The Thread Episode 2: When it Comes to Customer Service... Were Chatbots Just a Fad?

November 11, 2019

When it comes to customer service, were chatbots just a fad? 

CGS recently conducted a consumer survey of 1,000 US consumers where they asked them to weigh in on their likes and dislikes around customer service interactions. Host Evan Aldo is joined by Richard R. Shapiro, who is the Founder and President of the Center for Client Retention and a leading authority in the area of customer satisfaction and loyalty.

For the full 2019 CGS Customer Service Chatbots and Channels Survey visit:

Hello and welcome to the Thread, a podcast about the latest tech and business trends in fashion, learning and outsourcing. In today's episode, I'm joined by Richard R Shapiro, who is the Founder and the President of the Center for Client Retention. He's a distinguished author and he's also a leading authority in the area of customer satisfaction and loyalty. Richard has conducted research with thousands of customers, from Fortune 100 and 500 companies for over 30 years. Richard, first of all, I want to thank you for being here today.


Thank you for inviting me. It’s a pleasure. I look forward to our conversation today.


So our topic today, as you know, is chat bots. So how about you talk about what a chat bot is exactly. And your experience with it.Are they good, bad or just a fad dying out? What's going on with them?


Sure Evan. You know, that's a good question. I've been in business thirty-one years and most of our clients are call centers. For years  companies that were clients of ours, always used frequently asked questions either on their web site or possibly through the IVR system. The purpose of that was to cut down on frequently having a human being answer questions that would be most commonly “Yes.” So a bot kind of replaces those frequently asked questions. Everybody, I think, likes to self-serve and that's probably their first choice. And companies like to cut costs. So it is a perfect marriage. But the bot that exists now, if the company keeps it updated and it works, it's great. If the company doesn't constantly update the questions and the answers, that would be relevant. I think it's an issue and that's where customers get frustrated.


So they're good as long as the companies put the money into it, that makes sense? Absolutely. So as you know, CGS recently conducted a survey of a thousand people, and surveyed over a thousand U.S. consumers, where they asked them to weigh in on the likes and dislikes around customer service interactions. Now, the two most popular things are chat and phone. Do you have any preference? Is chat better than phone, phone better than chat?


Well, when we talk about live chat, I think that's really the same as a phone. But when you talk about a bot, once again, that's an elementary process. But one of the things that I have been impressed with about bots is that when bots first came out, they were kind of like stoic or they just answer the question or say, did I help you? But now a lot of the bots automatically are programed to say things like, “good morning” or “good afternoon” or “thank you” or “I hope you had a good day.” And even though you know it's a bot you feel good about. It just sometimes puts a smile on your face, even though it's a robotic message.

So, you know, more relatable.


More personalized, conversational bots I think are good. And I think that makes sense.


Absolutely. So in the survey we found this year, 29 percent of people said they would choose chat over a phone. But what was very interesting at least to me and a lot of other people was that the year before 2018, around 50 percent, they said they would choose chat over phone. So it's going down. Less people are choosing the chat. And that's surprising since as technology increases, certainly more people would choose to chat. Do you have any ideas why that can be?


Sure. Well, I'm an analytical person. So what that tells me is that people do like chat bots. But probably over the last year, maybe the companies that they're using haven't really updated the messaging or updated the responses. So I know there's one company that I have a relationship with. It's a voice recognition system. And they claim that 99 percent of all the conversations are accurate. Then I said to them all, how can that possibly be? Because Alexa isartificial intelligence and half the time she doesn't know what city I'm asking for or what song I'm playing. And they explained to me that the way they develop their system, which should be the same way that bots are a refined, is that they have a human analyst who interacts or oversees the interaction between the customer and the and the bot or the customer and the speech recognition system. So that let's say, one of their clients is a hotel. The first time somebody was asked how many people are in your party and the person didn't say two. But instead it said, well, it's my wife and me. A human analyst said, well, that's two people. And then responded. So, you have two people in your party. So these are the types of things you need to train the bot and keep it updated. And if you do, I think that people will be willing to go back to using the bot and it won't be frustrating. And I think that statistic will probably start going back where people are comfortable in using a bot versus the phone.


