Why wouldn’t you want to give your customers a full-service experience in an automated fashion? Wouldn’t it be the most optimal outcome for both customers and providers? The obvious answer is yes, and this is why virtual assistants will take you there.
What we know as web self-service, is a collection of knowledge management and search tools that attempt to surface relevant content to end users so that they can discover resolutions without human assistance or channel shifting. Since the emergence of virtual assistants for businesses, people tend to place them in the same category as the previously mentioned self-service tools. Today, I will contrast a fundamental difference of approach and capability between search and virtual assistants - and which one is more suited to a given situation.
Search and knowledge management - masters of research
Foundationally, web self-service tools rely on search mechanisms. While they employ a variety of techniques (such as frequency of document access) to bubble up the most appropriate content, these are little more than crutches to assist users in their attempt to guess at term combinations that will result in the most relevant search results.
This is not to say search is a bad technology - quite the contrary – however, there is the consideration of the right tool for the job. When it comes to researching subject matter - search can't be beat. Recently, I took an interest in a technology called Node.JS. I did not have any particular question in mind. My objective was to immerse myself in the subject matter and understand it from multiple dimensions. This requires research - the act of locating and filtering my way through a variety of documents relevant to the subject- and search is excellent for this purpose.
However, search is not ideally suited for customer service, and here’s why. People are usually dealing with a specific need and not a desire to develop a broader understanding of some subject matter. Unfortunately for knowledge management tools and their novel tactics - the search paradigm remains a faulty foundation to addressing such specific user needs.
Customers do not have the patience to wade through document collections or learn each knowledge management application. According to the Harvard Business Review, up to 57% of inbound calls come from customers that start on a company’s website. This clear failure of the website as a self-service channel is also indicated by the growth of live chat. According to a Forrester report, chat usage rates as a customer service engagement channel have grown from 30% in 2009 to 43% in 2012.
The difference is that chat is not a "self-service" channel in the traditional sense; it is a means to identifying a solution to a specific problem or need through direct assistance. The question then becomes not a matter of self-service versus chat - that is not really an option. What is an option, however, is employing virtual assistants to automate your direct service channel.
Virtual assistants - better at directly addressing a specific need
Virtual assistants are not search - they have no comprehension of the contents of your web site or its associated knowledge bases. Their understanding is focused entirely on interpreting or understanding the specific needs of your end users. Once such understanding is codified, it becomes repeatable - virtually guaranteeing that all future users with that same issue will have their needs resolved in a consistent manner. It is a minor investment for a human subject-matter expert to provide an ideal response with links to supporting content a single time, especially when you consider that from this point forward your users will be serviced with the most optimal path to resolution possible.
When best practices are encapsulated in an automated way, the result can actually increase the efficiency for end customers as they will receive resolution more quickly. Charter Communications found that live operators averaged over 13 minutes to resolve a password reset, while the virtual assistant averaged just 6 minutes and 40 seconds - almost half the time to resolution and have significantly deflected the number of live chats they receive, experiencing an 83% drop in live chat volume.
What virtual assistants are then, is not a self-service capability but a way of automating a communications channel with an ROI that parallels the automation of the traditional voice channel with IVR - albeit with a notably higher success rate in the on-line chat channel.