Quora and “The 1-9-90 Rule”

Generally, sites built around user-generated content (UCG) are believed to follow the “1-9-90 Rule of Content Generation”; that is, 1% of users contribute the majority of content, 9% occasionally contribute content, and the remaining 90% rarely or most likely never contribute. The 90% mostly lurk on sites and consume content.

The 1-9-90 Rule isn’t hard and fast. It’s believed that of Wikipedia’s 32m unique visitors, over 99% only consume content. In fact, there are less than 70k active contributors (0.2%)’ and the most active 1,000 contributors (0.003%) account for about 66% of the site’s contributions and edits. For Wikipedia, it’s more like the 0.003-0.2-99.8 Rule. (cite)

Being a global nonprofit fueled by anonymous contributions, Wikipedia is quite transparent with these stats despite its incredible participation inequality. Whatever the reasons why, it seems to work just fine as Wikipedia is a Top 10 Website and has been translated and updated into hundreds of different languages. It has generated remarkable content built largely on the backs of a few people who can write well, footnote text, and organize topics.

In the case of Quora, which in sometimes thought of a social layer on top of fact, what is its participatory distribution among users?

Of course, I don’t know the breakdown percentages because these are kept private, though the company has hinted it was surprised by just how many users only lurk on the site. If I had to guess, my hunch is that Quora is more participatory, relative to Wikipedia certainly, and in general compared to other UCG sites. In fact, it may be as many as 3% of users who do the majority of the work, while 20% only participate occasionally and the remaining 77% users primarily lurk on the site. And, since Quora has given users more ways to participate beyond contributing, editing, and classifying content, as well as enabling anonymity for certain actions, it may be able to create a similar level of quality content as Wikipedia, with the added bonus that primary sources within specific topics can offered verified answers and commentary.

If true, it is certainly an achievement to have created a site that could defy The 1-9-90 Rule and encourage more participation among users. It could have its own 3-20-77 Rule. Quora has spent significant effort to built, maintain, and nurture their community of major contributors who work away on the site answering questions, reviewing grammar, marking threads with appropriate topic pages, and editing question context and the content itself. In return for this free work, Quora developed a status structure that recognizes these contributions. For instance, the site has “admins” who band together to create and enforce site controls, and there are also “reviewers” who review questions within certain topics. (I have personally participated in most of these activities, though I am not a “reviewer” or an “admin.”)

As a topic unfolds on Wikipedia, armies of contributors and editors mold the text and structure until it reaches some level of equilibrium. On Wikipedia, all users are basically equal, so the system doesn’t truly distinguish between someone like me or Martha Stewart when we input content on the Wikipedia page for “Cookies.” In this sense, Wikipedia is largely democratic, where any registered user can edit and others can accept or strike the edits. This churns behind the scenes. The 99.8% of visitors only consume whatever the current result is at that time because Wikipedia is not a participatory environment.

The dynamics are quite different on Quora.

As a thread unfolds on Quora, users ask questions, they are organized, edited, sorted, and distributed into various users’ feeds. Once set, any user can submit an answer and any user can up-vote the content to the top of the page, where it’s more likely to be read. Users can also flag or down-vote content which in effect hides the content from most consumers’ eyes. These actions are recorded in a log tied to identity. Any or all of these actions could be done with someone’s real identity or anonymously, though the admins and reviewers do try to monitor inappropriate behaviors and Quora retains the right to limit or revoke access for site violations.

In order to encourage answers of high quality and/or authority, Quora has created controls on the site to encourage proper grammar and language. Quora has also made it clear it will assign reputation weights to users based on a number of signals, such as up-votes received within a topic, endorsements, or number of followers, among other data points. This is captured in Quora’s controversial “PeopleRank” algorithm, and is designed with the intent to help the best and most relevant content to surface to consumers.

And, so, here’s where things could get tricky for Quora…

At some point, when the site begins to attract new users and encourage them to participate, many within the original 3% who have been so instrumental in creating, installing, and enforcing quality controls that make Quora so great today may have to take a backseat as contributors to others as more domain-specific knowledge comes onto the site. Using the Wikipedia “Cookies” example and Martha Stewart, on Quora that same scenario would virtually drown out the voice of most users who seek to either contribute content or make edits, even if Martha’s content on “Cookies” wasn’t that good, relevant, or accurate.

As the “3%” become slowly marginalized into a more of a editorial role, administering topics, correcting grammar, classifying tags, and voting, the current 20% (myself included, on occasion) continue to contribute every now and then, but often can be intoxicated by “celebrity” contributions, constantly up-voting any content, oftentimes no matter the quality or accuracy. These reactions to celebrity behavior end up distorting information on the site for the 77% who consume.

If the goal of Quora is to be a place for thoughtful questions and (expert or insightful) answers, the real experts will only contribute occasionally and aren’t likely to always be receptive to scores of edit notifications and reclassifications, especially if the person registering those improvements aren’t expert in that topic to begin with. If a new user who is expert tussles with suggestions from non-experts, it may reduce their incentive to participate.

To control for these delicate scenarios, Quora’s PeopleRank eventually will give more weight to answers contributed by those not in the 3% percent, in favor of those who can establish topical authority, because if it doesn’t, the site is at risk of being governed by a small group of 3%, which could turn a public forum into an echo chamber. The danger for Quora, however, is that this could slowly unsettle the “3%” who have built a reputation by contributing so much time and sweat into building the site so far, and without those active contributors, who will keep the system organized?

To be fair, this is a very delicate issue and not easy to manage. Many companies, young or mature, wouldn’t be able to deal with this nimbly while also building a cutting-edge technology company. And, it’s impossible to quickly readjust participation inequality ratios as if it were an unsavory line of code. The only thing a site can do is to decide for itself what the best participatory distribution is for its current and long-term needs, and then create incentives for user to strike that balance.

Can Quora maintain its currently participation dynamic of 3-20-77 as it grows? I’m not so sure. I’d like to think more users will adopt the site and participate, but what is likely to happen is that as the number of unique users increases (it’s currently a small base, relative to other social networks) is that its participation ratios will likely push back toward a more traditional 1-9-90. Thus, the site will become less participatory as its user base grows in number.

Oddly enough, as Quora scales, becomes less participatory, and approaches The 1-9-90 Distribution, the content on the site could actually improve, assuming the company is (1) able to draw enough good experts in key topics and (2) conditioning the 3% to what their roles will be moving forward.

This is the hope. But, this is only theoretical.

In order to get here, the site has to make it easier for busy experts to contribute and to keep recruiting more of the 3% foot soldiers. The site may also have to convince the 3% to take more of an anonymous role in editing and classification (as contributors do within Wikipedia) so that their energies are focused on what the site needs most from their skills.

My concern is that Quora already has such a strong yet fragmented community of 3% (sometimes, I’m in the 3%, and other times in the 20%) that it could be very difficult to collectively change so many users’ trained behaviors on the site. My hope, on the other hand, is sort of counterintuitive — that Quora actually become less participatory. If it can make that transition as it grows, I wager the service provided by the site and its content will actually increase in quality. Put another way, as Quora becomes less participatory among users, the signal could increase enough to drown out the noise.

Finally, it’s worth noting that this isn’t a problem that machines can address. It is a human issue, and it will take humans to address it.

About Semil Shah

Official contributor to @TechCrunch (since Jan 2011); from July 1, will begin EIR with @JavelinVP

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 49 other followers

%d bloggers like this: