ClusterShot: How a Startup’s Pet Project Became Profitable

clustershothomepageWhen Canada’s Silverorange started developing ClusterShot, a web site through which people can upload and sell their photos online, in May of 2008, it was nothing more than a side project members of the 14-person web development firm worked on in the evenings and on weekends. Today the service consumes up to 25 percent of Silverorange’s development time and is profitable.

ClusterShot earns revenue from selling annual Pro subscriptions, which cost $20 and provide users with their own personal photo store. So how did ClusterShot achieve profitability just eight months after it launched last November? I spoke with Silverorange CEO Dan James to learn about the company’s recipe for success.

If there are enough squeaky wheels, maybe you can build a business to provide the grease: The idea for ClusterShot was born when Silverorange employees got fed up with unsuccessfully trying to sell their photos on sites like iStock and Shutterstock. Such sites have strict photo requirements, James explained, which makes it hard for non-professional photographers to get their images added to their inventory. So Silverorange decided to build a site on which anyone could upload and sell any photo they wish.

When it comes to creating a web marketplace, make failure an option: Silverorange didn’t stop focusing on the core of its business — web development — to create ClusterShot, but built it in its spare time and gave it only as much attention as was absolutely necessary. ClusterShot’s success, after all, depends on whether people are willing to pay for photos on the site, so rather than risking everything on a service it didn’t really need to build, the company took a methodical approach.

No one gets it right the first time around, and true entrepreneurs are prepared to readjust: Silverorange made two mistakes with ClusterShot. First, it underestimated the number of photos users were going to import, and was forced to completely rebuild the site’s image processing system within the first month once it was clear that the average new user’s upload time was a full two days. (ClusterShot can now import thousands of images per day; some 350,000 photos currently reside there.) The second mistake was design-related — Silverorange originally used a light background and font color on the site, which didn’t make the photos pop. So it changed the background to a deeper color.

The inmates can’t run the asylum, so pick a leader: ClusterShot was originally a group-directed project, but it lacked focus. So James decided to apply the model that Silverorange employs when it’s developing a site for a client, whereby one staff member is assigned to represent the client and take the lead during the design and development process.

You got this far listening to squeaky wheels, but don’t let individual complaints drive your business: At first, Silverorange would focus on making changes to the site based on individual user feedback, which resulted in what James described as “some hastily made features that really only satisfied one user.” It now pools and ranks user requests and makes changes to ClusterShot based on such aggregated feedback instead.

Note: Silverorange designed the current version of the GigaOM site.