How Conversational AI Diminishes Customer Service Costs, not the Experience
Moving customers from traditional to digital channels where AI-powered Automation can kick into gear.
Why Live Channels need to give way to Digital Ones
Here’s a typical scenario we’ve probably all encountered. You’ve purchased a product online and although the status says it’s been delivered there’s no sign of it despite waiting patiently for a day or two in case there’s some kind of mix up. What do you do?
Right! You go looking for a way to contact customer service. This can involve a multi-click hunt around a website to find a phone number, a contact form, an email alias, or a human agent lurking behind a friendly-looking pop-up that declares they can help you instantly, i.e. live chat.
Welcome to the world of customer service where the most popular channels of communication are still the old traditional ones of the phone, email, contact forms, and live chat.
Now consider how you communicate with friends, family, and even colleagues at work. You probably use text messaging, social media apps, or messaging channels like WhatsApp, Facebook Messenger, Slack, and Teams more frequently than you use a phone call or email. Why? They tend to be faster and friction-free.
So why then are businesses still driving us to 1–800 numbers, customer service email aliases, contact forms, and live chat agents? It doesn’t seem natural for a business to point their customers to traditional channels rather than the more convenient and common digital channels like SMS, social media, web portals, mobile apps, and messaging channels. But they do! And therein lies not only the major cost driver in customer service but the challenge of meeting today’s expectations for speed and convenience.
For some organizations, digitally transforming customer service wasn’t a priority or was on a slow burn — until COVID hit. Now the world has shifted dramatically with a renewed focus on automating customer interactions to improve service outcomes and customer engagement while reducing the cost to serve. And conversational AI technology is enabling this new form of automation.
Here are four simple benefits of moving customers to digital channels powered by AI.
1. Going Digital Means More Self-Service
Phone calls, email, contact forms, and live chat have one pitfall in common. They are all dependent on the availability and skills of live customer service agents. This means that they carry high costs as each minute an agent needs to respond to a customer and conduct the necessary tasks to resolve their issue incurs cost. In Gartner’s 2019 Customer Service and Support Leader Poll they identified that:
“Live channels such as phone, live chat and email cost an average of $8.01 per contact”
While the cost of resolving customer issues via phone, email, or live chat varies by industry, region, and by the unique characteristics of a business, this average cost of over $8 per contact is prohibitive. As a result (and especially in times of crisis such as with the current COVID pandemic) self-service models have taken off. Savvy businesses are increasingly shifting customers towards digital channels where self-service can be more easily enabled, and where the volume of calls, chats, and contact emails can be reduced.
The same Gartner report highlights the cost-benefit of self-service channels which really makes them a no-brainer.
“Self-service channels such as company-run websites and mobile apps cost about $0.10 per contact.”
2. Digital Channels Open the Door to Conversational AI
Besides the fact that traditional channels are very labor-intensive, they are also more challenging when it comes to automating them.
An IVR can help better route a phone call but it is still highly dependent on a human at the other end of the interaction. Emails can be lengthy and subject to interpretation, requiring a lot of manual time and effort. Even contact forms that are well-designed to categorize customers' issues fill up agent’s email inboxes. And finally, live chat is powered fully by agents during business hours. Even though they may be able to handle some concurrent conversations, it still has proven to be a costly channel.
In order to make customer service agents more efficient, contact centers have deployed different technologies over the years in an effort to bring more automation to interactions and optimize the average handling time (AHT) while balancing it with customer satisfaction measures (CSAT, NPS, etc.).
The IVR, for example, has helped better route customers to agents that are trained in a particular subject matter and can even take them to automated responses or to a voice mailbox to leave a message if so chosen. Technology investments have also been focused on the agent desktops where multiple applications make information more freely available to them so they can respond faster and more accurately. But the fact remains that there is still a lot of costly time spent by human agents responding to common queries that, thanks to artificial intelligence (AI) technology, can now be alleviated through automation.
Digital channels like websites, mobile apps, SMS, and other messaging apps that support text-based interactions lend themselves better to automation powered by Conversational AI.
Think of a chat you’d have with a person via SMS or WhatsApp — it’s more like a natural conversation with a stream of short messages making up the conversation.
This makes it easier for an AI assistant to detect and extract the information needed and gather any missing pieces necessary to take appropriate actions. Intent detection and extraction of information get trickier in a lengthy email or voice message so can be more challenging for the AI model.
3. Digital AI Assistants Automate Engagement
Like IVR, an AI bot or digital assistant can be deployed as the first point of engagement that the customer has in trying to tackle their issue or find the relevant information, while still providing a path to human agents. However, unlike IVR, an AI bot can often fully resolve the customer’s issue, without involving a human agent. An AI assistant understands and responds to a customer using natural language, similar to how a live chat agent carries on a conversation.