So companies they got to stay updated on it. That’s why. Because companies are getting a little bit lazy.


I mean. It is also possible a lot of people are on their mobile devices, and I think when you're on your mobile device, you know, checking out a Web site and if the telephone number appears, if the customer or the company actually highlights that number and you can just click on it, and if you think you're going to get a rep right away, that might be another reason as well. The visibility of the telephone number. If their expectations are that they might quickly get a person on the phone.

Fast, instant gratification. Absolutely. So I want to ask you a question. Very interesting question. We asked 1,000 people if you had to pick one word to describe artificial intelligence when it comes to customer service. What word would you chose?


Frustrating or frustrated? Listen, my wife is constantly yelling at Alexa. We do have a place in Florida where we feel very fortunate as well as in New York. And we could be asking what the weather is in Florida. And all of a sudden it will give us, let's say Fort Lauderdale weather or we didn't ask Fort Lauderdale. So here is where a case, artificial intelligence is kind of using our history, but their assumptions are incorrect.


So, you're not far off from what most of the people said, the most popular choice was maddening, 13 percent in every single other choice was pretty much negative with one kind of neutral one being futuristic.


I'm sure my wife would say maddening.


So what do you think could be done to fix that, to make people more positive towards that? You just mentioned earlier about a better statistic with the ninety nine percent of accuracy. Do you think if it gets to that point, more people will be?


Well, looking at the forest from the trees, there's so many choices out there for consumers. So overall, when customers get frustrated, you're really telling them you don't want to help them. I mean, and these days, with so many options and customers not being as loyal as they used to be, if you have a service where you have a lot of competitors oryou have a product where you have a lot of competitors and your customers get frustrated, they're going to start to look for your competitors. So I think you have to really make that the end game. Do I want to keep my customers? Do I want to keep them happy? Do I want to get good reviews? Do I want to get good testimonials? And to me, any phone call interaction is really an opportunity with that company to build a relationship - a stronger relationship. So I would make sure, whatever the process is, that it's well-defined, but in all cases make it easy for that consumer or that customer to get to a live agent if that's the only way the problem can be solved.


That makes sense to me. Makes a lot less sense to me. So another stat that was interesting to me in the survey was 71 percent of people would be less likely to choose a brand if they knew that they didn't have any human agents available. I know a lot of friends who just don't buy stuff when they know there's no one to talk to. You know, no actual person to talk to. Why do you think that is?


I think it's you know, what I kind of refer to before… that if somebody thinks I think everybody has a pretty good feel for whether something can be fixed by artificial intelligence or human being. But a lot of times you don't know. But if you do know a company doesn't have any kind of staff or you hear from your friends that they don't have, they don't allow you talk to a person. I think that in a lot of cases, it can help you make the decision to perhaps go to a competitor. But I know I had a case where the credit card company,by accident, I paid the wrong credit card. I had two credit cards with this company. I knew analytically that there was no way that I could transfer one money, the amount of money that I had a credit on to the other, to the other charge. And I had to call. Yet when I called the company. I had a recording saying, all of our agents are busy. The lines are long. Please, self-serve. And I knew what was probably going to happen. And I pressed zero. And immediately I got to a rep.


So in complicated situations, There's always, for your side of the credit card, something like that. You always need to talk to an actual person. The AI hasn't gotten that good yet to be able to do that completely on your own. What I want to add, you have any ideas for when you know, when we'll get to the point where AI could do like all of this stuff, the 99 percent you're talking to. It's a matter of years.


