It's that time of the year again when the colorful characters of speech recognition technology come out of the woodworks and converge upon the Big Apple in herds to attend SpeechTek—the industry's premier speech technology and solutions event. This will be my 10th consecutive SpeechTek conference and my first time attending as a vendor.
I remember in the early 2000s when Nuance was blue, Nortel Periphonics was the leading IVR platform and all the buzz was around open standards like Voice-XML with a little bit of SALT, X+V and MRCP. At this time the discussion was very technology-focused as the market was still immature. Then came the era where emphasis was on reusability and development effort as the focus shifted more towards speech applications. This was when RDCs, packaged speech applications and SCEs were the headliners at SpeechTek. Then came the time when IVR became more multi-faceted with supporting outbound and dynamic prompts. CCXML, proactive communication, SIP and personalized IVR were the hotbeds of activity. Somewhere in there was a short-lived industry movement towards standardizing IVR menus in response to the growing popularity of Paul English’s GETHUMAN. The movement fizzled out quickly and never amounted to much. In more recent years the focus shifted to ID&V and richer query to transaction engagement. Headlining were voice biometrics, mobile search and speech-enabled virtual assistants (like Siri).
Now the big question is what will be the prevailing theme in this year's SpeechTek? It seems from the SpeechTek website that there will be much discussion on the role of speech recognition in today’s age of ubiquitous computing (smart devices, cars, game consoles, TVs, etc) from a consumer perspective. However, the presentations, discussions and sponsoring vendors are still largely skewed towards customer self-service in the enterprise. This has been the case for the past decade and it makes sense, as the majority of spending on speech recognition falls in the enterprise market for customer service.
But, when it comes to customer service, the experience should not just focus on speech as a UI and self-service mechanism but rather focus needs to be on integrated channels of customer engagement with speech providing just one interaction layer. According to Ovum research, nearly three-quarters (74%) of today’s customers use at least three channels when interacting with a company for customer service.
While today’s customers expect an omnichannel experience (one in which context is persistent across all channels) most companies still exist in a multichannel world where interactions are siloed and context is not passed on from one channel to another.
Businesses have to get smarter. Supporting omnichannel is one step in the right direction. Another critical component is anticipating what customers want to do (in real time) and successfully guiding them to achieve what they set out to do with the least amount of effort. This is where it gets interesting. In order to make it simple for customers to connect with companies in today’s omnichannel age businesses need to have a big data-centric approach where they leverage insights gleaned from every aggregated customer interactions from calls, chat, web, location, social media and in-store transactions in conjunction with CRM data. Done the right way businesses can then predict with precision the intended customer journey whatever the channel and optimize each touchpoint to improve the experience.
Next week at SpeechTEK, I’ll be presenting in the "Giving Customers a Unified Digital Experience" session with Mike Monegan, (7.ai’s VP of Applications) where we will discuss the intersection of big data and omnichannel and how using prediction will provide a unified, channel-agnostic digital experience. Mike will also be on a panel “The Changing Role of Speech in a Multichannel World.”
I hope to see you all there.