by Rob Cheng (@robcheng)
CloudSpokes hit a big milestone this week, passing 50,000 members after just 19 months — hitting that mark 5x faster than Dreamforce! Of course, CloudSpokes could never have grown as fast as it has without Salesforce and other public cloud pioneers laying the groundwork for us. So we wanted to take this opportunity to reflect on the surprisingly deep relationship between the coupled evolution of cloud computing and crowdsourcing.
The Cloud Brought Crowdsourcing to App Development
Prior to the information age, large scale crowdsourcing was mostly applied to solving physics or math problems because those were the only problem domains where you could expect consistent and reproducible results regardless of geography. Once the Internet made it possible to easily share data globally, innovative companies took advantage by crowdsourcing previously intractable problems like discovering new deposits of gold, understanding the structure of a key AIDS enzyme, or improving prediction algorithms.
But when it comes to application development, access to data isn’t sufficient. You need access to code, configuration, assets, and more importantly, a place to deploy and test your code that mirrors the production environment. This last requirement has traditionally been so difficult even for internal development teams that entire industries have sprung up to help solve the impedance mismatch between dev, test, staging, and production environments. That’s where the cloud comes in — not only have SaaS offerings created standard environments for developers to target, but cloud platforms and infrastructure have made it easier to clone production environments with perfect fidelity, even while limiting access to sensitive data. For example, while Salesforce.com makes it easy to create complete sandboxes from production orgs, CloudSpokes clients find it’s often faster and safer to simply duplicate the subset of object metadata needed and have the competitors deploy that metadata into their developer orgs. Similarly, standardized EC2 AMIs and Heroku Git repositories help ensure consistent and reproducible results with crowdsourced code, enabling the power of crowdsourcing to be applied to a whole new set of problem domains, from automating back-end business processes to developing innovative mobile interfaces.
Crowdsourcing is the Extension of Cloud Principles to Development
While it’s tempting to view the cloud as simply an enabler of crowdsourcing, the relationship is actually much deeper: crowdsourcing is the natural result of applying cloud architectural principles to the development process. One of the core characteristics of cloud is elasticity: instead of paying high upfront (capital) costs for fixed capacity, you can pay variable (operating) costs based on capacity that scales according to demand. Similarly, crowdsourcing provides a model where you can replace high upfront labor costs with variable costs and more importantly, pay only for results instead of paying for inputs.
But just as with cloud computing, cost is only part of the equation. Public cloud vendors are able to rapidly innovate because customers share a common multi-tenant platform, so they don’t have to devote time and resources to support hundreds of possible on-premise configurations. With crowdsourcing, customers benefit from multi-tenancy in a much more direct ways. Drawing from a multi-tenant pool of shared development talent means you can staff every project with technology specialists instead of the IT generalists that are often necessary due to limited headcount. Because crowdsourced challenges attract multiple participants and submissions, you can reduce the risk of single sources of failure in your development process. And crowdsourcing allows you to combine the best aspects of multiple independent approaches within a competition model that drives innovation for your applications.
Crowdsourcing Will Become Just Another Benefit of the Cloud
If you scratch beneath the surface of almost any Internet success story, you’ll find that crowdsourcing has played an important role: Google’s pagerank analyzed user-created hyperlinks to improve search, Amazon’s recommendation engine dramatically increased online purchases, Yelp crowdsourced reviews, YouTube and Wikipedia crowdsourced their core content, and even the ubiquitous social media hashtag was originally invented and popularized by the community. Crowdsourced content so fundamentally reflects and reinforces the architecture and principles of the Internet, that today it’s hard to even find a site that doesn’t utilize user posts, pins, uploads, comments, votes, shares, or tags in some way.