Amy asks, “How can we use AI to improve customer service? What are some ways to practically implement it?”
Customer service is everything from the purchase onwards in the customer journey, and companies are faced with dual mandates: make service better, but make service cheaper. What are some practical applications of AI? The keys will be around the two fundamental types of problems AI solves best: understanding and predicting.
– What customers are saying to you, especially in non-text channels (but also text, like mining your CRM)
– What customers to each other, from social conversation to search intent
– What the profile of our MVCs is
From our understanding, we will predict:
– What customers will be high need vs. low need, and positive vs. negative
– What customers will have the highest lifetime value
– When customer needs will be highest
Can’t see anything? Watch it on YouTube here.
Listen to the audio here:
- Got a question for You Ask, I’ll Answer? Submit it here!
- Subscribe to my weekly newsletter for more useful marketing tips.
- Find older episodes of You Ask, I Answer on my YouTube channel.
- Need help with your company’s data and analytics? Let me know!
- Join my free Slack group for marketers interested in analytics!
What follows is an AI-generated transcript. The transcript may contain errors and is not a substitute for watching the video.
In today’s episode, Amy asks, How can we use AI to improve customer service? What are some ways to practically implemented customer service? Let’s start by defining that customer service is everything from the purchase moment onwards. So from the moment the purchase, the customer gives their credit card says, this is the thing that I want b2b, b2c doesn’t matter. And everything after that in the customer journey is customer service. So that means ownership, which is things like warranty and service and support to loyalty, which is and retention, which is all about making sure that the your your product or service still continues to deliver value to the customer. And ideally, ideally, value continues to increase longer the customer own something. And of course, evangelism, helping the customer if they’re happy to talk about owning the product or service and looking for people who are having those conversations already. So those are the things that we’re going to lump under customer service. Companies traditionally have had dual mandates when it comes to customer service and not necessarily good ones, the dual mandates are of course makes service better. But in a lot of cases, companies have been tasked with making service cheaper, because corporations see it as a cost center rather than a value center. Now as models of attribution analysis get better. We’re starting to see people thinking about customer service as a driver of up sells, for example, particularly in b2b if a customer is happy, they’ll buy more from you are FM analysis and techniques like that help with that. But there are still a lot of companies are stuck in the old mindset of sort of customer services, just a cost, keep the cost is low as possible, instead of making the customer deliriously happy. And they’ll buy more and recommend more. So what are some practical applications of artificial intelligence in this space? The key to the application of AI and machine learning and specific revolves around the two fundamental tasks that they would be tackling which is understanding and predicting to understand we would use machine learning to dig into things like what our customers saying about us, especially in in social conversations, and then search intent. For example, if I am an owner of a whirlpool washing machine, and I’m looking up a whirlpool warranty or Whirlpool, whatever error code, guess what, I have some service issues, right, I have some support issues that I need to have resolved. If I own CRM software, I’m looking for the for how to do a certain thing. Sir, type that into Google, I would want as if I was the company, I would want to be using machine learning to analyze the data and understand relationships between different types of searches, relationships between searches and social conversations, at what point does somebody stop searching and start complaining? Those are all things that I wouldn’t want to know. I would want to know what are the profile of our most valuable customer is what using techniques like clustering, categorization, dimension reduction, what are those, the the aspects or the variables of a most valuable customer, and that goes way beyond traditional RF m analysis. Traditional our FM analysis just looks at you know, recency of purchase frequency of purchase monetary value of purchases, and that’s okay. But if you have the ability, with machine learning to look at 100, 200, 300 dimensions of the customer, their demographics, the psycho graphics firma graphics, and you can then put together a real comprehensive picture that customer you should. And most of all, this is where machine learning is a shining beacon customer service, it is understanding what customers are saying to our companies, especially in non tech channels, I did a project recently, for a company where they said a 17,000 recorded calls in an audio format from their call center. And we had a I transcribed them and then had AI, digest that down to understand what are the key issues that these customers were calling about. But you don’t have to use advanced stuff like that. Even just mining, the data within your CRM is so valuable, and companies don’t do it. What’s in your customer service inbox, customers, companies just don’t look at that. And you can use AI to understand
once you understand, then you can predict things the kinds of things you want to predict, for example, would be what customers are high need versus low need. If you have to reduce costs and customer service, you definitely want to know hire low need and who is high and low value, right? So what are the types of if a customer is high need and low value, and you have the ability to predict that type of customer, you can say let’s not market to those people versus low need high value let’s market to those people as much as we can use machine learning to isolate, understand and build a model for predicting that based on
characteristics of data that you get, you can predict what customers have the highest lifetime value.
Again, if you go beyond our FM and your understanding, and then build a models who predict this customer matches 42 of the 46 high value indicators, let’s make sure that, you know we we target them really, really cleverly and smartly so that we win their business. And finally, now what are the what are the types of sentiment in conversations? And can you create a model of cases that were resolved successfully versus cases is that we’re not resolved successfully? If there are things like sentiment and tone in conversations that people are having with the chat bot with the operator with the customer service forms? Can you predict when something’s going to go off the rails and intercept that early on, so that you that that person never gets to the stage where they cause a PR incident for example. So those are understanding predict the two basic use cases for machine learning and AI that will help customer service get better. You can do this with the data you have today. What you have may be messy, it may need to be structured and cleaned up and engineered but the beauty is most companies have been collecting this data. It’s in your customer service inbox is in your CRM, it’s in your call center. You just haven’t been using it if you’d like help course I company trust insights will be happy to help with that. So great question. important question, important question and one we’re not talking about enough. As always if you have any follow on comments, please leave in the comments below. Please subscribe to the YouTube channel and the newsletter. I’ll talk to you soon. want help solving your company’s data analytics and digital marketing problems. Visit trust insights.ai today and let us know how we can help you
You might also enjoy:
- Almost Timely News, 17 October 2021: Content Creation Hacks, Vanity Metrics, NFTs
- Best Practices for Public Speaking Pages
- How to Set Your Public Speaking Fee
- What Is The Difference Between Analysis and Insight?
- Retiring Old Email Marketing Strategies
Want to read more like this from Christopher Penn? Get updates here:
Get your copy of AI For Marketers