White Papers

Copyright 2013 TechVentive, Inc., All Rights Reserved - Unauthorized reproduction, storage, transmission or quotation strictly prohibited. 8 And, the regression testing gets even more difficult when you consider that not all users of these systems are on the same release and patch levels of their technology stack components. Just because the application runs fine with version 11.4 of the RDMS (relational database management system), doesn't mean it works correctly with version 11.0 or 10.5. Moreover, a single patch applied to one of these versions could render the new application software inoperable. As a result, the testing team of an on-premises vendor must test potentially hundreds of combinations of environments to ensure that the software works and works as desired. All of this regression testing takes time and budget. The single technology stack of a multi-tenant SaaS solution is a clear advantage for the vendor and the customer. Vendors win as they can roll out more new functionality faster. Customers win as their solution can evolve fast and robustly. Both win as fewer 'surprises' await the customer because some anomalous bug crept into a customer's system. Customer support costs should be substantially higher for an on-premises vendor because one-off bugs are often difficult for the software vendor's support team to replicate and maddening for the software user to experience. Customer service in an on-premises software company must possess additional people with skills in the different hardware and systems software products that their customers use. They need people familiar with multiple versions of the application software, too. They may have to purchase/license additional hardware and software to replicate a customer's issue. All of this costs money and may re-direct monies intended for R&D into Cost of Revenue expenses for the vendor. The bottom line is that customer support is likely to be more expensive, time consuming and complicated for a vendor with a non-multi-tenant SaaS solution. The proof of all of this inefficiency is evident in a major application software vendor's suite. Its software can be deployed on-premises, on-premises in a private cloud, hosted on a third-party's data center or in the vendor's multi-tenant SaaS cloud. If a company chooses any but the last configuration, it will not receive upgrades and updates as quickly as the users on the multi-tenant version. Why? Vendors need additional time to regression test the solution for other computing platforms. We believe multi-tenant solutions: - are more efficient from an R&D spend perspective. - deliver functionality faster and with more frequency. - produce fewer bug 'surprises'. Additional Considerations Readers should note that absolute maxims are not possible in defining the advantages/disadvantages of different software deployment models. For example, a few multi-tenant software vendors have solutions that invoke one copy of the software yet keep each customer's data physically and logically

• Cover