Hello fellow CX practitioners,

I hope your customers are happy and your loops are closed (IYKYK) …

With all of the craze around chatGPT & autoGPT, I started to think about the ways that this technology can be used in customer service & CX.

Today’s newsletter aims to answer one question for you all:

What practical applications will conversational AI have on #CX and customer service over the next 10 years ..?

Check them out below 👇

10 Practical Ways #ConversationalAI Will Impact Customer Experience 🔮

Disclaimer: These are my thoughts only, based off what I have read & researched. I hope they can help you in generating ideas for your own business.

#1 - Customer Service Written Responses

A recent study by NNg found that ChatGPT users were able to write 59% more documents in a working day than people who do not use ChatGPT.

When you apply this principle to customer service, it’s evident how conversational AI would enable customer service agents to a) respond to customers in a quicker manner b) likely provide a more empathetic and well curated response. These efficiencies could be observed in any written response - emails, webchats, social media, product reviews etc.

Expected Business Impact: Lower cost to serve; happier customers.

#2 - Intelligent FAQs

Think about the most recent FAQs that you had a look at. I bet that they might have “kinda” answered the question, but not exactly. This is because customers often have difference nuances which can’t always be answered by a static FAQ.

I expect conversational AI to change this. As we’ve seen with ChatGPT, conversational AI leverages natural language processing (NLP) which makes it intelligent enough to modify answers in accordance to context or different situations.

Intelligent FAQs will go beyond a simple keyword-based FAQ, leveraging conversational AI to provide a well-thought answer leveraging a combination of all of the information that has been feed into the tool.

You can read more about intelligent FAQs here.

Expected Business Impact: Reduced contact centre volumes.

#3 - Ask Me Anything

Conversational AI is going to be able to turn any type of data into a simple output.

If you want it to look through all of your data and tell you the amount of times that a customer has said “f**k” in a survey, it will be able to do so.

If you want to figure out how many surveys your platform sent in the last 5 years, it will be able to tell you in plain English.

You know how Google is able to spit out different links when you ask it a question? Conversational AI will be able to spit out precise answers when you ask it something. It will trawl through all of the data that you share with it and will leverage this information to give you a plain-english answer.

Expected Business Impact: Quicker speed to insight; better decision-making.

#4 - Text Analytics & Sentiment Analysis

Text analytics & sentiment analysis has been around for a long time HOWEVER conversational AI has changed the level of access that we all have to this technology.

This point is better articulated in the form of a screenshot. Within 10 seconds, I was able to produce a sentiment category, NPS comment & text analytics topic associated to the customer comment 🤯.

Now the haters will question the accuracy. Putting that to one side for a moment, this technology has made text & sentiment analytics accessible to the general population for little to no cost. Pretty crazy, right?

Expected Business Impact: Greater speed to insight. Automated scoring for 100% of your unstructured interactions.

#5 - Predictive Analytics

Once again, predictive analytics has been around for a long time however it has not been accessible to a non-techie.

I expect this to change, and let me explain why.

Conversational AI removes the need for us to do any programming. We can simply tell the machine what we need in plain English.

Imagine this scenario. I have 10,000 survey responses which all have an NPS score and an associated customer comment. I can upload this file to a conversational AI tool and give it the following prompt:

• You are now NPS-GPT. I want you to use these 10,000 survey responses and associated NPS scores to predict future customer comments that I provide to you.

  • Once trained, you can upload customer comments who don’t have any NPS scores, and have the AI tool predict the specified outputs.

Now, I am probably over-simplifying the complexity of what needs to happen in the above situation, but the real point that I want to make is that it has suddenly become exponentially easier for non data-scientists to get a steer on predicted outputs. As the technology evolves further (and you use more data to train it), the accuracy will continue to improve.

Expected Business Impact: Ability to predict & mitigate churn before it happens.

#6 - Fraud & Confirmation Bias Removal

Conversational AI can be used to detect and prevent fraud by analysing customer behaviours and patterns, at scale.

It can leverage previous interactions from other customers to proactively flag language or patterns which have previously produced fraud or dishonesty.

A practical use case for this could be in Sales. Quite often, Salespeople are known to have “happy ears” - they believe that the sale is going well when the facts are that it isn’t.

Conversational AI could be used to analysis back and forth communications between customer and salesperson to determine whether the customer behaviours and language being used can be associated positively or negatively towards a sale.

Expected Business Impact: Early detection of risk factors; removal of confirmation bias.

#7 - Email Tagging

This application is similar to point #4 but I thought it warranted it’s own section.

With AI email tagging, humans can save the time required to read every customer email by having AI-powered programs scan emails, tag them, and direct them to the right office. This would help the service reps save time and focus on the more complex tasks that cannot be completed without human intervention.

Expected Business Impact: Improved productivity and workflow management.

#8 - Productive Hold Times

According to this article, IBM have data which states that 52% of customers hang up on the customer support line, as they do not want to wait for a customer executive to address their issue.

Now whilst conversational AI probably can’t solve long queue times, it can definitely assist with taking a message for customers.

Imagine this scenario - after XX minutes waiting on hold, a conversational chatbot asks if you would like to leave a voicemail explaining your situation. This voicebot automatically transcribes the conversation and creates a ticket within your CRM for future follow up.

In addition to this, the customer receives an SMS summarising their problem and is able to converse with the chatbot providing further details until an agent can pick up the issue. In some scenarios, conversational AI will be able to resolve the issue before any agent intervention is required.

Worth a shot, right?

Expected Business Impact: Improved speed to resolution.

#9 - Augmented Messaging

While chatbots are great at troubleshooting smaller issues, most aren't ready to tackle complex or sensitive cases.

This is where augmented messaging comes into play. This AI tool identifies opportunities where human agents should step in and help the customer for added personalization.

Here's an example of how this interaction could play out":

Expected Business Impact: Improved chatbot effectiveness; lower cost to serve.

#10 - Quick & Easy Translations

Last but not least, conversational AI is going to make it extremely quick to translate something from one language to another.

In fact, we might not even need to translate the comment or text. We can just ask conversational AI to explain what this meant, but in our desired language.

E.g. ChatGPT, read this French text and explain (in English) what they were talking about. The below screenshot actually BLEW MY MIND 🤯 - I asked Google to produce a randomly generated set of words in French. ChatGPT automatically recognised this:

Expected Business Impact: Low cost customer service becomes easily accessible in non-English speaking countries.

Follow me on LinkedIn 👇