The Future of Conversational Intelligence in Customer Service and AI

The Future of Conversational Intelligence in Customer Service and AI

This blog aims to give the reader a window into the future of Conversational Intelligence software. To learn about the future, we have to visit the past and define some terms such as what conversational intelligence is, why it is important, and use cases for the future.

What is Conversational Intelligence?

Conversation Intelligence refers to a category of software products that ingest inbound or outbound digital customer interactions and generate business insights for organizations. This includes phone calls, emails, chat messages, and social media interactions.

How long has Conversational Intelligence been around?

The first commercial software providers in this category, some of which are still in business today, were founded back in 2005. These providers had to use early natural language processing (NLP) and machine learning algorithms and deployed hardware-based solutions in the contact center to extract insights from phone calls.

What’s changed in the last twenty years?

Besides my waist size and hairline J, a lot has changed in the last twenty years. As it relates to Conversational Intelligence, it is mobile, social, the cloud (AWS/Azure/Google), and large language models (LLMs) for artificial intelligence (AI).

The mobile era brought new customer expectations on how they interact with the enterprises and organizations (i.e. colleges/universities, local government, and state government) in their life. If I can download an app in thirty seconds and it works perfectly in the device in my pocket, why can’t everything give me that same responsiveness? There are large generations of consumers that text more than they make phone calls, so if your organization can’t communicate thru SMS messaging, you’re at a strategic disadvantage in this generation.

Social networks connected families, friends, and strangers in new ways and became the dominant communication medium over the last decade. If I’ve had a poor customer experience that hasn’t been resolved, I’ll tweet at the CEO of the company my grievance and in almost all cases I’ll be on the phone with an executive at that company with a resolution in 24 hours. While that doesn’t look great on the leader of their contact center, it’s a reality that must be integrated into your customer listening programs.

The cloud is the most important equity and inclusion invention of the 21^st century. If you build your applications natively in the cloud, you can reduce the cost of development and support by over 70%. Why would anyone pay extra to have an on-prem conversational intelligence platform?

Last, but very much not least, we have the invention of AI LLMs. Starting two years ago with mind-blowing demos from Open AI with their Chat GPT model 3, the world awoke to a new era of computing. The cloud vendors spent a combined $60B in capex last quarter, most of which went to buy Nvidia chips. If your Conversational Intelligence platform is not designed to take advantage of this capex spending, well you’re already falling behind.

The next question you should be asking is Conversational Intelligence really important?

The simple answer is yes. It costs a lot of money to acquire a new customer. If you keep that customer, the profits made on customer lifetime value (CLV) deliver a solid ROI on the acquisition cost. However, if you don’t keep that customer due to a poor customer experience, your customer acquisition costs will eventually kill your business.

A study released a month ago by Khoros shared that the number one determinant of brand loyalty is a positive customer experience. “If you want to build brand loyalty, not lose it, ramping up your brand’s customer service efforts is the best place to start. Customers overwhelmingly agreed (83%) that they feel more loyal to brands that respond to and resolve their complaints.”

Is Conversational Intelligence really important? Absolutely. If you don’t know what’s upsetting your customers, if you don’t know what they are saying about you to their friends and families, how can you provide them with the best customer experience possible? This isn’t just a commercial problem. Student enrollment suffers from negative customer experiences, and cities and towns are losing their tax bases as people move to cities and counties that do a better job of taking care of their constituents.

What are future use cases for Conversational Intelligence?

 The analysts will tell you the major feature sets are conversational analytics, agent scoring, agent coaching, compliance, and trend analysis. They are correct, and if this is all you are looking for, then pick a cloud-native solution with the best pricing, preferably consumption-based and not agent-based.

We started SparkCX to build new innovations, to reimagine what great customer experiences could be, and deliver technology that works so organizations can focus on their core business.

Here are some use cases we’re excited about:

Agent Gamification – Once you put a conversational intelligence solution in place, you should see most of the benefit in the first six months assuming you’re acting on the insights you’re gaining. While some of the benefits of those changes accrue to agents, we’re interested in improving the agent experience and making happy agents, because happy agents correlate much more with happy customers. We are working with early customers to reward agents for delivering the best outcomes, helping their peers, and contributing to a more positive workplace.

Generative AI for the business user – Many Conversational Intelligence implementations are designed to improve the contact center they are deployed in, but they don’t harvest business insights to help improve marketing, sales, product development, etc. We are tuning our generative AI models to uncover business insights and proactively share them with business decision-makers.

Hyper-vertical use cases – Can you imagine a world where there are only 4-5 large language models to rule them all, or do you foresee a world where different models are trained on different data sets to achieve specific outcomes? We imagine the latter. While we’re excited about the GenAI LLM we leverage today, we imagine a world where smaller more performant and cost effective models may do a better job for a hyper-vertical use case. For example, analyzing a doctors recorded patient diagnosis requires different level of understanding than a consumer complaining about a coffee maker that stopped working.

About SparkCX

 SparkCX's mission is to deliver customer experience solutions enabling people to forget about technology and invest more time taking care of each other.  The founders are passionate innovators who want to make a real and sustained difference for their customers.