Post by Sean O’Connell, Manager of Product Marketing for Cisco’s Customer Contact Business UnitRemember the great”Seinfeld” episode, in which Jerry’s neighbor Kramer attempts to imitate an automated voice self-service system for”Movie Phone,” and then after not being able to identify the dual-tone multifrequency selections (aka Touch-Tone),”upgrades” himself to support automated speech recognition? (The punch line being”why don’t you just tell me the movie you selected?”). Sadly, the reality is that Kramer’s hilarious imitation does a better job of than some (most?) of the speech recognition systems in operation today.In the area of customer service, which offers companies the potential for competitive (dis)advantage, many today offer their customers a variety of self-service options. Typically, one of those self-service options is the interactive voice response (IVR), or voice self-service, system. If you’re like me, perhaps you prefer to use self-service systems to avoid the potential hazards of live customer service, such as long wait times or being transferred from agent to agent.Voice self-service systems have been around for many years. Most systems today are based on traditional dual-tone multifrequency (DTMF) technology -you know,”press 1 for sales, press 2 for service”, etc. In recent years, more and more companies are beginning to roll out automated speech recognition capabilities as a feature of their IVR/voice self-service systems. These automated speech recognition systems bring the promise of a more intuitive, user-friendly, and flexible user interaction. But all too often I’ve found that the speech recognition systems do not meet the challenges of recognizing human speech (particularly in environments with high levels of background noise, such as airports). Which begs the question, what are the considerations for businesses as they begin to deploy automated speech recognition systems?In a recent Webcast by Dimension Data and Cisco, the question of speech recognition’s maturity was discussed–citing a recent study which contrasted views from both vendors and consumers. Not surprisingly, vendor views on speech recognition are more bullish than those of consumers. But what is surprising is how dramatically different the views can be. Speech recognition, for all its hype and promise, does not work in a vacuum in the IVR/voice self-service environment. In fact, one of the conclusions of the study is that technology isn’t even the most important consideration; rather it’s understanding consumer expectations and motivations that will drive voice self-service usage.It’s important to remember that speech recognition systems do not work alone, and must be thought of as one piece of a larger voice self-service strategy -or better yet, the even larger customer service strategy (consisting of numerous multi-channel self- and assisted-service customer touchpoints). When you begin to think about the bigger picture for voice self-service, that’s when Kramer’s automated”Movie Phone” won’t outperform today’s systems.