According to a report by Forrester, Software-as-a-Service (SaaS) will likely surpass an adoption rate of 50% within the next year, with adopters using an average of four SaaS solutions. Given the persistent growth of companies such as Salesforce.com, NetSuite, Google, and Amazon — the latter of which will take in $5 billion from its cloud computing business this year alone — the Forrester study is hardly surprising.

In short, the cloud model shows no signs of slowing, providing opportunity for all types of cloud-based SaaS vendors to take advantage of a lucrative market. Companies are increasingly reliant on SaaS for mission-critical applications such as email and other collaboration tools, customer relationship management (CRM), enterprise resource planning (ERP), and finance.

However, there are challenges: with so many vendors in the market place, consumers can easily switch services if their demands for 100% availability and uptime are not being met. For example, a recent Aberdeen Group survey said that one quarter of users will abandon a Web application after just three seconds of delay.

To avoid customer attrition, SaaS vendors need to focus on the optimization and scalability of their operations in order to keep current customers happy and allow for future company growth. Many vendors attempt to address application performance issues through investments in traditional application performance management (APM) solutions and more hardware.

However, these investments alone seldom satisfy service delivery requirements long term. This is because in most cases, poor database performance is at the heart of poor application performance, and the database tier is frequently overlooked when assessing performance. Only by accurately diagnosing and resolving database performance issues can SaaS vendors more easily and economically optimize application response times.

Ensuring application performance: the wrong (costly) way

To adjust to rapid growth, many SaaS vendors address application performance issues by adding more server resources rather than re-architecting their operations for success. After all, horizontal and vertical server scaling is easy with commodity-priced Infrastructure-as-a-Service (IaaS) and automated virtualization solutions. However, these quick fixes fail to fully address application performance issues long-term.

Application performance issues typically originate at the database level because SaaS solutions continuously collect, analyze, and/or deliver vast amounts of data that must be processed instantaneously. Waiting for this data to be processed is a significant contributor to the time customers spend waiting on latent applications.

While modern databases such as SQL Server, Oracle, Sybase, and DB2 are built to handle large volumes of data, it is critical to properly configure them to produce the response rates their customers require.

However, databases are harder to scale than other tiers of the SaaS delivery system — resulting in the failure to properly configure this critical component of their operations. Instead of getting to the root cause of an issue, vendors will often fine-tune applications using data collected from APM tools alone without ever resolving the root cause of the problem.

To illustrate the impact, consider the fact that many leading SaaS vendors provide access to community-based portals that provide customers with visibility into performance and security issues. A quick survey of a few of these portals will show that issues are often occurring within the database.

Ensuring application performance: the right (cost-effective) way

To significantly improve cloud-based application performance, SaaS vendors should correctly tune SQL queries and properly configure database systems. Not only does this ensure better application performance, but vendors will see an increased return on investment of initial hardware purchases.

For example, for a SaaS company with $20,000 in current database server and storage infrastructure costs per month that anticipates growing to $75,000 per month in a year’s time, it’s not a stretch to estimate immediate savings of $10,000 per month, or a total savings of $285,000 over 12 months. Not only does this produce a 50% reduction in server cost, but also an improvement in application performance.

However, assessing database issues is challenging because of increasing complexity among database management systems, server resource provisioning, and SQL queries. Combined with constantly changing workloads and more dynamic server environments due to virtualization, it’s clear that ad-hoc or lightweight database performance solutions fail to deliver actionable information.

By implementing a legitimate database performance management tool to resolve database performance issues, organizations can accurately pinpoint a database’s contribution to slow application response time, in both development/testing and production situations.

In conclusion, the rapid adoption of cloud-based SaaS solutions has produced a tidal wave of companies looking to capitalize on new market opportunities. However, SaaS vendors will have to truly overcome obstacles around availability and scalability in order to ensure long-term success.

SaaS vendors cannot afford to let performance issues linger, nor can they risk making costly investments in added infrastructure that fails to improve application response times and reduce latency issues. They need to seriously consider adopting a unique approach to performance optimization via database performance analysis to more easily and economically optimize their application response times and maintain a strong customer base.