While details are scarce, you can expect the fog of marketing confusion that presently surrounds Microsoft's current client/server CRM and enterprise resource planning (ERP) to continue, as Microsoft tries to battle a broad number of competitors in different markets. Microsoft keeps wobbling between saying its business applications are for the mid-market and for the enterprise.
Already, Microsoft has said Dynamics CRM will target "small businesses... traditionally underserved by the lack of flexible and cost-effective CRM solutions." However, Microsoft was also quoted saying Dynamics CRM will appeal to government agencies "consternated" by hosted services. Microsoft has promised a "private database" for every customer "not a pooled data model."
While details are apparently still being worked out, you should expect Microsoft to make a strong integration play. Dynamics CRM will tap the same meta-driven configuration tools used by all versions of Microsoft CRM, while integration with Microsoft Office and Outlook will make life easier on partners and customers. You should also expect Project Snap, integration between Dynamics and Office/Outlook, to play a role in the architecture, and for everything to run on the evergreen SQL Server.
A major weakness of hosted services has been lack of integration with software from Siebel, JD Edwards, PeopleSoft and Oracle. SAP has been at pains to point out its hosted service shares the same architecture as its traditional client/server product.
Salesforce.com has tried to resolve integration in two ways. First, by exposing its metadata and information using web services and Ajax, and - secondly - by encouraging partners to deliver tight fitting applications. Examples of the later, include Informatica and Scribe Software providing software to migrate and synchronize data between Siebel and Salesforce.come and Dynamics and Salesforce.com respectively. Additionally, Salesforce.com is now reaching outside CRM with an OEM version of its service for developers to build new applications capable of integrating customers' data minus the familiar Salesforce.com CRM tabbed interface.
Microsoft has had more than six years to win over the types of customers now flocking to Salesforce.com, NetSuite and Sugar CRM. It has chosen to spend the better part of the last half decade trying to merge its various CRM and ERP offerings on a single code base under Project Green. That project has suffered numerous set backs.
Microsoft's architecture also promises to throw up the scalability issue. Microsoft's decision to "give everyone a database" will mean its service will face an infrastructure hurdle, as Microsoft will essentially try to apply a client/server architecture to an online service. Fans of the Microsoft online service model need only look at the performance and reliability issues that sill hit Hotmail.
One of Salesforce.com's successes has been its use of a large, mission-critical database - Oracle - on scalable hardware. Microsoft's approach promises to be costly, using a lot of Intel-based systems, meaning it is unclear how Microsoft can pass on costs to the price-sensitive customers snapping up Salesforce.com, starting at $995 for five users per year or $65 per user each month. That will be particularly imporant for Microsoft, after Ballmer last year promsied to give Salesforce.com a "run for its money." ®