StumbleUpon is one of the more interesting services in the social media realm. While it has its social elements for sure, it is much more about the content.
As a follower of the search industry, it’s particularly interesting to observe how content quality is handled with StumbleUpon – an area where Google has struggled and other new search engines have popped up with different approaches to solving. For example, Blekko takes a human curation approach to cleaning up search quality.
While we don’t often think of StumbleUpon in terms of search, it is similar in that it’s a tool you use to find content. The difference is that in many cases with search engines, you’re looking for very specific things. With a content discovery service like StumbleUpon (and even Google’s new Sparks feature in Google+) you’re looking for things related to certain topics, but not for anything in particular.
With a new feature that StumbleUpon is testing – the explore box – this “looking for information” aspect of StumbleUpon is emphasized even more. It allows you to drill down to very specific keywords – even more narrow than the topics StumbleUpon has become known for. And with the nature of StumbleUpon, users drive the highest quality content or at least the most appealing to the highest amount of visibility.
The company discussed how users can improve the quality of their own Stumbling experience in a blog post this week.
“As a human-curated discovery engine, StumbleUpon relies on our users to create a collection of the best content and information on the web,” said StumbleUpon Content Operations Program Manager Bruna De Goes. “We auto-magically identify most low-quality content, but sometimes nothing can beat the human brain’s ability to tell good content from bad.”
StumbleUpon users have four ways they can give content a “thumbs down”. These include: not for me, report spam, duplicate content, and block this domain.
Imagine if Google had these options on each search result. They did recently start letting users block domains from their own search results, and that data, Google said does influence overall search results in “high confidence” situations. But imagine the anarchy that might ensue if competitors and disgruntled customers were able to just hit a spam button on each search result. Blekko does that. What if Google let you hit a “duplicate content” button? How often would that be misinterpreted and send the wrong message to Google?
If you look at how StumbleUpon defines duplicate content, it explains it as: “you’ve stumbled this before. This option is useful especially when viral copyrighted content (photos or videos) are embedded in web pages for the purpose of getting a lot of traffic.”
What’s to stop a user from reporting the original as duplicate content, however, if they saw another site using it first? De Goes explains duplicate content slightly differently, saying: “You’ve already seen this content, and it looks like it was pulled from another source without referencing the original.”
It would be nice to know more about how StumbleUpon enforces its reports. It has a page about disputes here, but it doesn’t give a lot of insight into such enforcement.
You may recall how some application developers were affected by blocks on Facebook, despite having solid reputations among users. Facebook’s algorithm simply banned the apps altogether, affecting entire businesses. It would be a shame if sites were being majorly impacted by wrongful spam/duplicate content reporting. I’m not saying this is happening with StumbleUpon, but they could probably do more to make the process clearer.
Here’s what StumbleUpon considers to be poor quality content, according to its guidelines:
Duplicate content – The content of this website exists somewhere else on the internet and is not original. (Examples: A viral YouTube video embedded on a page that has already been submitted with its original YouTube URL. A photo slideshow that “steals” popular photos.)
Low quality pop ups – Pop-unders, shaky-flashy etc.
Ad heavy page – The website contains nothing but ads above the fold.
Pages that list search engine results
Sites that require a login to access information – Abrupt call to sign up that is generally intended to get your email address.
Low quality page in general – This website has low quality web design. It does not look attractive.
Photo URLs – The only content on the page is the photo, with no way to navigate back to the source.
Here, StumbleUpon tells users to read the Wikipedia article on spam to identify it.
StumbleUpon has over 15 million members according to StumbleUpon social media manager Katie Gray, who said as much on Quora in late June. That’s apparently less than Google+ already has (which bodes well for Google’ sparks feature, which in my opinion is a direct competitor to StumbleUpon), but nothing to shake a stick at.