Unless you are a member of the small fellowship of netizens who keep a Google News search for “Wikipedia” bookmarked, chances are you missed out on a truly strange but totally real press release last month from a software company called InstallAware. Its headline: “InstallAware, the Only Alternative to InstallShield, Fails to Get Its Wikipedia Article Published Despite Years of Trying”.

Another reason you may have missed it is because no one picked it up. As far as I can tell, The Wikipedian is the first to write about it at length.^[1]And, this being The Wikipedian, I do mean at length. If it’s a failure as a press release, it is a fitting capstone to a failed attempt on the company’s part to attain the exalted status of mattering on the Internet these days: having a Wikipedia entry of one’s own.

Before I go any further, I actually want to recognize InstallAware for trying, however imperfectly, to do the right thing. Instead of trying to sneak an entry into Wikipedia, they used the Articles for Creation (AfC) process as it is intended. Wikipedia has a big enough problem with anonymous PR activity that—regardless of other mistakes the company and its consultants made along the way—Wikipedians should be grateful they tried to follow the rules and use the appropriate channels.

That said, this is a hot mess of a situation. Eight times since September of last year, using two separate Wikipedia accounts, InstallAware has submitted a draft entry, in slightly different versions, to AfC for review. Eight times they have been rejected, with some reviewers offering a couple of jargon-laden phrases to explain the reason, or nothing more than the required template. Most companies in this situation would slink away, dejected and angry. InstallAware seems to feel that way, too—but took a different tack.

♦     ♦     ♦

The press release itself suggests a blithe unawareness of InstallAware’s position on Wikipedia, not to mention whether anyone would care. It includes cringe-worthy bravado such as “InstallAware’s significance is beyond question” and a ham-handed critique that “Wikipedia is out of touch with its original egalitarian ideals,” says the company founder himself, and even cites unrelated research by the Wikimedia Foundation’s own Aaron Halfaker in support of its claims. As InstallAware sees it:

InstallAware has been repeatedly frustrated in its efforts to contribute an article about itself to Wikipedia. InstallAware, the largest independent software installation vendor for Microsoft Windows, hired a specialist and conducted months of revisions, which ensured that the InstallAware article had more quantity and quality of citations than InstallShield, a similar product which does have a Wikipedia article.

There are two arguments here: a) InstallAware deserves a Wikipedia entry on its own merits, and b) InstallAware deserves a Wikipedia entry because InstallShield, a rival software tool, has one that contains fewer citations than the one they prepared for InstallAware.

To evaluate the first argument, we must consider the guideline Wikipedia editors use to determine whether a given subject should have its own page: Notability. As far as guideline names go, it’s an undeniably loaded word. If one is told “sorry, you’re not Notable” you can understand why they hear “you aren’t important enough”.^[2]When discussing this topic with non-Wikipedians, I prefer to use the term “eligibility”. It’s a delicate message that is too often delivered with a one-size-fits-all template.^[3]“Don’t template the regulars”, says an essay familiar to Wikipedians. I think this is backward: don’t template the noobs.

The generalized version of the Notability requirement^[4]“If a topic has received significant coverage in reliable sources that are independent of the subject, it is presumed to be suitable for a stand-alone article or list.” is very broad and susceptible to interpretation based on wishful thinking. But it isn’t meaningless: it basically says that Wikipedia outsources its judgment about a topic’s significance to sources it considers reliable, which must have written about the subject more than once and with enough information to write a satisfactory entry. When it comes to extant organizations, this often means mainstream and industry news publications.

To apply this standard, we must consider the draft itself.

The draft is at least an honest attempt to reshape a press release into something resembling an encyclopedia entry. It simply tells the company’s story, plus contains some additional information about its software products. Of course it still makes lots of mistakes: toward the end it reads increasingly like a brochure, offering simply too many product details for an encyclopedia. It actually boldfaces InstallAware like that, which is pretty silly in the press release, and completely absurd for an encyclopedia entry. It’s almost a surprise there aren’t little ® symbols after the name of each product.