By either deflecting calls to a messaging channel where a bot can engage with a customer, or by putting it in the customer service page or contact form of the company’s website or mobile app, customers can choose to self-serve via the bot rather than wait until an agent is free to take their call or conduct a live chat conversation. Hence the term “automation-first”. This is a nod to the fact that bots can do a lot to help, but there are also things that they may not be trained for, or simply can’t do well (empathy, for example!).
Using the power of Natural Language Understanding (NLU) and Machine Learning (ML) technologies, AI assistants can understand a customer’s intent and are trained to respond and take necessary steps to resolve their issue. Whether it’s seeking order status information, tracking a delivery, reporting an issue, onboarding with a new policy, making a payment, updating account details, or finding information, these conversational AI solutions either eliminate human intervention by providing full automation of common requests, or they partially automate, prioritizing the customer and/or the issue, and gathering pertinent information before passing it on to a live agent, or scheduling a call back to resolve the issue.
4. Supporting Live Agents, Not Replacing Them
Rather than replace human interactions altogether, AI assistants can take the burden of automating or partially automating routine customer requests, leaving the unique or complicated issues for live agents to deal with. This alleviates a lot of pressure on contact center agents and benefits the business by reserving skilled human resources to deal with more complex customer transactions, sales, or complaints. By using AI to triage customer engagement, overall agent productivity increases as they don’t have to deal with repetitive and simple tasks.
This automation-first approach using Digital AI Assistants provides the best of both worlds. When the bot can’t respond or resolve customer issues, or if a customer asks to speak to a human, there is always a path to a live agent so that the customer isn’t left hanging or frustrated. This can either be through integration with the organization’s live chat system where the bot can escalate immediately to a live chat agent, or it can be the bot gathering necessary details and scheduling a callback during business hours. This helps a business achieve an optimum balance between cost-reduction and customer experience, especially in times of crisis such as the COVID pandemic.
An Easy & Powerful Conversational AI Solution: Deflect, Automate, Escalate
Circling back to the beginning of this article where I summarized some of the pitfalls of customer service centers using traditional labor-intensive channels as the primary means for customers to initiate their service requests, let’s explore a strategy for deflecting customers away from these and towards a more efficient digital and AI-powered solution.
The term call deflection may have historically negative connotations around call avoidance but today the term digital or contact deflection is about giving customers options to shift to a more efficient channel where they can self-serve and choose to have their issues dealt with by a digital AI assistant.
Deflection
When a customer calls they can be given an IVR option to choose to continue the engagement via a messaging channel on their mobile device where a bot is made immediately available to them and offers a range of common customer service options that enable self-service.
If the customer initiates a query via the organization’s customer service webpage, mobile app, or contact page, equally a digital assistant can be made visible on the page, encouraging those that are willing to engage with it to do so. The net effect of this is to offer them the digital alternative to sending an email, completing a contact form, or initiating a live chat session — in the knowledge that they get immediate information that is likely to resolve their query rather than have to wait for a human agent.
Resolution
A well-conceived customer service bot solution is designed to handle some of the most frequent issues that customers encounter. This could be requesting a document or form (which the bot can provide for them to instantly upload to their device), tracking an order, scheduling an appointment, updating account details, or responding to queries about a product or service.
The bot detects the customer intent using AI and resolves the problem if possible.
Deferral
While a good percentage of customer requests may be resolved by a bot, it’s natural that not all of them can be. There will invariably be unique issues and complex queries that the bot hasn’t been designed to handle and a good fallback position is essential. If the issue isn’t immediately urgent it may be appropriate to defer it. In this case, the bot can triage the customer request, gathering relevant details and information and preparing a work ticket for a customer service rep to act on at a later time.
Escalation
Not all customer queries are a priority but some are. For this reason, having a path to handover to a live agent is ideal. The bot can be designed with rules in place to prioritize the escalation, integrating securely with the company’s live chat system for seamless handover with the necessary chat history and information for the live agent to act on.
Conversational AI works best when there is a path to a human so the concept of it replacing humans in the contact center or customer service department goes against logic. What it does though is to take the burden of handling the repetitive things away from your employees and allowing them to handle the more complex issues or the customers that are more comfortable with a human versus a bot.
There’s no doubt that advances in AI technologies in the area of natural language processing and machine learning give businesses new opportunities to automate customer interactions not only in service scenarios but across the whole customer life cycle.
If this is something that you’re business would like to learn more about please reach out and our ServisBOT team will be happy to chat and show you some great examples of conversational AI at work.
