Today’s consumers are familiar with Interactive Voice Response (IVR) systems—commonly referred to as “phone trees''—and their benefits and downfalls. A well-constructed IVR feels seamless and gets a consumer right to where they want to be… but a bad one can lead to non stop tapping of the 0 or 1 key, all the way to shouting down the phone when the automated robot just doesn’t get it. Read 6 ways building a modern IVR propels your business to new heights.
Those in the contact center or customer care community know that the technology has actually been around since the early 70s. When IVRs were first released as a feature, they promised great things, such as massive call deflection and call steering rates.
However, they were often really difficult to both configure and maintain, with some companies deploying entire teams of people to make regular changes to meet continual shifts in demand.
Frustration among consumers was obvious, with many bypassing IVR trees altogether by pressing zero or just pressing whichever buttons seemed like the shortest path to speak to a human. However, they then needed to be transferred between departments because they had ended up in the wrong place. Operationally, this introduced other challenges. Agents now had to deal with frustrated and disgruntled customers, who’d been trapped in the labyrinth for what felt like forever, only then to be asked for their details again.
Speech recognition in the IVR world then hailed an end to the “press 1” requirement. Consumers could simply say what their reason for calling was, or answer a yes/no question. The infamous “I SAID YES” era commenced, with technology and speech recognition platforms still coming of age, trying to understand multiple languages, accents, and idiosyncrasies.
While these systems were often more effective, the upgrade in the quality of the experience required a lot of time, effort, and investment to make subtle changes—not to mention the cost of hardware and port licenses that had to be maintained. Many companies began to ask themselves: how effective are these systems given the time and resource required?