Phil Wainewright, who has been covering SaaS as long as anyone, brings up some great points about Salesforce.com’s continued hooplah around multi-tenancy in his latest blog post.
The first major point is that multi-tenancy is more of a marketing gimmick than it is a “secret sauce” for Salesforce.com. In fact, as the post infers, there are parts of multi-tenancy that may not benefit users at all. And since Salesforce.com built its infrastructure a decade ago, prior to the explosions in virtualization and open source technologies, they were forced to create theirs in such a way.
But in my opinion, one of Salesforce.com’s greatest skills is spinning their limitations into a marketing differentiator.
Now, the talk at Salesforce.com is all about the cloud. Fine. But where does multi-tenancy fit into the emerging view of cloud computing as one of openness and interoperability? Or, more directly, how important is multi-tenancy when you are free from the binding restraints of high priced, server-based Oracle pricing?
I would argue (and have in the past at length) that multi-tenancy is a non-factor in the truly open cloud. You could go so far as to argue that Salesforce.com’s multi-tenancy model is a lock-in strategy that is anathema to what the cloud should be representing.
In a truly open cloud – you should be able to deploy the SAME software configuration in a vendor-operated cloud as you could your own private cloud (Wainewright hints at this, noting that larger enterprises are seeing smaller scale clouds as the way to go inside their IT setups), or even in a public cloud (say, hosting your CRM system inside a Rackspace or Amazon-powered environment.)
But Salesforce.com cannot allow this. Not I do not say “will not” allow this – as I don’t think the Salesforce.com lock-in strategy is 100% intentional. Based on the limitations they had in 1999 – they built on a multi-tenant architecture, one that now supports a platform and web sites in addition to the core SFA apps they first devised. This is a highly functional environment – sure – but offers little in my opinion in terms of long term risk mitigation or flexibility for deeply customized environments for the very enterprise consumers they are trying to target these days.
I will certainly be typing up more on this concept, as my ideas around the open cloud and how the present SaaS CRM providers fit in develops in my mind…stay tuned.