and advice to ensure COI compliance, and make it safe to follow the rules. Yes, there are plenty of how-tos on pages scattered around the website, but if Danny Sullivan is right about one thing, it’s that Wikipedia is confounding to the uninitiated.

Five years ago, Wikipedia was definitely not ready for this. Today I think it is. And I wouldn’t necessarily call it traditional public relations, and certainly not marketing, because Wikipedia is a unique medium with its own rules. I suggest thinking of it as Wikipedia relations, or wiki relations for short. Hesitant Wikipedians should see it as a mark of how far the project has come: while volunteers remain the core of Wikipedia’s community, there is room for professional representatives of outside interests to work constructively in this space.

Returning to Jimmy Wales’ comments above, ethical PR firms and COI editors do exist. With some effort by the Wikipedia community and the Wikimedia Foundation, more can be encouraged, and Wikipedia would be better for it.

Announcements, Wikipedia culture, Wikipedia etiquette

The Grande Guide to Wikipedia

Tagged as , , , , , , ,
on May 3, 2011 at 8:14 am

In line with my cryptic tweet of yesterday afternoon (owing to an early scoop by The Next Web) here’s the big reveal: in the past few months I’ve been working with the marketing automation company Eloqua and design firm JESS3 (with whom I worked on “The State of Wikipedia” video) to write a new entry in their “Grande Guide” series of how-to manuals. Of course, I wrote about Wikipedia: “The Grande Guide to Wikipedia”:

Eloqua Grande Guide to Wikipedia

Because Eloqua’s audience is marketers, they are also the focus of this guide. One of the first (rhetorical) questions raised in this guide is this: “Is Wikipedia a marketing opportunity?” The answer, more or less, is: “No, but…” While trying to use Wikipedia as a marketing tool is one of the surest ways to find yourself in trouble with Wikipedia editors, there are times where it is appropriate for someone who works with or for a company to make positive suggestions and even some non-controversial edits.

This subject makes Wikipedians understandably nervous. As evidence, consider the many tens of thousands (perhaps hundreds of thousands) of words expended on debating the propriety and rules surrounding paid editing—without coming to a resolution. The result is a confusing place where contributors with a financial interest are not exactly welcome, but also not disqualified. It can be very confusing. As Eloqua’s Joe Chernov writes:

It’s also important to note that we worked hard to preserve the integrity of the Wikipedia community throughout our Guide. We aimed to share how Wikipedia truly works, so that marketers can understand and appreciate it – not so they can game the system. We hope and trust that respect comes through in the content.

I hope you’ll read “The Grande Guide to Wikipedia” and, whether you’re a marketer curious about Wikipedia (more than a few of you, I know) or a Wikipedia editor skeptical of marketers (and not without reason!), I hope you’ll learn something new.

Biographies of Living Persons, Influencing Wikipedia, Wikipedia in the news, Wikipedia Optimization

A Potential Supreme Court Nominee Probably Edited Her Own Wikipedia Article. Is It a Big Deal?

Tagged as , , , , , , ,
on April 13, 2010 at 9:30 am

leah_ward_sears_wikiNew York-based media blog Gawker is reporting that Leah Ward Sears, former Chief Justice of the Georgia Supreme Court and potential nominee to the United States Supreme Court by President Obama, edited her own Wikipedia article in late 2008 and early 2009.

While the possibility exists that someone else used her initials, last name and year she became a state Supreme Court Justice as a username, it usually turns out that this type of account is exactly that person. Gawker is focusing primarily on an edit she made that was favorable toward herself:

On May 6th, a user named LWsears1992 edited Leah Ward Sears’ Wikipedia page, adding the clause “Based in large part on her highly regarded record” to a passage about how she defeated an opponent in the 2004 race for Georgia Supreme Court. (Georgia is one of eight states that have the sort of weird policy of electing Supreme Court justices.)

This is technically correct, but not exactly right. While Gawker does have a screen shot of an edit by Lwsears1992 “adding” this, all she did was restore a phrase that had existed on the page since June 2005, added in the first place by a technology consultant in Atlanta. The phrase was removed again a few days later for lacking a source, and Lwsears1992 did not press the case further. Not that Sears should necessarily be making direct edits on matters of disagreement, but these are considerations that few Wikipedia outsiders understand.

