Big Oil's Wikipedia cleanup: A brand management experiment out of control

When BP was accused of improving its environmental record on Wikipedia Jimmy Wales stood by the oil giant's practices on Wikipedia. That's because unchecked image cleanup by Big Oil's PR reps - with readers none the wiser - is standard operating procedure for the "Internet's encyclopedia."
Written by Violet Blue, Contributor

When BP was accused of having a direct hand in improving its environmental sections on Wikipedia, readers were unaware that nearly half the corporation's page had been written and vetted by the oil giant.

The news broke just as BP obtained an emergency April 5 hearing in its Deepwater Horizon spill trial in New Orleans, to fight what it calls "fictitious claims" for victim compensation. At stake for the oil giant is blame and financial responsibility for the disaster.

BP Deepwater Horizon Wikipedia

Editors on Wikipedia were strongly divided about how BP had been facilitating changes made to its Wikipedia page. CNET reported that BP's press representative "Arturo BP" was not touching BP's page, but instead relied on other editors to make changes. Arturo BP was as a proxy to vet facts for Wikipedia from "experts within the company." BP was quick to say it was not breaking Wikipedia's rules.

Indeed, companies can and do manage their Wikipedia profiles. Wikipedia page management is often viewed as brand management. What's unclear is where the line between corporate brand management, community and fact and fiction lie. 

However, the Wikipedia community doesn't see BP's page management as so black or white. Editors fought each other on Wikipedia discussion pages about the nontransparent corporate content; were the rules really being broken while they were technically being followed? 

Wikipedia founder Jimmy Wales responded by loudly defending BP's actions. Wales fired off blistering statements saying the oil company was behaving exactly according to the way Wikipedia's rules are supposed to work. Wales specifically outlined that BP's use of Wikipedia was exemplary, 'going above and beyond what is required in order to be very clearly in compliance with best practice." 

Chevron tried, and Wikipedia failed 

The results of BP's "best practice" deserve a close look, along with another big oil company working on Wikipedia: Chevron.

ConocoPhillips and Chevron have had PR people working on Wikipedia since at least 2009 (working together actually, since Chevron owns Conoco). "ChevronJustinH" - Justin Higgs, Chevron Spokesman and Media Advisor - has deleted his user page, but his record remains and shows a legacy of attempting to act in good faith within Wikipedia's rules - that has gone largely unchecked.

Beginning in February '09 ChevronJustinH clearly identified himself and suggested an edit on the talk page of an article about one of Chevron's refineries.

When no one makes the edit, he does it himself.

Higgs doesn't do very much more until April 2011, when he starts working on the main Chevron article. The pattern is the same; he posts suggestions on the talk page, waits for feedback (sometimes weeks, sometimes days) and then makes the edits to the article.

If there is feedback or indication that anyone is reading ChevronJustinH's suggestions, it is not in the open. This goes on until March 2012, when he posts a message to the "Wikiproject Energy" talkpage, actually asking for help reviewing his suggested edits.

Hello all, Justin from Chevron here.

For a while now, I've been posting to the Chevron Talk page about improving the accuracy of the entry. During this time, I've been a little surprised that I've been the only active user there. I've posted edit requests for nearly a year and hadn't received responses. After time passed and my notes went ignored, I went ahead and addressed minor inaccuracies that were factual in nature.

Given, the recent discussion that's been taking place about how companies interact with Wikipedia, there's a bit more clarity around the escalation process. Thus, I'm putting out a humble plea for COI [Conflict Of Interest] help.

(...) in an effort to further improve the entry, I'm asking for someone (or group of folks) without a COI to adopt this entry and work with me to continue to better the page.

He gets no response.

The next time Justin from Chevron suggests anything he puts his suggested changes on the userspace subpage belonging to ConocoPhillips Digital and Social Media Senior Analyst Sara Orsi (Saraorsi), and again posts a message on the talk page.

After three years of Chervon's PR lead asking for help, one editor finally responds and fixes up how some of the references are formatted. That same editor, Lexein, engages Higgs in discussion on the talk page about the neutrality of the proposed edits - the discussion has not developed since February 22.

While this is a welcome sign, and Higgs endeavored to act in good faith while getting nowhere pleading Wikipedia editors for help, Chevron's Justin Higgs has been hands-on editing Chevron's Wikipedia Page for nearly three years.

Chevron's Justin was up against a system that had long ceased to serve both Wikipedia article subjects, and Wikipedia readers - who believe the Chevron page is not produced and maintained by Chevron. 

BP tries to find the line

In what Jimmy Wales cites as the example of best practices in the case of British Petroleum, the experience is much different.

The "Prudhoe Bay" subsection of the BP Wikipedia article (one of nine that include rewrites by Arturo BP) is the most recent change to be incorporated into BP's Wikipedia Page. (A tenth was under discussion when CNET reported on BP's Wikipedia involvement.)

In Alaska's Prudhoe Bay, BP had an enormous oil spill in 2006, and then another one in 2009, among other incidents in the area. Both Alaska state and U.S. federal government filed criminal lawsuits against BP; the oil giant was put on probation for the 2006 spill and faced probation-violation charges for the 2009 spill when according to the Wall Street Journal, "The [U.S.] government said the terms of BP's probation included improving maintenance and safety of the Prudhoe Bay pipeline system and alleged that BP failed to comply."

