Real-Time Conversational AI: The Next Frontier in Intelligent Customer Experience

Written by
Andy Pandharikar
April 19, 2022

As the world becomes more digital, conversational AI is increasingly being used by businesses to power customer interactions. Solutions such as messaging and live chat provide an efficient and cost-effective way to offer real-time support to customers. By automating simple tasks and providing instant access to information, conversational AI can help reduce the workload of customer service representatives and improve the overall customer experience. 

In addition to providing customer support, conversational AI tools are helping customer support teams further enhance interactions with customers. Analytic tools are providing recommendations and suggestions on how agents and representatives can improve customer satisfaction scores and brand experiences through coaching and training.

With all the benefits, conversational AI isn’t without challenges. Natural human conversation is dynamic and a constantly changing form of communication. From languages, dialects, and accents to slang and visual formats such as emojis and memes, there are a lot of influences on daily communication. Human conversations can also be fragmented, or require context to fully understand its meaning. Conversational AI systems need to keep up. 

Why real-time conversational AI? 

You've likely used tools like Intercom, Drift, or Salesforce to chat with a business. When your customers chat with a support agent, they are creating data in real-time, but this data is typically only used in post-contact analytics to improve future interactions. 

But what if that data could be used to improve the conversation while you’re having it? What if an AI agent could see what your customer is saying as they are typing, and respond according to the exact issues or feedback provided at that moment? 

And what if human agents had intelligence alerting and assisting them in real-time with a variety of pain points related to the conversation - product issues, competitive comparisons, pricing sensitivity, bundled offerings, and a host of other potential key moments? What if an agent were  able to provide the most optimal resolution or recommendations on the fly?         

That's real-time conversational AI, and where we need to go with conversational AI.

Unstructured data and real-time conversational AI 

We must remember that data is the key to understanding our customers. Decades ago, most of the data businesses collected was structured. They’d get a purchase order from a customer, and they’d file it away in their system. That was it. But over time, as we started getting more data — unstructured data — businesses began to realize that they needed to change the way they operated in order to make use of this new information.

The same is true for customer experience today. We’re collecting more and more unstructured data from customers in the form of conversations. As they continue to multiply across touchpoints and channels, companies are thinking about using unstructured data to inform their customer experience strategies. 

Conversational AI tools like messaging and live chat are a great place to start, as they are built with systems and integrations to directly collect, analyze and enhance conversations. The next leap for conversational AI is to take advantage of the real-time data that these and other tools are generating, to provide agents and customers alike with real-time support. By having the tools to understand our customers in the moment, we can aspire to provide them with the best possible experiences. Unstructured data is a type of customer knowledge we hope to tap to achieve this level of understanding. 

Examples of unstructured data for real-time conversational AI

Let's look at a few examples of how real-time unstructured data could be used to improve conversational AI experiences.

1. Contact center teams can use real-time unstructured data to personalize their conversations.

By using unstructured data, contact centers can better understand their customers in different contexts and personalize their conversations accordingly. For example, if a customer calls with an issue related to product quality, real-time conversational AI may suggest learning more about their home environment, and upon learning that the customer has pets, provides suggestions on how other customers with pets have prolonged the durability of the product, in addition to offering a replacement. This can help improve the customer experience and increase customer loyalty. 

2. Contact centers can use real-time unstructured data to learn about customer preferences.

Unstructured data can be used by contact centers to learn about customer preferences, which can then be used by other teams. For instance, a customer shares details on why they love a product. Real-time conversational AI may capture this information and then flag it as a piece of content for marketing, and suggest to the agent to find out more about the customer’s needs.  

This information can then help businesses personalize marketing messages and campaigns.

3. Contact centers can use real-time unstructured data to address customer issues.

Contact centers can use unstructured data to resolve customer issues more effectively. For instance, suppose you're a consumer electronics firm that recently released a new digital camera. After users start interacting with your chatbot, you notice that a lot of people are talking about difficulties deleting pictures. Real-time conversational AI could automatically insert this product issue into the conversation flow and send customers to a corresponding tutorial. 

This information can help companies provide better support and resolve customer problems more quickly and effectively. 

These are just a few hypothetical examples of what real-time conversational AI can do, and where technologies are going. But the role of unstructured data in providing rich customer experiences is certain. By taking advantage of real-time unstructured data, companies can improve customer service, inform other functions, and better understand their customers.

In Practice: Google Agent Assist

To see how unstructured data is used by conversational AI in the wild, we can look at the Google Agent Assist. Google Agent Assist is a conversation assistant platform that uses AI to provide customer support. 

The platform assists human agents in conversation and in chats to help serve customers more efficiently. It can transcribe calls in real-time, recommend ready-to-send responses, and provide answers from a knowledge base. By bringing FAQs and other company documentation into a knowledge base, the platform is tapping unstructured company data to help resolve conversations more quickly.  

Google Agent is just one example of how unstructured data can be used to improve customer support experiences. As conversational AI becomes more popular, we can expect to see more businesses using unstructured data in similar ways to improve their customer service efforts.

From reactive to proactive: rethinking customer experience

Traditionally, contact centers have been used for reactive communications. A customer has a problem and they call customer service. The customer service representative then tries to solve the problem.

What if we could shift to a proactive model, where we could anticipate and prevent problems before they even happen? This is where real-time conversational AI comes in. By using unstructured data, both chatbots and human agents can become more proactive in their interactions with customers and can help prevent problems from happening in the first place. 

Further, these insights can empower the product innovation process. As customer service and marketing teams get to know their customers better, they can feed that knowledge back into product development to create products and services that better meet customer needs.

Return to blog