Disclaimer: All content in this article is the sole viewpoint of the author and is not endorsed by OpSource.

What are SaaS Key Drivers?
Oracle lays strong technology foundation for SaaS delivery

The delivery of Software as a Service offers several business and technical challenges for all parties involved - independent software vendors (ISVs), hosting services providers and customers alike. The technical challenges for SaaS providers include support for multi-tenancy, security, usability and service-level management.

The Oracle SaaS Platform -- consisting of database, middleware and systems management software -- provides a comprehensive platform that addresses these challenges and runs on grid. In simplest terms, SaaS refers to hosted on-demand delivery of applications, as opposed to on-premise. However, software as a service is not a singular model. It encompasses a spectrum of choices ranging from hosting traditional packaged applications at a remote site to multi-tenant shared applications. The Software as a Service delivery model is rapidly gaining adoption. In order to leverage this growth, ISVs need a platform that can meet the needs of next generation SaaS.

SaaS Drivers

The move to Software as a Service is driven by customer requirements and preferences. Customers are attracted to the promise of faster adoption, minimal impact and cost to internal IT and familiar web-based interfaces similar to consumer web applications resulting in faster adoption and minimal training costs. This demand for SaaS delivery is leading new ISVs to consider adoption of SaaS; and many existing ISVs are looking to offer their existing applications in the hosted model - either by hosting existing applications or by re-writing to a new architecture for SaaS.

ISV's can also leverage the following benefits of a SaaS delivery model.

• Access to new accounts/markets: With Software as a Service model, ISVs can target smaller customers who can well afford a subscription price and appreciate the minimal IT footprint
• Customer Insight: This is perhaps the most strategic competitive advantage from SaaS for ISVs. With SaaS, the ISV can monitor how the applications are used and develop deep knowledge of usage patterns, usability issues, etc. This insight can help in improved product roadmaps and also provide ideas for what new applications would be valuable to the customers.
• Value-added services: The shorter product lifecycle and ability to make frequent upgrades and fixes makes it much easier to bring new and incremental services (applications) to market and test them with real-life users.

Business Requirements

There are many challenges facing the SaaS provider as they look to build, deploy and manage these services. The following summarizes the key representative challenges and drivers for these ISVs:

• Service-level Management: Customers are increasingly asking for well-defined Service-level Agreements (SLAs) around performance, unplanned downtime, data backup and recovery, etc. The application design and the underlying platform must provide the necessary functionality and tools to meet these promises to the customer.
• Customizable Solutions: One of the most common concerns expressed UI can help ISVs sell to more customers and also partner with other vendors to offer vertical solutions.
• Low TCO to Scale: The ISVs need to be able to predict and manage the cost of operating SaaS solutions. Unlike traditional software vendors, SaaS providers must take into account the cost of delivering the service, up keep and ongoing maintenance and upgrades. These costs can rarely be passed on to the customer and directly impact the SaaS providers bottom-line.
• Time to Market: One key advantage SaaS ISVs have over traditional software vendors is the ability to expedite time to market. It is therefore important to have infrastructure and methodologies in place that allow the ISV to bring new functions and features to market in much less time than traditional software.
• Ability to Monetize: SaaS delivery model allows for incremental upsell at relatively low cost. ISVs want to be able to offer enhancements and complementary functionality or higher service-level guarantees without significant changes in the underlying application or infrastructure. It should be technically feasible to monitor and bill based on metrics such as usage.
• Security: The customers are often entrusting critical data to the SaaS provider and must be protected from theft and misuse. In addition, several regulations and laws apply to certain types of data (health records, financial records, government records) and in certain geographies. Breaches in security can not only hurt the end customer but could also damage the SaaS provider's reputation and expose him to the risk of litigation. Therefore, holistic security at all tiers is a prime business requirement.

Technical Requirements

The on-demand delivery adds several new business challenges and translates into technical requirements for the choice of the platform technologies.

The requirements that are new or require special emphasis for SaaS are as follows:

• Manageability: ISVs realize that service down time, poor performance and the inability of end users to complete critical business processes can be costly to their business. In the context of SaaS, where a service provider may have multiple instances of the application and underlying infrastructure running, it is imperative that the service level management tools can scale and support hundreds, even thousands of instances - in order to sustain and grow the business profitably.
• Security and Privacy: One of they key concerns of customers looking at adopting SaaS is security and identity management. The security infrastructure must ensure that customers' sensitive data and credentials are secure even in a multi-tenant environment. Effective monitoring and auditing tools should be available to track access and usage. Identity management also can allow the SaaS provider to interface with a customer's existing directory and provisioning infrastructure, and delegate everyday tasks like user management to the end user or the customer's IT department.
• Shared services and SOA: Standards based service oriented architecture enables the SaaS provider to build the application out as a set of services that can be re-used rather than a single monolithic application. The many benefits of SOA include faster time to market for new applications that can be composed rather than developed from scratch, greater interoperability with existing services within and outside the firewall, increased reuse, business processes that are easy to configure and evolve, and the ability to define and measure response-time and availability for specific services.
• Meta-data driven Customization: Customers often have business requirements and processes that may differ from those of other co-hosted customers, and may also not exactly fit the industry standards, if they exist. The on-demand applications must allow for easy customization and extension of user interfaces, data and business processes.
• Environment-centric design: With SaaS, the service provider can optimize the architecture taking into consideration its choice of operating system, databases and middleware. This reduces development costs, time to market and cost of maintenance over the life of the application, and can provide higher performance as well.
• Provisioning, patching and systems management: The SaaS provider is now responsible for bug fixes, upgrades and patching. With traditional delivery the customer could afford to schedule some downtime but this becomes increasingly harder or impossible to do if multiple customers are using the software often across different continents and time zones. The underlying infrastructure and application architecture should allow patches and upgrades with minimal disruption in service.
• Scalability, high availability and reliability: The underlying platform must provide for easy scale-out and scale-up of applications and shared services, so that more tenants and more per-tenant users can be added without degradation in service levels. The SaaS model means that even for applications that are not very mission-critical it is important to design for scalability, high availability and reliability as hundreds or thousands of customers could now be impacted by planned or unplanned downtime. Furthermore, a multi-tenant architecture and shared services add new complexity to problems like restoring data.
• Multi-tenancy: Many SaaS providers use single instance multi-tenant architecture in order to amortize costs across multiple tenants. However, this approach is suitable for many applications and customers, especially when targeting small and medium businesses, it is not the only architecture for SaaS. In fact, architectures for SaaS vary from hosted applications with multiple instances (ASP) to multi-tenant shared services models. The choice of the appropriate tenancy model will depend on customer requirements such as privacy, security, high availability, customizability, and comfort with the shared model.

Conclusion

SaaS is at an evolutionary stage currently and Next Generation SaaS delivery demands a robust platform. Visit the Oracle Technology Network for comprehensive details as to how the Oracle SaaS Platform meets these next generation SaaS challenges.