There was lots of talk about platform at SaaScon 2007, starting with SalesForce’s Jim Steele revealing Apex (SalesForce’s new development platform). As part of the presentation, a SalesForce developer presented an application he’d created using SalesForce’s Apex. The application allowed for tracking time-off requests, and included a custom time-off record, workflow rules for escalation, actual code to calculate holidays and weekends, and a calendar to show the time people were taking off.
The demo was very impressive, and definitely seemed to validate the Saugatuck keynoter’s suggestion that we will see a handful of what he called SIP (SaaS Integration Platform) stacks pop up from big players. A SIP stack would provide a SaaS platform where developers could write mini-applications that would run within the vendor’s SaaS infrastructure and could extend existing functionality. Developers would be able to develop sweet little bolt-ons and not have to worry about hosting their code themselves (or creating the core functionality).
It’s interesting to contrast this idea with something like SugarCRM’s development platform. SugarCRM can also be customized, but you have to download the code and host it yourself (I’m not sure how many vendors want to run your infinite loops in their multi-tenant environment).
Either way though, it seems that these platform providers will run into a conflict of interest by providing both a platform AND functionality (think Vista/3rd party spyware). It’s already playing out. SalesForce recently purchased a Web 2.0 document management provider to be the driver behind SalesForce Content. So where does this put the 20 existing providers of document management solutions within SalesForce’s AppExchange? I spoke to one of these vendors at the event that was very frustrated with SalesForce’s new acquisition. They had spent a ton of time and money to establish themselves as a premier provider of the service within SalesForce, and now they face the prospect of being obselete within SalesForce.
It will be harder to convince development companies to jump onto a platform when there’s no security that the platform provider won’t build the exact same functionality into the core of their application later. But, on the other hand - a platform provider shouldn’t be held hostage to 3rd party apps in their developer network, as it puts them at a distinct competitive disadvantage. It just seems to me that companies like SalesForce will have an extremely difficult time balancing the relationships with 3rd party developers with the need to put the best product on the market.
On the flip-side of the equation, a platform provider that provides very little or no end-user functionality, but rather focuses on the infrastructure and platform tools will also have a hard time attracting developers because the developers don’t want to have to start from the ground up. No way the SalesForce developer is going to create a fully-integrated time tracking system in a weekend without the huge head start the core SalesForce architecture provided (Users, Permissions, Layout, etc)
So, where’s the happy medium?
