For Better Customer Service, Offer Options, Not Apologies

SARAH GREEN CARMICHAEL: Welcome to the HBR IdeaCast, from Harvard Business Review. I’m Sarah Green Carmichael.

The internet has revolutionized customer service. You can take care of so many problems without ever talking to anybody.

It’s been great for consumers; it’s been great for companies. But it also means that the problems that do escalate to a phone call with a customer service representative are often the hardest to solve. Which is why it’s much more important for companies to get those interactions right.

Our guest today studies what separates a satisfactory interaction from an unsatisfactory one—and has new research that companies can learn from.

And I’m personally excited to talk to him because of my own recent, frustrating airline experience.

Jagdip Singh is a professor of marketing at the Weatherhead School of Management at Case Western Reserve University. His team’s research is featured in the article “‘Sorry’ Is Not Enough” in the January–February 2018 issue of Harvard Business Review.

Jagdip, thank you so much for talking to us.

JAGDIP SINGH: Oh, you’re welcome. I’m glad to be here.

SARAH GREEN CARMICHAEL: So, I thought I would start by telling you a recent experience I had and then getting you to react to it. Now, I will try to be brief, because I know these airline horror stories can just go on and on, and they’re probably only interesting to the person who experienced them.

JAGDIP SINGH: Sure.

SARAH GREEN CARMICHAEL: So, I recently was flying to London for a conference. And the day I was supposed to leave, I was just as sick as I have ever been. I had one of those terrible, 24-hour, like, violent throwing-up episodes. So, I went to the doctor. I got a note. I called the airline. I told them, I, like, can’t get on the plane. And they said, OK, that’s fine; call back when you’re ready to rebook, and we’ll wave the change fee. Great.

So, a couple of days later, I call back, and I’m trying to rebook. And I have to try to leave, you know, imminently, because now—I was supposed to get there a couple days before the conference to take some personal time. Now I’m just trying to get there for the conference.

So, I’m like, really stressed out, and I’m really, like, still recovering from this thing, but I just have to go. So, first, they’re like, you need another doctor’s note to say you’re fit to fly. OK. I go back to the doctor’s office. I get another note from the doctor. I call back to the airline. Of course, every time I call them, it takes 45 minutes to get through to a person. So, finally I talk to them again. They say, OK. That’s fine. We’ll wave the change fee, but to be on the next flight is, you know, thousands of dollars. And I’m just sort of stunned by this because of course, I thought getting all these notes would mean, you know, that there was some kind of medical waiver for stuff. So, I say, I have to call my manager, and I have to ask her about this; this is so expensive. Hang on. Call my manager; talk through it with her. We work something out.

So, after all of this, they do finally manage to get me on the plane. And as I am then on the phone with them, I am actually packing my bag while we’re sort of dealing with all of this. And I just ask them, I’m just curious to know what you think I should have done in this situation, because this is just, like, this whole experience has just been so bad, I’d like to know how to handle it differently should I ever be again stricken down by, like, a biblical level of throwing. And they’re like, we just suggest you make the flight. And I couldn’t imagine a good customer experience for anyone on the airplane to have me in that situation somehow on the airplane.

So, what stands out to you as someone who studies customer service?

JAGDIP SINGH: You know, these experiences are so common, and sometimes it just takes one individual to change the direction of these interactions. And yet, somehow, the airlines or the service provider is just not able to connect the dots. And, you know, the research we have done—this is just one piece of a series of work we have done—just show that problem solving remains a very difficult challenge.

SARAH GREEN CARMICHAEL: Yeah, so I would love to hear about the study that you and your team did where you analyzed these over 100 videos filmed at customer service desks at airports in the U.S. and the U.K.

JAGDIP SINGH: Our approach was to see if we can code the words that frontline and the customers use to express problems or to solve problems and to look at their non-verbal cues, meaning their facial expressions their body postures, to better understand what is being communicated during these interactions and what distinguishes a satisfactory episode from one that is not so satisfactory and that’s the basic data we used to come up with. The findings that we did find in our research.

