One of the real-time lessons we learned at Judy’s Book is that while UGC is great, without vigilant moderation of that content, you’re going to get a lot of noise in the system, which is not great. When we were aggressively seeking reviews, we relied exclusively on the community to flag posts and then reviewed those posts for compliance with our terms of use. This approach is ok, but doesn’t go far enough. We should have looked at every review and only approved content that was of value for review readers.

When we transitioned to deals, we adopted the same - anything goes, we’ll deal with spam - approach but quickly found that deal posters were flooding the system with deals that from our perspective were questionable. The problem was compounded by the fact that with deals, there is a financial incentive to posters - the more of their links they publish, the more likely they are to get paid. Initially, we tried to combat the problem with algorithms that gave higher preference to known posters, and more aggressive deal removal tactics. This helped, but we constantly felt like we were behind the curve.

The solution was a simple re-framing of the problem. Rather than let anything onto the site and then remove bad deals, we decided to only allow good deals onto the site. The way we did this was by creating a moderation queue. Anybody could post anything, but before anything came onto the site, it would have to be approved by a member of the JB staff. This simple decision completely eliminated our spam problem. A simple, streamlined interface made it quick and efficient to review deals and we saved ourselves a great deal of frustration dealing with bad deals on the site’s primary pages.

If you’re thinking about incorporating UGC into your site, I’d strongly urge you to consider a moderation queue approach. While you do sacrifice content velocity and incur a certain amount of overhead due to the review process, you gain a great deal in terms of content quality and less policing of the site’s content pages. While I can envision a few scenarios where quantity should come before quality, they are rare. When in doubt, err on the side of what will generate the higher quality user experience.