The growth of Conversational Commerce is emphatically customer driven. The population of 3+ billion smartphones are now in use for web browsing, search, messaging, making payments and carrying on conversations. These powerful devices are now augmented by a fast-growing population of “smart speakers” which have caught the attention and imagination of digital strategists spanning banks, packaged goods providers, beauty products specialists and mass media, in general.

Customers are already digital. Brands are scrambling to catch up.

It is fascinating to look at the mix of job titles that are taking an interest in how conversational technologies are deployed. At the “C” level, Chief Marketing Officers were among the first to take interest in new channels. Now there are Chief Digital Officers and Chief Innovation Officers joining the ranks. They are most often involved in setting the strategy for “going digital.” What’s been fascinating to observe is the groundswell of interest among individuals at the business unit level. As I scanned the titles of individuals registering for our Webinars on Conversational Commerce, I see “Community Manager”, “Customer Success Manager” and “Digital Product Manager” interspersed with more descriptive titles like “Disruptive Technologist”,  “UX Lead” and “Listener”.

These folks are not technologists or part of DevNation. They are business unit-level employees who find themselves to be both champions and beneficiaries of core conversational technologies. And the rest of this post, designed as a technology primer, is dedicated to them and the overall effort to democratize Conversational Technologies for brands and their customers.

Deconstructing AI

The effort starts with parsing “Artificial Intelligence” into relevant components and resources. When it comes to Conversational Commerce, Natural Language Processing (NLP) is central. It plays the role of deriving meaning from imprecise and meandering utterances or typed-in phrases. It is closely mated to, and actually a subset of Machine Learning, which is foundational for an automated agent to stay current to new terms and phrases and tuned to provide correct responses at scale. Because human judgment is required to keep all these automated systems on track “Supervised Machine Learning” is the indicated best practice, providing mechanisms for subject matter experts and knowledgeable agents to check and validate the responses of automated resources.

More recently Deep Neural Networking (DNN) emerged from the recesses of supercomputing with fuzzy, and opaque-by-design, processes that enable life-like virtual agents to converse in multiple languages and make sense out of conversational inputs that not only involve turn taking, but also take the sorts of unexpected turns or branches that we humans can often tolerate but stymie automated systems.

A Special Role for Speech Analytics as a Gateway Investment

This year Speech Analytics rockets up the list of transformative technologies. Products, services and vendors that fall into this category have long been associated with providing solutions that detect patterns in spoken input (mostly stored in the call recording apparatus behind contact centers) for the purpose of detecting root causes of agent failures or isolating those moments when a conversation falls out of compliance with standing regulations.

In the former instances, speech analytics platforms save contact center managers the time and effort of listening in on every call – which is an impossible assignment to begin with. Analytic lenses can be applied to stored voice in order to detect phrases or agent behaviors that correspond to negative outcomes, like subscriber churn, or positive outcomes, like measurable increases in customer satisfaction or net promoter scores.

We’ve completed field work and are preparing to publish the 3^rd annual assessment of speech analytics around the world (sponsored by Uniphore). One of the major findings, which will be a topic of discussion at the Conversational Commerce Conference in San Francisco in September is that executives in companies that have invested in and deployed speech analytics to support workforce optimization (WFO) and training of customer care agents in contact centers now recognize that the latest solutions are valuable resources for informing and training virtual assistants and chatbots.

In this age of “Big Data and Analytics,” spoken conversations are a deep reserve of “unstructured data” and speech analytics solutions, as their name says, perform rudimentary pattern matching to recognize or predict the intent of a customer or prospect and start the process of providing the proper response. That makes speech analytics one of a handful of technologies that companies have invested in that will retain their relevance and ultimately increase in importance as digital, optichannel self-service morphs into multiple, ongoing conversations.

You’ve Always Had the Power (To Listen and Understand)

There’s a Wizard of Oz moment for the individuals in charge of bringing Conversational Technologies into their digital commerce, customer care or optichannel strategies. They have already made significant investments in contact center, social marketing, CRM, Dot Com and mobile commerce infrastructure. Very few are starting at square one and my suspicion is that execs attending our Webinars and Conferences, especially the one with the title “Listener” will be pleasantly surprised to learn about the platforms and tools that are designed to make their lives easier.

Conversations are largely about listening and understanding, not coding or architecting. The democratization of Conversational Technologies is underway largely because solutions providers recognize that their ultimate end-users are people who want to use their own words to search for and find the products they want or to describe and resolve the issues that are of greatest concern to them. Whether they are community managers, UX specialists disrupters, or simply listeners, conversational commerce champions are going to find that their investments in Speech Analytics, CRM and other elements of contact center infrastructure, they’ve had the power all along.

Categories: Intelligent Assistants, Articles