SARAH GREEN CARMICHAEL: And what were those findings?

JAGDIP SINGH: The most interesting findings pertained to what we call solving work versus relational work. So, solving work is creatively, competently, and energetically generating solutions to the problem customers have. Whereas, relational work is showing warmth, showing empathy, and forging a personal connection with the customer. And what we were finding was that in these situations of problems where customers are working under time pressure, the solving work is the dominant factor in customer satisfaction. And No. 2, because customers only see what the service person is doing, they look at the words being used. And relational work—showing empathy or showing warmth or forging personal connections—the words used up in these kinds of work is actually perceived to be not very helpful because in their mind it distracts from the solving work that is essential for the customer to have good options.

So, that was the core of our finding. And in order to develop it further, we did a follow-up experiment in which we could add more detail to that idea. And what we found was that in the case of solving work what is important is not the ultimate decision, the choice the customer made or was given in that situation, but how many options, how many good options the customer had. So, this idea that just looking at the outcome can tell us how good that interaction was, was dispelled in our experiment because what we found was that the more creative and interesting options the frontline person was able to give to the customer is what determined to what extent the customer was satisfied with the exchange.

SARAH GREEN CARMICHAEL: I’d like to play a clip actually that you used in one of these experiments just so we can hear how relational work sounds.

SIMULATED FEMALE AIRLINE CUSTOMER SERVICE REP: Hi. May I help you?

SIMULATED FEMALE AIRLINE PASSENGER: I came on the 6 a.m. flight from New York via Atlanta, but my checked baggage is not here.

SIMULATED FEMALE AIRLINE CUSTOMER SERVICE REP: I am so sorry and am happy to help. May I have your boarding pass?

SIMULATED FEMALE AIRLINE PASSENGER: Sure.

SIMULATED FEMALE AIRLINE CUSTOMER SERVICE REP: I apologize for the inconvenience. Your baggage didn’t make the Atlanta flight and is not here. I am so sorry for—

SIMULATED FEMALE AIRLINE PASSENGER: That is unacceptable. I have a job interview at 1:00 p.m., and my baggage has all the materials.

SIMULATED FEMALE AIRLINE CUSTOMER SERVICE REP: I am so sorry for your troubles. I wish I could be more helpful in your situation.

SIMULATED FEMALE AIRLINE PASSENGER: This is so unfair.

SIMULATED FEMALE AIRLINE CUSTOMER SERVICE REP: I know how you feel. I was in a similar situation once. I understand how stressful it is for you.

SARAH GREEN CARMICHAEL Why use the robot voices?

JAGDIP SINGH: Ah. We wanted to use voices that were in some way as computerized so they would have exactly the same intonation and the differences between scenarios would not be traced to the specific voices used.

SARAH GREEN CARMICHAEL: So, tell me a little bit about why this relational approach of apologizing and empathizing doesn’t seem to work.

JAGDIP SINGH: Customers think that there’s got to be a creative way of handling problems. I mean, you lost your baggage; you have an interview—this is not an ordinary circumstance, much like yours: you’re sick, and you’re trying to be careful and respectful of people on the plane. And they think that the frontline employees can do better than just give them the standard, single option or spend a lot of time telling them things which are simply empathizing but not solving. So, I think empathizing by itself is not the problem. It is at the cost of solving. And because time is limited, and you are under time pressure, there’s a natural tradeoff between the two—relational work and solving work—that customers have salient in their minds; and when they see the agent using a lot of apology, a lot of sorry, they feel that time could have been spent much better to creatively, energetically, find some interesting solutions that will give me some choices in that matter. I think that’s what’s going on in the minds of the customer.

SARAH GREEN CARMICHAEL: Let’s take a moment now and just listen to a quick clip of what that problem solving sounds like.

SIMULATED FEMALE AIRLINE CUSTOMER SERVICE REP: Hi. How are you doing?

SIMULATED FEMALE AIRLINE PASSENGER: I came on the 6 a.m. flight from New York via Atlanta, but my checked baggage is not here.