I think it's probably a matter of years because one of the companies that has had a terrible track record for a AI. Boeing 737 Max. I mean, basically, Boeing, through artificial intelligence decided that should override the human interaction and human thought. So I think we're far away from A.I. taking over, which I think is a good thing. I think that the human brain still can be empathetic and sympathetic as well to be knowledgeable about how to answer the questions.


That's interesting. That's good to hear. I don't want to AI to take over everything. So what I want to ask you now is about the human bot hybrid solutions. The customer starts with the bot, then they get to a chat person, if the bot can't fix the problem, then if the chat person can't fix the problem, then they go to an actual person on the phone. Do you think this is the best solution?


I think it's a good solution. But I think everything has to work. I did a mystery shopper for a department store where I looked up the most expensive engagement ring and it was thirty nine thousand dollars. And I said that I want to order six engagement rings. All of a sudden a bot came up and said, can I help you? And I wrote, I want to buy six of these engagement rings, can I get a discount? And then it was a human chat agent who did go to the next level and said, oh, you want to buy six engagement rings? And I said, yes. Now, they didn't say why. Which probably would have been inappropriate. And then the chat agent who is live said, OK. I can't help you, but call this telephone number and you'll get the fine jewelry department. And when I called that number, which I figured was going to happen. I got just the general recording for press one for parking, press 2 for hours. And this sale, potentially even though it was made up could have been one hundred eighty nine thousand dollars worth of merchandise. So too many companies, even if they refer the customer to a telephone number, that telephone number should go directly to whatever the department is. So bot, chat agent and then directly to a department or a person that can help you. Either automatically or giving someone the direct number.


When you have the phone at the end, you have to actually have a person there. So when we look at the age demographics from the survey, we found a few things that were really surprising. 25 to 44 year old respondents are two and a half times more likely to use chat or messaging over the phone, which they're younger. So you wouldn't find that too surprising. But when you compare that to 18 to 24 year olds, they prefer phone more than the people who are twenty five to forty four and even more than people who are over 65. Any ideas why that’s possible?


The only thing I can think of is something I referenced earlier that I think a lot of younger people are on the phone constantly where once again, if the telephone number is there and they think they can get a hold of a person, they're calling that number. That's the only thing that they are at a phone more than their laptop or some other device. So that's the only thing I can think of.


I couldn't think of anything. I mean, I kind of like to use the phone over other methods.


Unless they'd rather talk to strangers over their parents.


Our last part of the show, we always like to bring down to Earth just a little bit. So I want to talk about customer service excellence. The golden standard. Yes. I want to know what an example of a time where you were just so impressed with customer service, the most impressed you've ever been and what made it that way?


Sure Evan, I appreciate that. So recently I had a problem with my printer where it wouldn’t print black ink. Actually, I had the same printer that I basically threw out because I couldn't figure out how to fix it. So now I bought a brand new printer and the same basic problem happened. So I went onto the site and I did find that problem listed. And all of a sudden I did go into a bot situation. I wrote as my response, I am ready to throw this printer out the window. And two seconds later, a live chat agent came on and spent about fifteen minutes with me. Going back and forth. Didn't matter if it was phone or chat. It's the same kind of services so far as I'm concerned with. So friendly, so competent. But what totally impressed me is two days afterwards he actually called me to make sure that it was still working. So I think, you know, the lesson of that is not only fix the problem, but follow up with the problem to make sure the customer is still happy.


The follow up that shows people care definitely 100 percent. Richard, I want to thank you so much for your insight. I've learned a lot about customer service from you. It's been a great pleasure interviewing you. To everyone tuning in thank you so much for joining us. By the way, the consumer survey report that we've been talking to can be found in the YouTube video description below or searching 2019 CGS Customer Service Chat Bots and Channels Survey. That's it for episode two of the thread. Be sure to subscribe to this podcast on YouTube and your preferred podcast source. We're exclusively sponsored by CGS: an applications learning and outsourcing company that supports clients most fundamental business activities. For more information, visit I'm your host Evan Aldo: let's stay connected.