In total, Lwsears1992 made 36 edits to Wikipedia, all of them relating to this particular article. So how did she do? Did she make the page better or worse, overall? To find out, I went through each and every edit, starting with the article as it appeared before she started working on it, November 3, 2008 and concluding with the article after she completed her work, on November 13, 2008. Here is what I found:

Better:

  • The fact is that Sears is being called out because she attempted to be transparent about it. However, it’s probable that she made a single edit an hour before her first editing session from the IP address 167.192.61.254 in Atlanta, Georgia. Unfortunately, she screwed up a template, rendering the “Infobox” sidebar a mess of code. But I count this as a positive, because of what happened next. Once she had caused this error, she created an account and undertook the task of fixing it. Not only did she do so, but approximately a third of her edits were devoted to getting this one thing right.
  • She uploaded her own photo, taking the time to release it under two free licenses, the old GNU license Wikipedia used to use for everything, and the Creative Commons license it uses now. She experimented with the sizing of the photo she added, including trying it at full size before settling upon 155 pixels wide, which is the width still.
  • She added useful context, such as noting that her resignation from the Court would coincide with the end of her term; this is unambiguously more useful than simply ending the sentence on “she will resign from the State Supreme Court at the end of June 2009.”
  • Chances are good she made the article sturdier in the long run, changing the article to read that she was the “first” African-American female Chief Justice in a U.S. state instead of the “only” one. Assuming this is correct, the former will always be true though the latter assuredly will not be.
  • She tried to protect her own page from vandalism by experimenting with templates meant to indicate the page cannot be edited in some circumstances. But as she was not an administrator, she couldn’t do this anyway. Once she saw it wasn’t working, she took them down. One could almost file this as a negative, because trying to get a page locked from editing is a sure sign of not understanding Wikipedia. On the other hand, changing your own mistake is a sign that you do. I’ll give her the benefit of the doubt here.

Worse:

  • She didn’t cite any source for the claim she is the first African-American female Chief Justice in a U.S. state, making the claim difficult to verify. Anytime one makes a claim of superiority or “firstness,” it helps to source the claim to avoid the dreaded “[citation needed]” tag.
  • She didn’t provide any edit summaries for her work, making it tedious to click through each and find out exactly what she did.
  • She made some changes that didn’t make the page better. In one edit, she edited internal site links embedded in the phrase “Chief Justice of the Georgia Supreme Court” so that instead of directing people to articles about Chief Justices and the GA Supreme Court, it would go to a non-existent page that she probably assumed existed.
  • She also removed internal links to the names of her appointer (Zell Miller) and predecessor (Norman S. Fletcher) for no apparent reason; she also removed the link for “Chief Justice of the Georgia Supreme Court” — perhaps after noticing that it did not lead anywhere. Odder still, she did replace some of this information, including Miller’s name, but removed Fletcher’s name after having initially sought to add it. In any case, he is back in the full article today.

What is the value of adding her photograph vs. removing the name of her predecessor? What is the value of adding new details which are presumably correct, but not citing independent sources? How bad is it to edit your Wikipedia article without seeking consensus of other editors? How should one seek to change their articles on Wikipedia in any case?

These questions and more like it have been coming up more often in recent months. It’s a subject recently addressed by the Wikimedia Foundation’s Jay Walsh in an interview with PR Week. It’s a subject that others are discussing, from law firms in the UK to PR firms around the world. It’s s a subject I weigh every day as a consultant on matters of Wikipedia, and in an article I just published in Politics Magazine.

My answer regarding Leah Ward Sears is that, she made the article better, but not much. She did not go about it the right way, but the right way is non-obvious to most, and the burden is on Wikipedia to make its rules understood by outsiders. While some of her edits were self-serving, they were of a mild sort. At most this was a venal sin, not a cardinal one. Gawker is turning this into a “gotcha” story on the implied theory that interacting with one’s own Wikipedia article is never acceptable. This is a myth, one widely believed and one propagated by many at Wikipedia simply to keep people from meddling with their pages en masse. This is understandable, but it won’t work out in the long term.

If Sears is Obama’s nominee and is further confirmed to the Supreme Court, perhaps it will help put an end to this kind of “gotcha”. I doubt this is significant enough to come up at confirmation hearings if she is nominated, and it should not be. But I will concede that would be kind of entertaining.

Image via Sears via Wikipedia.

Congressional Wikipedia Edits

Did Rep. Hinojosa Get a Free Pass on Biased Wikipedia Edits?

Tagged as , , , , ,
on March 30, 2009 at 9:21 am

This weekend, former Tacitus and RedState blogger Josh Treviño asked over Twitter:

Do you think Rep. Rubén Hinojosa (D-TX-15) had his staff edit his own Wikipedia entry? So blatant: http://bit.ly/TSbMm

The answer is yes, and yes it is and I see what could be regarded as something of a double-standard here. First of all, Hinojosa’s congressional staff was quite up front about editing the page in the first place, when they contributed from a House-registered IP address in early 2008:

I am a staffer for Congressman Hinojosa in his Washington office and have found some mistakes. We have edited them and added the official biography from our website. If there are questions, please [sic]

Not unlike Twitter, you only get so many characters to explain your edit in the edit summary. This was not a bad way to go about it, although another editor left this note on the Talk page associated with the IP address for the Hinojosa staffer:

ATTN: Staffer for Congressman Rubén Hinojosa

The edit was constructive in the sense that it did not delete unfavorable information from the page. At least, not all that unfavorable. For example, they changed

Finally in 2002 he was elected once again after running unopposed.

to

Finally in 2002 he was elected once again.

which it’s debatable whether this is even the right phrasing, but there is no question they removed (what may or may not be relevant) context.

