This article originally appeared in the October 2012 issue of CUSTOMER
Optimum (News - Alert) contact center performance is essential for satisfying the needs of today's always-on customers, but it can be difficult to maintain in the face of constant change. Unfortunately, these changes – whether they include the introduction of a new service or an upgrade to existing technologies – must be made in highly complex, multi-vendor environments where network issues are extremely difficult to detect and diagnose.
Preempting issues with pre-deployment testing is not a luxury. The chance that a customer might have a negative experience is too great a risk to take. In fact, Gartner (News - Alert) estimates that up to 70 percent of enterprises that do not perform adequate network assessments will experience unplanned additional costs and time delays. The following tests will help companies ensure optimum contact center performance and a great customer experience:
Before any change is made to a contact center, it is extremely important to document performance up front. To fully characterize performance, baseline testing should emulate real-world customer behaviors and traffic volumes, whether those behaviors and traffic relate to e-mail, chat, phone, video or data communications.
Voice Quality Testing
Unlike traditional phone systems, today’s IP-based solutions carry voice traffic along the same network that supports video and data solutions. While this generates tremendous cost savings, it leaves contact center solutions vulnerable to voice quality problems, especially under heavy load conditions or when traffic spikes in other areas impact voice transmissions.
To accurately assess voice quality, tests must be conducted bi-directionally, or from the customer to the agent and back again. At the same time, companies should also assess the quality of IVR prompts and responses.
Poor voice quality issues cost U.S. businesses more than $34 billion per year, so it is extremely important that companies preempt these problems from the start.
Regression testing ensures that existing services do not regress or degrade as changes are introduced.
To successfully implement a regression testing program, companies need to accurately map out existing services. This step can be extremely challenging for organizations that are constantly updating their contact center systems – and their underlying technology – in response to events such as product launches, cost-cutting programs or mergers and acquisitions. In other cases, the loss of legacy expertise has created gaps in understanding operations as a whole.
Many companies must first conduct an end-to-end audit of existing services before they can begin a major project such as a migration to SIP or a CTI (News - Alert) implementation. In some cases, the very tools that companies will use to perform regression testing can be used to quickly unravel complex routing schemes or branched IVR functionality. This work can be leveraged down the road to create automated regression test scripts.
Toll-Free Number Sweeper Testing
As contact centers evolve, so does their utilization of 800 numbers. Continuing to maintain toll-free numbers linked to outdated campaigns, services, products or departments can be extremely expensive. This test quickly determines which numbers are actively being used and which are out of service.
The last mile of any project is the most difficult, and the only way to determine if a new service or upgrade is ready to roll out to customers is to test complete contact center performance from end to end under predicted load conditions.
Best practice methods recommend a methodical, end-to-end approach to pre-deployment testing. That means first validating real-time voice and video services. Once these systems are stable, companies can add data traffic and test service quality under these new conditions. Many times the additional data traffic generated by chat and CTI applications will affect voice and video quality. Once the quality meets expectations, companies can begin testing the functionality and performance of each application (IVR, routing, chat, CTI, agent desktop) in a step-wise fashion until the entire customer experience is validated at expected usage levels.
Developing a test plan that accurately reflects expected user behaviors and traffic patterns allows companies to uncover issues and interoperability problems before they can impact live systems.
Companies that follow best practice project management and pre-deployment test methodologies are better able to deploy new customer services on time and on budget.
Typically, these companies allocate approximately 5 to 12 percent of their project costs for pre-deployment testing. However, the cost savings of detecting issues in the lab compared to the production environment is 100 times lower. More importantly, it ensures that customers are delighted by the effort, not frustrated by dropped calls, garbled speech or chat delays. It’s clear that the return on investment of implementing such pre-deployment tests sufficiently justifies the cost.
Edited by Brooke Neuman