Give Customers What They Want: Great Self-Service

Donna Fluss

Delivering an outstanding customer experience (CX) is a top goal for companies, as it is an essential and measurable differentiator between otherwise commoditized products and services. Executives have known this for years and have started to make investments in their service organizations and contact centers to improve the CX, particularly if the investment has a quantifiable payback.

Self-service is one of the areas that has captured the attention of C-level executives, particularly now that it has become the preferred method for first-line support for many customers. Customers are demonstrating a desire to help themselves, regardless of the channel in which they interact with any organization, whether it is a company or government agency. As self-service solutions can deflect a large percentage of inquiries and interactions from live service and sales resources, they provide an ideal method for improving the CX while reducing operating costs.

DMG estimates that more than 90 percent of the self-service solutions in use today are outdated and need a complete overhaul of their technology, functionality, and user experience. Many of these voice-based self-service solutions were built more than 10 or 15 years ago, when the technology was limited. Organizations are still using touch-tone and rudimentary speech-enabled interactive voice response (IVR) solutions and prompters. Given the substantial enhancements and innovation in the area of self-service technology and consumers’ growing preference to help themselves, it’s time for organizations to update these customer-facing applications.

Solving the Problem

Artificial intelligence (AI)-enabled omnichannel intelligent virtual agents (IVAs) are the future of self-service. These solutions are game-changers because they allow customers to converse naturally instead of going through a series of nested questions and options. Industry best practice is for companies to identify the information, functions, activities, and transactions their customers want to handle by themselves and automate them with intelligent conversational omnichannel solutions.

Intelligent virtual agents are powered by sophisticated speech technologies, AI, machine learning (ML), analytics, and more to enable them to mimic human cognitive functions and interact in a conversational manner in voice and digital channels. Intelligent virtual agents leverage machine-readable, context-aware knowledge bases (or other data sources) to retrieve the information needed to respond in a personalized and contextually relevant manner to human questions or input. Machine learning continuously improves their accuracy and effectiveness over time based on knowledge gained from each interaction, which is assimilated and leveraged in future conversations and transactions.

As importantly, virtual assistants (VAs), which use many of the same technologies as IVAs, are being used to provide agents/employees with context-aware guided support and information. Both IVAs and VAs are proving effective at intelligently handling inquiries from external customers and internal employees.

IVAs are intended to support omnichannel interactions (voice, chat, messaging (e.g., Facebook Messenger, Telegram, WhatsApp, etc.), mobile apps, text/SMS, email, social media, and other digital channels) and enable seamless migration between channels with retained context. If needed, this includes automated escalation to live agent support, in the customers’ channels of choice, routed to the right resource to resolve the issue.

Intelligent virtual agents can use both explicit and implicit measures to identify the need to transfer to a live agent. Explicit methods include client-defined business rules, sentiment thresholds, and customer requests, and implicit ways can be based on the IVA’s determination of customer intent, negative sentiment-oriented language, and/or customers’ past preferences. When necessary, an IVA initiates a warm transfer by passing a transcript of the self-service interaction, browser history, cart contents, and, in some cases, predicted intent, to an appropriately skilled agent via screen-pop or an integration with the CRM system. Once transferred, the IVA can perform double duty by providing the agent with context-aware guided assistance on the topics and tasks that are relevant to the specific customer interaction, along with suggested responses, in a channel-optimized format.

Finally, the transcripts from the escalated interactions can be retained and used to further train the IVA or, in more technically advanced solutions, the IVA can learn from agent interactions by modeling how to respond to questions, new intents, and new language, in real time, using a supervised machine learning approach.

The growing preference of consumers for self-service is giving organizations a great opportunity to automate the handling of inquiries, transactions, and tasks that do not require the cognitive capabilities of live agents. However, organizations need to realize that this is not a field of dreams where customers will use anything they are given.

Enterprises and government agencies need to dedicate the time and effort to take a fresh look at all of their self-service capabilities. This includes building an organization-wide self-service strategy that aligns with CX goals, finding all the departments and functions where self-service should be used, asking customers/constituents what they want to handle by themselves, and selecting the right technology and implementation partner (who may be the same) to build the applications. It’s time for companies to listen to their customers and enable them to help themselves, particularly as doing so will enhance their brand and bottom line.