Conversational AI — It’s all about Automation.

Enterprise AI
Jul 15, 2020 · 9 min read

Conversational AI may sound like a daunting term that brings to mind something that is complex, costly, and only for very specialized use cases or industries.

However, the reality is that many businesses are embracing conversational AI solutions in the midst of the COVID-19 pandemic as a means to more easily transition to efficient digital operational models, leveraging digital AI engagement and automation technology.

While automation may not seem particularly new or exciting, this recent wave of AI-powered technology is creating very timely opportunities for business leaders to fast-track their digital transformation efforts and better respond to the dramatic impacts of the COVID-19 crisis.

And once they realize the compelling benefits of conversational AI there is no going back.

COVID-19 Spurs Automation-First Approaches

The year 2020 has made it especially critical for businesses to prioritize automation and digital engagement initiatives powered by AI bots.

As the COVID-19 pandemic crisis hit, business operations were instantly impacted in ways that may alter their path for a long time, maybe even forever. With staff deployed from office spaces to remote working from home at a time when some industries like travel and insurance saw over 400% increase in incoming customer calls, an urgent spotlight was put on digital operations and cost-efficiencies.

Source: Pindrop: 5 Insights Of Covid-19 In The Contact Center

Even though digital transformation efforts were long underway in many businesses, many found their initiatives were lacking or insufficient to meet the demands of the new COVID era.

In responding to the crisis, interest in AI technologies that reduce operational costs and optimize scarce resources has flourished. Two key AI-powered technologies that have become central to advancing automation are Conversational AI and Robotic Process Automation (RPA). While these two emerged as distinct technologies, they share the common goal of reducing or even eliminating costly manual steps in a process.

RPA applies machine learning to highly repetitive and rules-based business processes, using software robots to mimic human actions in large scale and repetitive processes, eliminating much of the previous need for manual actions and keystrokes. In the absence of Application Programming Interfaces (APIs), RPA can automate tasks, for example, by taking data from a spreadsheet and inputting it into legacy backend business systems. Generally applied to large-scale backend processes like invoicing, claims, and loan processing, RPA is responsible for automating repetitive tasks at scale, especially where APIs are not available.

Conversational AI, on the other hand, applies natural language understanding (NLU) to understand customer conversations and then automate the actions that need to be taken to solve their issues. The focus is on using natural conversation to more accurately understand user intent and context, respond in an intelligent way, and automate the tasks needed to solve the user’s issue. When integrated via APIs with business systems, these conversational solutions offer businesses a more personalized approach to handling customer interactions at scale, reducing reliance on human agents for less complicated tasks and lowering the cost to serve.

It’s easy to see how conversational AI can bump up against RPA as front-end customer interactions trigger different workflows, some of which are automated by the customer-facing AI assistant and others by RPA robots.

While automation has become a pressing goal in these times, it is an automation-first approach that we believe brings the most success. This takes into account the fact that not everything can be fully automated and a path to escalate to a human is always a wise move, either in the case when the AI can’t fulfill a customer need or the user requests a handover to a human. By putting a digital AI assistant to work at the customer interface and in a suitable digital channel, automation can kick in to handle the conversation and resolve the customer issue. And because a bot can understand natural language, provide personalized responses, and automate workflows, it can either fully resolve a customer issue or gather enough information to hand it off to a live agent for further handling.

The Rise of the Conversational Interface

History has shown us that new technologies often emerge enabled by advancements in and accessibility to other technologies. For example, our mobile phones wouldn’t be what they are today without progress made in areas of technology like computer processing power, communications, touchscreens, batteries, and cloud technology.

So to understand why Conversational AI has emerged as such a force for digital transformation it’s interesting to take a look at how different interfaces have evolved and matured to support how customers engage with businesses and search for information.

Over the past three decades, different interfaces have supported us as consumers, from websites and email to mobile and social media apps, and more recently to conversational interfaces. It is the recent shift from graphical interfaces towards conversational interfaces that is monumental for business. Now we can interact, transact, search for information, and purchase products and services using voice or text-based chat on our preferred digital channel and around the clock.

Advancements in natural language processing (NLP) and natural language understanding (NLU) mean that digital AI assistants can understand context and intent in a chat, extract the relevant information in order to execute next steps, and respond to customers in fluid and seamless conversations.

When integrated with business systems via APIs, AI bots can authenticate customers and access their information securely, providing the added benefit of conducting personalized conversations at scale. They can also work collaboratively with human workers and hand over to them if needed or if requested by a customer. It’s the integration of AI bots with business systems and processes that make them way more powerful than simple FAQ chatbots.

Four Conversational AI Solutions to the Rescue

It’s important to think of conversational AI solutions in terms of the complete customer lifecycle, and not just isolated customer service interactions. Digital AI assistants can help a customer from the moment of their first interaction searching for information and engaging in a sales cycle, keeping them engaged throughout and encouraging them towards converting to purchase. As the COVID crisis hit, organizations quickly turned to conversational AI solutions to help them surmount some new and unexpected challenges. Here are four powerful conversational AI solutions that are fast-tracking their digital transformation and positioning them for success now and for the future.