On February 25th 2013, Arturo at BP posted a new version of "Prudhoe Bay" in his userspace, and started a discussion on the BP talk page about replacing the current subsection with this draft.

After comments from a single Wikipedia editor, the "Prudhoe Bay" section was replaced in its entirety by content written by Arturo BP on March 1st - with no suggestions having been made to change any part of Arturo's text (Although a note was added to the talk page after the BP-Wikipedia story broke in the press March 21st, identifying Arturo at BP as a "connected contributor," there was no such indication at the time).

One look at the previous version and the BP version, and it's easy to see that this isn't just a rearrangement of a few sentences or words changed here and there through negotiation with Wikipedia editors - this is a wholly new version.

The original version of "Prudhoe Bay" deals with three issues - a major oil spill at the Prudhoe Bay site, a water leak at a separation plant, and a methanol leak. The new version also covers the three issues, but the water leak is gone and a second, smaller oil spill now has its own paragraph.

In Arturo's version currently on the BP Wikipedia page, a confusion between the major leak of over 200,000 gallons in March 2006 and the minor leak of under 1000 gallons August 2006 is cleared up.

The newly added claim that "there was no impact upon wildlife" is sourced to an Alaksa Department of Environmental Conservation spill report that says only "No impacts to wildlife have been reported at this time" ("At this time" being just over a week after the spill).

The May 2007 water leak section is now gone. The missing section had read,

In May 2007, the company announced another partial field shutdown owing to leaks of water at a separation plant. Their action was interpreted as another example of fallout from a decision to cut maintenance of the pipeline and associated facilities.[261]

Arturo clearly points out he removed the section, saying the removal was due to the item's source (a USA Today article) not actually mentioning the leak, and it "did not have any environmental or safety impact."

However, that citation was being used - albeit poorly - to anchor an important piece of information about BP and Prudhoe Bay: the widely-reported assertions by investigators at the U.S. House Energy and Commerce Committee that BP had "cut maintenance of the pipeline and associated facilities" and launched a criminal investigation into whether the maintenance cuts had lead to the spills mentioned earlier. According to CNBC, Committee leader U.S. Rep. Bart Stupak said the water spill, "was a further sign that BP cost cutting was to blame for the poor state of infrastructure at Prudhoe Bay." The water spill was after Guardian UK reported that the congressional committee had, "demanded an explanation from Britain's biggest oil company for documents suggesting that managers considered turning off the flow of anti-corrosion chemicals to save money."

Also not included, from an Alaska Dispatch timeline from 2011:

Nov. 9, 2009: An 18-inch flow line ruptures at BP's Lisburne field, spilling nearly 50,000 gallons of an oil and water mix onto the tundra about half a mile from Prudhoe Bay. Warnings, including sensors that showed drops in temperature and even alarms, began going off but BP operators failed to investigate or troubleshoot the cause of the alarms for months.

The October 2007 methanol spill is handled differently in each version. In the original, methanol is identified as "poisonous to plants and animals" and its use is explained (it is used to clear ice).

Finally, in the current (Arturo at BP) Wikipedia version this information has been removed and there are new quotation marks around the words "toxic spill."

As of this writing, the Wikipedia editor who approved of Arturo BP's "Prudhoe Bay" section has taken a closer look as well, now regretfully calling the new version "far from accurate."

In light of the experiences had by both Chevron and BP, perhaps there now should also be quotation marks around the word "rules" at Wikipedia.

Is objectivity possible, or lost?

If it's true that Wikipedia currently provides 88% of Americans with their knowledge, then this is a much bigger problem than two big oil companies rewriting their own public histories within an extremely inconsistent system.

Wikipedia's struggle with objectivity, between brands and community, is a curious experiment of our generation. Wikipedia subjects should have the right to manage their public information and branding, and editors must have the right to call BS on them - if the editors are truly objective, and if the guidelines are evenly practiced and policed. But that's not what's happening here, and it doesn't appear to have been happening on Wikipedia for some time.

For the bemused outsider, it's not unlike observing thoroughbred breeders: the trick is to closely supervise the genetic mixing between horses with strong traits, to retain the good traits and filter out the bad. Unsupervised, close inbreeding magnifies the weak points in addition to the strong points. But in this case, we don't get horses that are really fast and really crazy. Instead the result is a closed and broken culture where the quest for objectivity is nothing more than an open content pissing match, led by a bellowing, uninvolved landed digital gentry from a bygone Internet era.

The one who has truly lost here is you, the reader. Unless someone can gain control of Wikipedia and clean house, the "Internet's encyclopedia" will disintegrate into an intellectual bag of candy mixed with hand grenades, and be remembered by those actually seeking facts as a historical curiosity; once brilliant as the glittering, refined resource that once made it so valuable.

ZDNet has reached out to BP and Mr. Wales for comment and will update this article upon response. Post image: U.S. Coast Guard working to put out BP Deepwater Horizon fire, Wikicommons 2010.

Editorial standards