“Conversational self-service”? You may say that sounds an awful lot like chatbots. Well, there are certainly some similarities between the two, particularly the use of NLP to enable conversational interactions, but self-service automation entails a lot more than just responding with cute answers to canned questions. In this post, we explore some key considerations for a full-fledged customer self-service automation platform. Hopefully, the differences between such a platform and a simple bot-building tool will become evident as we go along.
There is no dearth of research to substantiate that there is growing desire for immediacy in customer service¹. Customers no longer want to spend time in long IVR queues waiting to speak to the next available agent. Instead, they prefer to use online research, knowledge bases, peer-to-peer communities, and other self-service options to locate answers.
Self-service isn’t a new concept – it has been around for a while in the form of online FAQs, self-serve kiosks, and mobile apps. However, the exposure to slick experiences offered by customer-obsessed digital natives, such as Amazon or Uber, means that traditional self-serve approaches that are complex to use and limited in reach are no longer sufficient.
Consequently, organizations now need to rethink their self-service strategy.
In this post, we look at how enterprises can elevate their CX by enabling AI fueled self-service experiences over customers’ preferred messaging channels such as SMS, Facebook Messenger and WhatsApp.
Conversational self-service: A deeper look
The speculation and hype around chatbots is at its peak. The ability to automate customer interactions over customers’ preferred communication channels such as SMS, Facebook Messenger and WeChat has put chatbots driven customer service squarely at the center of all CX related discussions.
However, delivering seamless conversational self-service is not as easy as it sounds. Following are the key requirements for facilitating an enterprise-grade, conversational, customer self-service experience:
Identify contextual customer intent
Identifying customer intent is arguably the most critical aspect of delivering conversational self-service. If you can’t understand what the customer is trying to achieve, the rest has little or no value. Many NLP tools try to extract intent only from the current customer input, i.e. one message at a time. Such an approach works only for simple FAQ or Q&A based chatbots. For broader self-service automation, intent should be derived based on customer input over multiple interactions as well as context derived from previous interaction history and customer actions across touchpoints.
Channel-agnostic, yet channel-aware experience
Customers today use multiple channels to interact with businesses. A comprehensive platform should allow you to extend consistent self-service automation capabilities across all channels. This requires two key capabilities:
(a) The conversational intelligence needs to be abstracted away from the interaction channels so that a common intent engine can be used to parse customer messages coming from various channels. Such an approach eliminates the need to build separate chatbots for each channel and/or replicate changes in multiple places.
(b) The response engine needs to be channel-aware to leverage the advanced interaction capabilities provided by some of the new channels. For example, while a video can directly be shared as an embedded file on WhatsApp, on SMS you can at best share a link to the video. An advanced example would be proactively shifting the conversation from one channel, say SMS, to another, say Facebook Messenger, when a photograph (e.g. photo of the customer’s driving license) needs to be obtained or a customer’s location needs to be shared with a field service engineer.
Authenticate customer identity without breaking the flow
When the self-service flow involves dealing with sensitive information exchange or high value transactions, it becomes necessary to establish the identity of the customer. For seamless customer experience, it is important to carry out this authentication flow over the same channel where the customer started the interaction.
Integrations with backend systems for context-aware service and end-to-end automation
We have all had incidents wherein we had to narrate the entire conversation history to multiple service representatives because the IVR connects you to a new agent every time you call to follow-up. With conversational self-service, you have the opportunity to eliminate such issues and delight customers with context-aware experiences. This needs integration with backend CRM and customer service systems to check if there is an open issue for the customer who has just reached out. If yes, personalize their experience by eliminating redundant intermediate steps and addressing the core issue at the very start.
Further, facilitating end-to-end automation of customer requests – with the necessary back-end system integrations – is a critical platform capability when it comes to self-service. E.g. allowing customers to change their address, activate a new service pack, move money between accounts, change direct debit dates, etc. are all core services and the self-service experience would be incomplete without automated fulfillment.
Fig. Key capabilities of an enterprise-grade conversational self-service automation platform
Seamless handover from chatbots to live agents and vice versa
Customer queries or requests that chatbots cannot handle should be transferred to live chat agents for further handling. In such scenarios, the ability to direct conversations from multiple channels to an agent console while retaining the context becomes a key capability. Likewise, once the agent has addressed the issue at hand, the conversation should be handed back to the chatbots for further engagement as the customer desires.
Growing self-service scope over time
The underlying platform should allow you to expand the scope of self-service without going through major re-engineering cycles. One of the ways to achieve this is by deploying a ‘manager bot’ that identifies the intent of the customer based on initial message exchange and hands over the conversation to a ‘child bot’ responsible for handling a particular use case. Such an approach allows for adding new self-serve cases and/or making amendments in selective use-cases without having to switch-off or re-engineer the full solution.
Learning from previous conversations
Even when you have built your service experience based on extensive conversation data and multiple rounds of internal testing, it is natural to come across new conversation patterns and unhandled cases. A good self-serve automation platform should allow you to analyze past conversations, identify patterns and address them easily.
Proactive as well as reactive customer engagement
Traditional customer self-service channels such as online portals and mobile apps have focused on serving customers after they have identified a need and/or problem. While it continues to be a key requirement, such reactive engagement alone is not sufficient in today’s age. Organizations need to proactively anticipate and fulfil customers’ needs. Messaging based self-serve automation is now making it possible by allowing you to trigger a contextual message, based on a business system event such as an appointment falling due, a missed payment, upcoming direct debit, expiring contract, exhausting data pack, etc., on customers’ preferred messaging channels. Customers can simply reply to these messages and go through the self-service flow.
AI powered conversational self-service can lead to tangible benefits in both pre-purchase as well as post-purchase conversations. In the pre-purchase phase, self-service allows customers to locate new products and offers, get answers to specific product queries, locate details of associated warranties and support, and proceed with the purchase quickly without human intervention. In the post-purchase phase, self-service can be used for seamless on-boarding and installation, customer education, feature/product usage tours, and automated in-life complaint/query handling.
Along with CX improvements, self-service automation also leads to significant cost and efficiency benefits for businesses. It allows customers to find answers to their routine queries themselves enabling agents to handle higher value customer interactions that require sophisticated personalised intelligence.
However, selecting the right platform to power your self-serve automation is key to leveraging these benefits.
Stay tuned and share with us how you are planning to use conversational self-service to improve customer experience. You can find out more about our Platform-as-a-Service approach to self-service automation on our website home.
Request for IMIconnect trial.
1. Research by leading Forrester Analyst Kate Leggett: Your Customers Don’t Want To Call You for Support and Online Self Service Dominates Yet Again. Why? Its An Effortless Way To Get To Your Answers