As the pandemic hit, many businesses saw huge surges in inbound requests to their customer support centers. From mortgage payment deferral requests, to travel insurance queries or general virus-related helplines, customers needed urgent assistance. Without sufficient agent capacity to handle extra demand, many needed to quickly deploy a contact deflection solution that shifted customers from inefficient channels like phone, email, and live chat towards more automated self-service digital channels.

Once a customer is shifted to these channels, the benefits of automation using a digital AI assistant become evident as it can handle requests and either fully or partially automate the tasks required to fulfill the customer’s intent. There are a number of possible treatment paths once the digital assistant has control of the contact:

  • First, the digital assistant will try to resolve the query using AI and trained responses.
  • If the AI can’t resolve it, the bot will try to gather enough information to handle the query in an offline manner. We call this contact deferral. This allows the contact center to smooth the peaks and troughs normally associated with operations.
  • If the contact can’t be deferred or if it is a high priority, the contact is handed over to a live agent.
Source: ServisBOT: Contact Deflection — A Path to Automation

Using this triaging approach allows the contact center to prioritize and handle many more queries in an efficient manner while lowering the cost of service.

What these businesses have discovered is that fairly simple AI-powered contact deflection solutions can be launched in a short time and result in immediate positive impacts; productivity shoots up, the cost to serve drops, and customer satisfaction increases as their issues are handled faster.

Thinking beyond automating typical customer service interactions, take the example of where a conversational AI solution helps automate an online sale and increase sales conversions. Why would a web-based online process need a bot you may ask? Surely something like an online quotation form is already automating the workflow — right?

If you’re an insurance provider you know that some customers drop off throughout an online quotation process, others complete the form inaccurately due to a misunderstanding of terms, and others try to guess what the best answer is to lower their rate. So using an AI bot to keep an online customer engaged and informed throughout the sales process or re-engaging with those that fell off is a sure way to increase sales conversion rates as can be seen in this chatbot case study. And when your digital advertising costs are high this is especially valuable.

In this use case, the bot can assist a customer throughout the online quotation and sales process. explaining terms and answering questions, customizing the offer and price to their needs, and re-engaging proactively with those that dropped from the process. Not only does this enable more out-of-hours automation of a sale but it diverts routine queries from contact center agents so that they can focus on more complex customer issues and close these sales.

Onboarding new customers is a critical process for industries that require a multi-step journey to get customers started using a product or service. It can be a costly and complicated activity, especially in cases where copies of documents need to be collected from the customer and manually uploaded and reviewed by an agent. This can often be the most time-consuming part of the onboarding process — contacting the customer for copies of documents and chasing them up to respond in a timely manner. It is especially difficult to connect with the customer when consumers are no longer answering phone calls from unknown numbers.

It’s also not uncommon for customers to complete some onboarding steps in face-to-face settings. For example, new mortgage customers often choose to go through onboarding with a loan officer at the nearest branch. The pandemic, however, has put an abrupt end to in-person onboarding and forced companies to leverage better digital AI engagement models where they can apply automation techniques.

Take the case of onboarding an auto insurance customer. This involves some tedious gathering of proof documents such as the customer’s driving license, banking information, no claims bonus or proof of previous insurance, social security number, vehicle identification or registration number, among others.

By implementing an AI digital assistant, the insurance provider was able to dramatically decrease the manual effort and costs involved in getting customers started with their new policy. Before deploying the conversational AI solution, human workers handled dozens of emails and document attachments per day, gathering all the different documents, validating them, and collating them so they could be attached to the customer policy record. With agent productivity increased 8x due to this new onboarding process, agents can be put to work on other higher-value tasks.

Conversational AI solutions are not only for inbound use cases. AI assistants can also act in a proactive manner, reaching out to engage, or re-engage with customers. Now that the pandemic has shifted everyone to virtual channels, it is even more important to be able to connect with customers through these channels and keep them engaged in ways that nurture the relationship and create loyalty during and beyond the pandemic.

Proactive outreach is relevant for business use cases like collections, onboarding, and renewals, but it can also be effective in re-engaging with users that have dropped off during a sales or onboarding process or to check in on customers to see if their needs have changed as a result of the crisis.

Collections is an especially sensitive topic now as businesses try to collect overdue payments or offer deferred payment plans to customers that face financial difficulties. A mortgage provider, for example, can deploy an AI bot that reaches out proactively in a customer’s messaging channel to notify them of overdue payment and advise them of eligible mortgage relief or forbearance options. The bot can handle much of the conversation and be integrated with the live agents where it can escalate more complicated issues to them.

Digital AI assistants can work alone on a messaging channel to automate a specific customer interaction; they can support a customer to complete a web-based journey more efficiently, or they can help a live agent resolve a customer issue.

Whatever channel, format, or use case they are implemented for, the key benefit of Conversational AI is automation.

And now more than ever, automation really matters.

Conversational AI Solutions

Automating Digital Engagement

Conversational AI Solutions

Insights for businesses on deploying conversational AI assistants to automate digital engagement and improve business results.

Enterprise AI

Written by

Passionate about digital and AI technologies that advance how businesses engage with their customers and employees to create superior value for all.

Conversational AI Solutions

Insights for businesses on deploying conversational AI assistants to automate digital engagement and improve business results.