Those problems can be dealt with. The problem we can’t is the sources. The very first draft cited a couple of press releases, and called it good. Once InstallAware was informed they needed better references, what else did they do but add more. Unfortunately, these weren’t much better and the long list now included only makes them look desperate. Among the sources included: InstallAware’s own website (several times), the founder’s own resume, websites of InstallAware partner companies, SEO zombie sites, even Wikipedia itself.^[5]In Wikipedia circles, this is just about the funniest mistake one can make. One thing is definitely missing: serious coverage from recognized news publications. As one Wikipedian put it:

Sorry but it seems the currently listed sources are simply not enough, what is needed is solid in-depth third-party sources such as news (any time of news is acceptable except press release and trivial passing mentions). If there’s not enough, then there’s simply not enough for a solidly acceptable article.

That sounds right. I ran my own search, on Google News and Lexis-Nexis, and I just found press releases. Curious whether they were aware of these issues, I reached out to InstallAware via the media contact listed on the press release. Over the course of a few polite if pithy emails, I got a better understanding of where they were coming from.^[6]Among the questions I asked: Did they know about the WMF Terms of Use disclosure requirements? No, that one had escaped them, and earlier this week the consultant added a disclosure notice to the draft discussion page. An unfortunate error, but a fair course correction. What was their goal with the press release? “To pressure Wikipedia into accountability and rationality.” Which, fair enough.

Did they know about Wikipedia’s Notability requirement? Yes, they said, and pointed me back to some of the citations to InstallAware’s own website I had ignored. It turns out several of the articles they believe support their eligibility are quite old and no longer online. To make them available for inspection, InstallAware simply scanned them and posted them to their own website, without making this clear to anyone.^[7]A simple rule of thumb: don’t expect Wikipedians to read past their initial assumptions, let alone spend time correcting your mistakes.

Did they attempt to discuss the article with Wikipedia editors? Yes, the specialist—whose expertise it turns out is SEO, not Wikipedia—had posted messages on an involved editor’s discussion page, and also the AfC help desk. Click through and you’ll see an unformatted wall of text that is a chore to read. So, guess what: no one read it.

You can find InstallAware’s collected list of sources here, and evaluate them as I have. I think you’ll find, as I did, that the longest article was written by the founder himself, while others are republished press releases, brief mentions in blog posts, pages on commercial websites, and a short product review. There is one bylined Microsoft publication that might be useful if better sources also existed, but it’s still a borderline call. Overall, it is not: InstallAware does not have sufficient coverage to meet the Notability requirement.

Let’s turn to InstallShield.

What’s interesting is that it sure looks like InstallAware has a point here. Indeed, the entry for InstallShield has only two citations, and one of them is actually a press release. The page was flagged more than four years ago for requiring additional citations. Otherwise, the page seems appropriate enough. It isn’t excessively detailed, and it’s reasonable to guess the article was created not by the company that sells it, but by Wikipedia editors who knew both about the software and also how to develop an entry. In fact, the entry has existed since 2004, long before Wikipedia was a place to be seen.

But that doesn’t seem fair. Should the page be retained simply because it has been around so long? Is the page somehow more acceptable because it was probably created without a marketing goal in mind? In strict terms the answer should be no, but in reality the answer is often yes. There is, in fact, a double-standard for content^[8]and behavior, but that’s another post on Wikipedia: what the policies and guidelines say is ideal, and what Wikipedia editors will actually tolerate. This double-standard is consecrated by the long-established and completely necessary, yet unofficial compromise rule, called “Other stuff exists”, which basically says: just because we have some bad articles that is not a rationale to create more bad articles. InstallShield has been, er, shielded by these circumstances.

Then I ran the same search for InstallShield as I had for InstallAware. The results did not bolster their argument. Although the InstallShield entry contains inadequate citations now, they definitely exist. Some of the stories are quite old, so they are not online, but it’s my opinion there is enough substantial reporting to justify their inclusion. There’s Crain’s Chicago Business in November 1997 with “Installation-software firm set for leap into corporate arena: raising money to push beyond vendor market” and InfoWorld with “Installation software vendor to ship enterprise version” from June 1999, and more. The software has received less press recently, but the snarky IT news site El Reg has mentioned it twice in news stories this year. Taken as a whole, it’s my professional opinion^[9]this note is either an appropriate disclosure that Wikipedia engagement is a key service at my firm, Beutler Ink, or an unsubtle advertisement that InstallShield meets the Notability requirement. “Other stuff exists” need not apply.