SIMULATED FEMALE AIRLINE CUSTOMER SERVICE REP: Let me check this right away. May I have your boarding pass and baggage tag?

SIMULATED FEMALE AIRLINE PASSENGER: Sure.

SIMULATED FEMALE AIRLINE CUSTOMER SERVICE REP: I see that there was a weather-related delay. Your baggage did not make the Atlanta flight due to insufficient connection time.

SIMULATED FEMALE AIRLINE PASSENGER: That is unacceptable. I have a job interview at 1 p.m., and my baggage has all the materials.

SIMULATED FEMALE AIRLINE CUSTOMER SERVICE REP: Let me see what I can do to get it here for you as soon as possible.

SIMULATED FEMALE AIRLINE PASSENGER: This is so unfair.

SIMULATED FEMALE AIRLINE CUSTOMER SERVICE REP: I understand. Let me see how to get your bag here at the earliest.

SARAH GREEN CARMICHAEL: OK. So, it actually, goes on a bit more from there, but in the kind of same vein. You can clearly see what the differences there between the two styles. Why isn’t it possible to both be empathetic and apologetic and problem-solve at the same time? Why does it have to be a tradeoff?

JAGDIP SINGH: The tradeoff has two elements in it. One is just the time. There’s a limit to time. These interactions don’t last very long. The highest was about 8 minutes or 10 minutes. So, within the limited time, the words is a fixed quantity; and solving work, solving words, are different than the words that I use to communicate empathy and relational words. What we’re finding is that the customers look at the agent and hear the agent and make some inference about how hard they are working to solve the problem. So, if they use a lot of empathy words but actually are focusing on solving, that’s feasible. But the customers do not see that, and they think that if they’re using a lot of relational words, it must be not paying attention to the solving work. So, that’s one area of tradeoff that customers have in their mind.

The second is that is a whole body of work in psychology which shows that competence and warmth, the two dimensions of behavioral dispositions among humans, they are negatively correlated in general. So, in other words, competence, which is like solving work, warmth, which is like relational words, tend to be not positively correlated in general. So, the notion being that these two qualities are hard to do and simultaneously in any setting; they are not easy to multitask. So, that also plays into this thinking. In our own work, what we’re finding is that there’s actually a lot of truth of the psychological research in customer service as well. The people who are very good solvers tend to not be overly empathetic or engage in relational work. They keep focused on the careful cognitive effort that’s needed to look for these options.

And that truly is the dilemma of customer service work: when there are no problems, empathy or relational work are very important; but the moment a problem arises, the frontline agent has to shift to a different style of operating, which is a problem-solving approach where solving work has to dominate and relational work has to be kept to the minimum. And I think customers that are OK to saying in the beginning, I am sorry for what happened; let me see what we can do to solve. After that point, customers expect you to be spending all the time you have on solving and giving them interesting options.

SARAH GREEN CARMICHAEL: So, I think it’s important to note though that although we call this the solving attitude, in your second scenario, the problem solving does not actually solve the problem. The bags end up getting there after the woman’s job interview. You know, in both cases the outcome is the same. How come the person going through this experience would be more satisfied with the problem-solving approach when the problem’s not solved.

JAGDIP SINGH: That’s right. As you can see, this experiment was designed to achieve the same outcome, and we wanted to see does it really make a difference in the mind of a customer who looks at these scenarios, looks at the options which are given, and the outcomes, and are we able to figure out differences in customer satisfaction. And the answer is yes. It’s not the outcome. It is the creative generation of plausible, useful options. So, in the second scenario, the remaining part of that interaction is dedicated to agent giving the customers multiple options.

SIMULATED FEMALE AIRLINE CUSTOMER SERVICE REP: I have a few options. I can have your bag on the next direct flight at 2:25 p.m. and delivered by 5:30 p.m.

SIMULATED FEMALE AIRLINE PASSENGER: That won’t work. I need my bag before my 1 p.m. interview.

SIMULATED FEMALE AIRLINE CUSTOMER SERVICE REP: OK. Once the bag arrives, I can expedite delivery for a $25 fee which I will waive. But you still won’t get the bag till 3:30 p.m. If that doesn’t work, I have some other options.

