StumbleUpon: The Google For People Who Aren’t Looking For Anything! – Harnessing Collective Intelligence

You sit down at your computer with the entirety of human knowledge at your fingertips; from how to build a rocket, to what Geoff had for lunch, but cannot think of a single thing to do.

With over 2.4 BILLION people in the world using the internet, spending a total of 35 BILLION hours online per month, this has become the modern day curse for the regular internet user. Luckily for you there is StumbleUpon, which aims to finally give you the personalized web experience you need, to fill in all those unwanted real-life hours that you used to spend outside.

What is StumbleUpon?


StumbleUpon is a type of web search engine that finds relevant webpages for users by studying their interests and reception to other webpages. It does this by allowing users to initially choose categories regarding their interests and hobbies, then providing users with webpages that fit into those categories. As a user ‘Stumbles’ through pages, they have the option of giving a Thumb’s Up or Thumb’s Down to each page. This feature allows StumbleUpon to personalize websites given to users even further by making connections with their preferences to the preferences of similar users and using this to provide more relevant webpages. Thus creating a unique and personalized journey through an assortment of webpages catering to each individual.

Through this post I would like to discuss how StumbleUpon successfully Harnessed the Collective Intelligence of its users to become the unique powerhouse of personalized user experiences that has gained it extensive notoriety and over 10 Million Subscribers in just over 10 years. I will also discuss the possible downfalls that could hinder the growth of this company.

So What is Collective Intelligence and Why should it be Harnessed?

Collective intelligence is a shared or group intelligence that emerges from the collaboration and competition of many individuals. – Wikipedia

This is an extremely useful process for aiding rapid business growth through user participation and collaboration, by providing an architecture that allows users to edit, upload or create content. When a Business can harness the collective intelligence of a large group of individuals, then the data that is created quickly becomes more valuable then the mechanism used to create it. This is evident in examples such as Wikipedia, Amazon and Facebook, where the website itself holds little value, but rather the user’s data is what makes these businesses unique. After all without Collective Intelligence these websites are just empty white pages.

Harnessing Collective Intelligence - The Shortcut to rapid business growth.

Harnessing Collective Intelligence – The Shortcut to rapid business growth.

StumbleUpon has also used the Collective Intelligence of its users as the basis of the success of its business model through 5 key aspects of best practice which will be explored in-depth below.

1. Rewarding the User First:

When first viewing StumbleUpon, its simplistic design (a stumble button and a thumbs up/down button) along with quick and easy signup, provides the perfect template to minimize the barrier of adoption and allow users to begin ‘stumbling’ within minutes, providing a rewarding first experience for users. By using the intuitive thumbs up/down process for evaluating content it has also provided a universal and simple user-interface mechanism to gain information from all its users to provide better matches. This ease for users to accomplish their goals allows a rewarding experience for the user where their needs are met first before any secondary benefits for the company.

2. Setting Network Effects by Default:

Network effects occur when a service becomes more valuable as the amount of people using it increases. This is evident in StumbleUpon as an increase in users means both an improvement in the connections made between interests and websites, as well as more webpages being added to the database. StumbleUpon gains network effects via voluntary addition of webpages by users and indirectly as a side effect of users liking and disliking webpages. The addition of webpages by users although voluntary still provides benefits to the users self interests aswell as webmasters will want to add their own webpages and users may want to share their favourite webpages.

3. Involve Users Implicitly and Explicitly:

StumbleUpon’s seemingly primary goal is to create a personalized experience for each user via their interests and opinions of relevant content. User’s are able to improve the personalization of the websites they view by choosing the categories they are interested in and giving each website they view a thumb’s up or thumb’s down depending on their thoughts on the content provided. This allows the application to implicitly gain information through user’s preferences of pages and thus improving the connections between relevant interests. However users can also explicitly add content through adding webpages to the StumbleUpon database and although only a small percent of people will do this, it still amounts to a large amount of content daily.

4. Trust Your Users:

For Harnessing the Collective Intelligence of your users it is important to Trust Your Users, by providing an open context to allow participation through adding their own context and controlling their experience. This is no different in StumbleUpon as users are able to add their own webpages and evaluate websites negatively or positively. However it is important to note that a small element of control must remain, especially in this context as users are able to add any webpages and categorize them how they wish for others. This can potentially cause problems as users may upload unwanted material (such as adult content) in the wrong category, however this is corrected by ‘Natural Selection’ where wrongly marked websites are ‘thumbs downed’ as spam and correctly marked websites will thrive on top of the lists.

5. Facilitate Emergence:

When creating an application it is often difficult to predict the best design for the way users will utilize the software, therefore it is vitally important to provide a less upfront structure that will allow users to determine the direction your software takes. StumbleUpon began its creation with both the website stumbling design as well as a profile and blogging system (Facebook-style), however they were unable to support the quick user growth and as such had to choose what to cut back. So on October 24th they removed the blogging aspect and concentrated on the stumbling aspect as it was the core use of the application by users.

These aspects of harnessing collective intelligence for StumbleUpon proved to be highly beneficial for the company, however there are also possible downfalls and conflicts of interest that can occur which will be discussed below.

Walled gardens:

Walled gardens occur when a program has the ability to control the content that the users are able to use within the system and not allow users to view outside content. This is a possibility within StumbleUpon as the ‘mechanism’ for finding relevant webpages could turn into a ‘mechanism’ for showing the webpage which has paid the most, where only websites who pay could be viewed on StumbleUpon.

Privacy and Liability for Individuals and Providers:

Privacy and liability concerns could occur through leak of users private data and sharing of inappropriate webpages to users. StumbleUpon stores all websites that a user ‘Thumbs Up’ and displays this on their profile page, furthermore you can easily search through users you have on Facebook to find their profiles, this allows the possibility for embarrassing information to accidentally be leaked unsuspectingly.

Quality, not just Quantity, matters:

This can be the greatest downfall of Harnessing Collective Intelligence as not all users inputs are equal, it is important to have a way to distinguish between beneficial data and spam from users. However StumbleUpon allows users to mark webpages as spam or not relevant to their interests, so this allows this non-beneficial information to quickly be deleted or slowly become less frequent, with the higher quality data being used more frequently.

It is clear through this blog’s research that although there are downfalls to Harnessing Collective Intelligence, the benefits greatly outweigh these. It is clear through StumbleUpon’s use of Collective Intelligence that the information gained from a large group of people can be of great value to providing a unique user experience and can provide rapid user-based growth.

What are your opinions on the benefits and negatives of Collective Intelligence?

Join me next week where I discuss the importance of data as “The Next Intel Inside”.

Arnold Bach


Week 1

Welcome to my Web 2.0 Evaluation and Review Blog,

I am currently a student at QUT, studying Information Technology and Applied Science, and as part of the IT section of this degree have been required to construct a weekly blog on current Web 2.0 technologies.

So over the next 8 weeks I will find, review and evaluate different types of Web 2.0 applications and compare them to similar concepts on the market.

By reading this blog you will have the opportunity to trial the weekly application, as well as obtain an overview of the relevance, implications, issues and future directions we believe the application entails.

Thank you for taking the time to view this blog.