But there’s something else going on here. Most of InstallShield’s coverage is from the late 1990s, when they were an independent company putting out tons of press releases in an era with many more technology magazines being published than exist today. Do you think it’s possible that InstallAware is at a disadvantage because of the declining journalism industry, to say nothing of the utility of the press release? I do! As noted above, Wikipedia itself outsources many content decisions to the judgment of working journalists, of whom there are fewer than ever. Then again, maybe InstallAware just isn’t very interesting.

♦     ♦     ♦

None of this is to say all of the mistakes are on InstallAware’s side. Their errors are specific, attributable to individuals, and therefore simple to point out. What happened here was a failure of communication on both sides, and Wikipedia’s mistakes are long-term, systemic, with a collective responsibility that is all too easy to ignore.

Herewith, the most important mistakes I believe made on both sides:

• No one at InstallAware, and neither their specialist, bothered to learn much about Wikipedia. Neither disclosed their conflict of interest nor made any kind of introductory statement to the community about their intention and perspectives. When they finally did try speaking with editors, they didn’t keep it brief, and they didn’t follow standard conventions.
• On the other side of it, Wikipedia editors didn’t immediately offer useful feedback. Instead both InstallAware accounts received only templated messages. Even if they hadn’t responded, Wikipedians never tried to engage on a human level. Yet I noticed something else while researching this: in 2006 the consultant was in a similar COI situation, and at the time received a friendly response from an actual Wikipedia editor. Ten years later, Wikipedia is less hospitable.
• InstallAware was unwilling to reconsider that the sources it proffered actually fit the standards Wikipedians ask for. Over the past several months, at least one Wikipedian declining their submission enumerated the ways in which the various sources were insufficient for the purposes of establishing Notability. Maybe these justifications seemed arbitrary, but they aren’t, and InstallAware should have educated itself after the first couple of rejections.
• The Wikimedia Foundation has not done enough to publicize its disclosure requirements, added to the Terms of Use in 2014, to industry professionals. Likewise, Wikipedians haven’t made this clear enough. Even though the disclosure requirement is featured prominently at AfC, it’s hard to fault anyone for overlooking it. Wikipedia has so many points of advice, it can take years to get up to speed.
• Independent PR consultants take on too many projects they’re not actually qualified for. Relatedly, advice to companies: don’t hire SEO consultants to run a Wikipedia project. SEO and Wikipedia occupy adjacent spaces, as Wikipedia is famously a top Google search result for nearly everything. But the actual knowledge and skills involved in one or the other are vastly different.
• A successful Wikipedia consultant spends less time looking for ways to make something happen—the creation of a new article, for example—and more time looking out for things that may cause it not to happen.



• AfC doesn’t work when submitters don’t know anything about Wikipedia. I’ve heard a lot of complaints about AfC from Wikipedians over time, and they’re right that it often doesn’t work very well. The InstallAware situation is just a more emphatic version of the usual problem. Submitters don’t know what they’re doing, and Wikipedians are too busy to bother with bedside manner.
• Likewise, Wikipedia should define more clearly which kind of publications it considers appropriate for verifying notability for extant companies and organizations. Besides the General notability guideline (GNG), applicable Wikipedia guidelines such as Notability (software) and Notability (organizations and companies) offer some wise and unavoidably vague advisories. After all, Wikipedians can’t anticipate future situations which might be ill-served by too-specific rules. But it could be clearer, and it might not hurt to include some examples of acceptable sources and why they are.

Ultimately, the most salient issue in this whole kerfuffle is that InstallAware was unwilling to take “no” for an answer. But a close second is the fact that Wikipedia editors made only a half-hearted effort to communicate with them, and third is that AfC is just about impossible to navigate for anyone. InstallAware is only one company, and if this is actually the end of the road for them at Wikipedia, there will still be countless more companies asking for an entry after them.

Working with companies may not be what gets Wikipedians out of bed in the morning, but so long as the site remains one of the Internet’s top destinations, and maintains its famously low barriers to entry, it’s in Wikipedia’s best interest to improve these processes. Yes, that process is still mostly going to reject drafts of articles about companies, who won’t be happy about it. Hopefully, they won’t feel like they have to write press releases.