In the continuing discussion of cloud computing, salesforce.com occupies a unique place. While most cloud vendors have kept to one component of cloud computing, Salesforce has inserted itself into all areas and the company is using its flagship CRM applications as its first case example.
Most people would agree that cloud computing is constructed from several parts though it is more than their sum. Software as a Service (SaaS) is the best known element but by itself it does not constitute cloud computing. In fact, it is debatable whether any of the elements alone would constitute the necessary conditions for cloud computing.
Other elements include platform as a service (PaaS) and infrastructure as a service (IaaS), which can be thought of as the supporting structures for software development and deployment and the physical datacenter respectively, which every software system must have regardless of how it is delivered to the customer.
There are many new vendors calling themselves providers of PaaS and IaaS that are laying claim to the moniker of cloud computing but those claims are not credible if you adhere to the trinity model above. In most cases these vendors offer a conventional hosting service which they also call PaaS when, ironically, it would, be more appropriate to call them IaaS. However, with the low visibility of infrastructure it is doubtful a pure infrastructure company would get funding. To that point, the best known infrastructure company, Amazon, is self funding.
Other platform companies offer some amount of software to integrate disparate applications but their architecture is decidedly single tenant. Compared to the hyper-efficient multi-tenant architecture of the Salesforce cloud, they use a lot of resource.
On the other hand, Salesforce has developed a shared infrastructure based on multi-tenancy, a development platform (also shared) that it has used to build and enhance its SaaS CRM solutions.
In a recent analyst briefing tour event held in Boston, George Hu, Executive Vice President, Marketing, for Salesforce presented the fully articulated Salesforce offering including its Sales Cloud and Service Cloud solutions. Hu also introduced FinancialForce.com, a new company that has developed a full accounting package running natively in the Force.com platform.
Hu said that Salesforce had taken a minority stake in FinanceForce.com. Significantly, Salesforce has always maintained that it had no interest in developing an ERP function and true to its word the company has not. However, with minority ownership in FinancialForce, the point is moot — Salesforce is now a full competitor in the market for front to back office computing solutions capable of facing off at different levels of the market with NetSuite, Oracle, SAP, Sage and Microsoft.
Also at the event, Salesforce used, for the first time in my recollection, the words “enterprise cloud computing”. The phrase hardly needs definition but it shows a robust intent to take a fully articulated alternative IT paradigm to the enterprise. In case you missed the point they have another phrase to drive it home — “the real-time cloud” that implicitly contrasts with background batch processing still inherent in some conventional applications.
Together, the two phrases announce that this company fully intends to replace conventional computing everywhere it can. They may not grab the whole market and it will take time to transition from legacy computing — just as it has taken a while to dethrone the mainframe — but the direction of the market seems obvious. Oracle and Microsoft at least will not stand by and watch and I expect some important announcements coming from Oracle Open World (ongoing as you read this) and Microsoft’s analyst meeting in November.
With footprints in front and back office computing, SaaS and cloud computing, in a short ten years, Salesforce has managed to reinvent enterprise computing and put it on a path of greater reliability and cost effectiveness than what was available when it started out. Skeptics will point to this or that function or feature that goes unimplemented still and while their objections might be valid for specific instances, they are beside the point. Features are not capabilities, they rely on capabilities and the product set appears to have the wherewithal to enable smart people to build whatever they need for enterprise computing.
Good as all that is, the expansion of this paradigm is also causing a contraction in the size of the available market. Going forward, enterprises might spend hundreds of thousands instead of millions on various IT projects and while that is still a lot of money, it is less than enough to support the constellation of software, development — and especially integration — vendors in the market today. This kind of relative contraction is inevitable in maturing markets and a good indicator of how far we’ve all come.