June 2009
By Tom Edwards

Tom Edwards is owner and principal consultant of Englobe, a Seattle-based consultancy for geostrategic content management. He previously spent 13 years at Microsoft as a geographer and as its senior geopolitical strategist.


User-generated content

Let’s say that you’re reading a news story about a particular area of geographic conflict and you decide to investigate further. Without an encyclopedia available, as fewer and fewer of us seem to have them on hand these days, you quickly check out your handy online references. To your surprise, the article on this disputed feature seems to be an amalgamation of strongly differing opinions and ideologies, to the point where the article has been locked down from further editing. Such is the nature of the brave new world of user-generated content, where a content publisher forges a careful alliance of sorts with a wide range of contributors across very diverse locales and cultures. Depending on the intended purpose of the provided content, the end result can take on a life of its own, as it becomes the focal point for a silent yet fervent battle over “fact” and “truth” from divergent viewpoints.

In this new world of social networks, the blogosphere, online communities and the ever-growing notion of “crowdsourcing” factual information, the pure objectivity of content has become even more tenuous and fleeting. Why do I say more tenuous, as if to suggest that perhaps our trusted information resources weren’t objective to begin with? Well, quite frankly, they weren’t. As a cartographer, I can directly attest to this reality, at least as it pertains to maps. I’ve conveyed previously in this column that experienced cartographers realize the maps you read, and often implicitly trust, are actually the end products of a long process of empirical data passing through several stages of subjective human interventions based on a specific purpose, the technology available, the intended audience, and the organizational and/or cultural context. What could be called a “correct” map in one country might be called a complete fallacy in another, depending on the local geopolitical perspectives and even the national laws regarding cartography.

So is a typical map just another subjective, crowdsourced document? No, not at all! The key difference lies in the fact that all those human interventions are conducted by highly trained cartographers and spatial data analysts who are committed to a common goal of representing the “ground truth” as best as possible. The subjectivity in this example primarily comes in the need to generalize the information and symbolize it appropriately. It must be scaled down to fit a usable medium, and there are many methods to generalize and symbolize spatial data.

Thus, in my view, the real challenge of leveraging user-generated content is subject matter expertise and whether the contributed content represents actual fact or just speculation. The issue is really not that the content is being contributed from outside the information provider’s company because information publishers regularly contract for external expertise for specific places, languages and so on.

The issue is more about the unknown level of expertise and the general lack of metadata available on the social web about a specific contributor’s level of competence to properly discern, process and comment on any particular subject.

The processes of writing a book, creating a map or building a video game have worked well for so long, generally with reliable results, because it is a collaborative process of highly-skilled individuals. Their expertise provides the necessary checks and balances. In the “old days,” an encyclopedia was written by a wide range of subject matter experts (SMEs), scholars and researchers, who together created a sense of credibility and reliability. Some cultural differences will no doubt arise even in these situations. I recall working on a digital encyclopedia product wherein users from Korea claimed that all the Asia-related articles were written by the Japanese -— which wasn’t true, but they perceived an obvious bias.

There is absolutely nothing wrong with opinions about facts. Even highly qualified scientists can have widely divergent opinions and conclusions about the same set of data, but they have their educated reasons for the difference. And if that’s the case with quantitative, scientifically-derived data, you can only imagine the degree to which this complexity extrapolates into the more qualitative realm of geocultural information.

However, as we’ve already seen in our dynamic twenty-first century society, the fine line between fact and opinion seems to be getting increasingly thinner. Opinion is really nothing new, as most major news outlets, whether print or broadcast, have always left room for editorial opinion from respected reporters and analysts. This decades-old model is swiftly losing ground to the more prevalent notion of relying upon the collective wisdom of the masses to provide the “truth,” which is the very model upon which some web-based references operate. The idea is that the facts will be put to the test by a wide range of users with vastly different opinions and eventually the truth will be solidified. People will test the contributed content against some level of objectivity and a sense of honor. For the most part, this seems to be working, but areas of stark difference will always prove to be “battleground” topics — geopolitical disputes, religion and so forth. At what point does the collective wisdom of the masses trump the viewpoint of an SME? That actually remains to be seen, as this user-generated information model is still very much evolving.

Mitigating risks

How does all this reflect on the geocultural risk that I’m so fond of discussing here? From my experience, it’s tough enough to deal with geopolitical and cultural content when you’re doing so from a position of expertise. The point of the many guidelines I have conveyed in this column is to mitigate potential risk and minimize negative user feedback. When user-generated content around sensitive issues is allowed, it effectively eliminates the veil between a careful strategy for how a company represents geocultural content and the onslaught of government and public backlash.

In other words, when a company opens the doors to user opinions that may conflict with the target markets, the company puts itself in a precarious position of having to constantly justify this decision to one government and angry local group after another. This is absolutely not to say that user-generated content is to be avoided! The point is that it has its place and its purpose, and I believe more companies need to proactively strategize on if, how and when to leverage it in the context of their products and services.

Thus, if you’re an information provider or publisher who is considering the allowance of user-generated content or if you’re opening up a new avenue for the generation of such, I’d highly recommend at least two pieces of guidance. First, I strongly suggest that information publishers understand the possible roles of user-generated content alongside well-researched content from SMEs. I believe that both forms of information have their purpose and utility, but never one at the exclusion of another. The best user-generated content that I have seen fill in the blanks in areas that are otherwise not covered well - activities, art forms, geography and cultural information on a very localized linguistic or geographic scale. In some cases, I’d implore companies to resist the urge to add features just because everyone’s doing it. This may not be the right move for your content.

Second, as is evident with most user-generated resources today, there is an absolute need for moderation of some form. Even if your intended purpose is to provide a forum for raw opinions, it should be moderated by qualified individuals.

Any content that is posted to your website or resource should be reviewed and approved in some manner. Such moderation could be very proactive, in which no contributed content is ever seen until approved, or it can be more reactive (which is the more commonly utilized model), where contributed content is considered innocent until proven guilty, and if so, is promptly removed. There are many possible models here, but something will be necessary unless your forum for user-generated content is overtly intended to be pure opinion and subjective speculation, as in the ongoing issue of setting a good signal-to-noise ratio in content.

In the end, I’d like to clarify that I’m certainly not trying to discourage content developers and providers from shying away from user-generated content. Quite the contrary — the door that technology has opened into the hearts and minds of millions of users is not something to be overlooked or squandered. The real challenge, however, is leveraging these millions in a way that is meaningful and respectful, both to the contributors of the content as well as your end-users. Despite the popularity of blogs, Facebook notes, Twitter updates and all kinds of random net-based information, there is still something to be said for structure and strategy.