However, I am not sure why adding material from his official bio is “constructive” when Wikipedia explicitly forbids plagiarism. Using information from the bio would be one thing, if there was at least a citation. Instead, are a few examples from what they added:

In Congress, Rubén Hinojosa is regarded as a champion for the disadvantaged and has distinguished himself as a strong advocate for education, housing and economic development. His primary goal in Congress has been to reduce the chronic unemployment rate in regions of the district.

And:

As chairman of the Education Task Force for the Congressional Hispanic Caucus, Congressman Hinojosa ensures that federal education policy never loses sight of the youngest and fastest growing population in the country – Hispanic Americans.

And:

On the House Financial Services Committee, Congressman Hinojosa is widely recognized as a leader on issues affecting the underserved, from banking to housing.

Favorable impressions of a subject can be attributed to independent observers, but it should never presented as a true-because-Wikipedia-said-so fact. Yet as Treviño noticed, it has largely remained intact in the year-and-change since. In fact, one editor did stop by a few days later to clean it up, but only slightly.

A partisan job? Probably not — that editor was a retired aviation engineer from Bristol, England. Nevertheless, it’s worth asking whether a Republican congressional staffer making these kind of mistakes would have received the same benefit of the doubt? If one takes into account the case of former Rep. Gil Gutknecht (R-Minn.), whose staff was caught in mid-2006 making similar changes to his Wikipedia bio, this seems unlikely. Then again, the staff of former Rep. Marty Meehan (D-Mass.) was called out for doing something similar earlier that year, so the answer is not so simple.

The bottom line here is that Rep. Hinojosa’s page needs some major work to bring it back in line with Wikipedia standards. If nobody gets to it soon — and in fact the page has remained unedited for two months now — I may just have to get in there and fix it myself.

Wikipedia Optimization

Searching for Wikipedia Assistance on Craigslist

Tagged as , , , , ,
on March 22, 2009 at 4:00 pm

Here’s an interesting request for Wikipedia assistance on Craigslist, sent to me by a friend and former colleague from my hometown of Portland, Oregon — coincidentally, also the birthplace of the wiki — just a few days ago:

some wikipedia help (SW Portland) Date: 2009-03-18, 11:01PM PDT I am looking for help with wikipedia. I could really use a good wikipedia editor willing to help and post what seems reasonable on Wikipedia within the rules. The topic is "string matching algorithms and structured data". You need to have a basic undertsnading of this topic. Your need to be an experienced wiki editor and have credit in the wikipedia community regarding such topics.

I’ve clipped a bit from the bottom, but it also includes this:

* Location: SW Portland
* it’s NOT ok to contact this poster with services or other commercial interests
* Compensation: $25+ depends on qualifications.

With the math knowledge requirement and low monetary offer, I am not surprised that the ad remains at the time of this writing.

The mention of compensation could well raise concerns among editors who are wary of financial interests influencing content on Wikipedia. While I am sympathetic to this point of view for the simple reason that they are often correct — people who are willing to put money against getting something changed on Wikipedia are likely to be willing to pay for edits that satisfy their interests but fall short of Wikipedia’s goals — this is also why the Conflict of Interest guideline specifically states: Where advancing outside interests is more important to an editor than advancing the aims of Wikipedia, that editor stands in a conflict of interest. How serious is the advertiser about following this? I’d say the phrase “what seems reasonable on Wikipedia within the rules” has to be pretty close, but what may seem “reasonable” to someone unfamiliar with Wikipedia guidelines may nevertheless conflict with them.

While this request appears to be small ball, it does remind me of the time when a Microsoft employee offered Australian programmer Rick Jelliffe money to edit a Wikipedia article of interest to the company. Presumably knowing he would be sympathetic, Microsoft instructed Jelliffe to use his best judgment, and the controversy only kicked off once Jelliffe himself wrote a blog post about it. Notwithstanding comments from the likes of Jimmy Wales saying he was “disappointed” in the situation, it is unrealistic to expect that interested parties cannot seek to correct inaccurate or incomplete information — which is what Microsoft says it was doing. Lost in the controversy was the possibility that IBM, Microsoft’s rival, may have had people anonymously weighting the article in question.

Ultimately, Jelliffe’s biggest mistake was not disclosing the arrangement on the article’s Talk page at the time of his edits. This may have meant additional scrutiny on the page, but that comes with the territory. And if anyone takes up this guy’s offer, I’d recommend they do the same.