SIMULATED FEMALE AIRLINE PASSENGER: Why can you not get my bag on an earlier flight?

JAGDIP SINGH: And when the customer objects that that won’t work, the agent provides another option.

SIMULATED FEMALE AIRLINE CUSTOMER SERVICE REP: Yes, an earlier option I have is an Atlanta –Houston connection. That will get your bag in Miami by 1:47 p.m. If I expedite, you will have it by 2:30 p.m.

SIMULATED FEMALE AIRLINE PASSENGER: That does not help. I have my interview at 1 p.m.

JAGDIP SINGH: The agent is constantly working with the customer to provide options that might be useful to the customer. The customer makes the decision.

SIMULATED FEMALE AIRLINE PASSENGER: I guess I do not have much choice. The 2:25 p.m. direct flight will have to do.

JAGDIP SINGH: That ultimately, maybe, a direct flight is better because baggage can get lost when they go through multiple points of contact. So, maybe it’s not going to make it anyway; I’ll figure out another way. But the customer leaves satisfied because at least they had a choice to exercise. They had multiple options. They came away thinking that the agent did do the work needed to give them these solutions to the problem that they had.

SARAH GREEN CARMICHAEL: This conclusion—that it’s not about the solution per se; it’s about how you get there; it’s about offering options to the customer— this is a slightly different conclusion than some other research previously had published which was in the article “Kick-Ass Customer Service.” In that research, those authors identified seven different customer representative types and found that the most effective in their study was one they called “the controller,” where you’re actually directing the customer towards one optimal solution and sort of taking control of the interaction. But this approach you’ve been describing here is very different. Why do you think there is that disparity?

JAGDIP SINGH: I think they are looking at personalities. We’re looking at behaviors. Their approach of controller is more top down, where the agent is competent and is more effective in directing the customer. We are more thinking in terms of bottom up, where the agent is working with the customer, being tuned to their specific situation their concern, and working with them to come up with solutions, and then letting them make that decision. We also differ a little bit in the sense of controller is one type of personality, whereas we believe in more flexible behavioral approach, where the agents have to express some empathy and relational work. We’re not saying don’t do it at all. We’re just saying keep it limited to the first few seconds of the exchange, and then move on to solving work. So, the shifting behavioral repertoire is what we think is most effective in these settings.

But I can tell you from my personal experience. Last night, I was to fly from Reno to Cleveland. And I arrive at the airport, and I’m told that my flight is going to be delayed by three hours due to weather and technical issues in Houston. And Miles missed my connection to come to Cleveland and I’ll have to stay the night in Houston. And I said, Well, now Sarah won’t be happy because I have to do this interview with her this morning. So, it turned out there was an agent at the front desk also named Sara.

SARAH GREEN CARMICHAEL: Ah!

JAGDIP SINGH: How interesting! Right? Now, she had the eye on the ball. She knew that I needed to be here in the morning, and she got to work right away. She hardly took any time to say I’m sorry or empathize. She said, things didn’t work out with the plane; let me see what I can do.

But she would not talk to me. She was just making a lot of facial expressions and working on the keyboard. I know she was working a solution for me, but I wanted to know what exactly is she thinking. Is she working for choices that will work for me or not? So, knowing my research, I began to engage her in discussion: Are we going through Chicago? Have you looked at that? Maybe that’ll work. Can we go through Denver? Now, she had a good idea. She said Pittsburgh is going to be a good one for you. Then you drive in the night, be back home by around 1:30, 2 a.m. in the morning, and that’s the best I can do if you don’t want to stay the night.

SARAH GREEN CARMICHAEL: Wow.

JAGDIP SINGH: And I didn’t know that she’s thinking of Pittsburgh. I just thought she was going to get me to Cleveland. So, I said, no; try Houston, Chicago. And she gave me all options, and that’s the interesting part. She gave me the option through Chicago. She created that; she made a car. She got an option to Houston and told me what that would be. At the end of the day, I said let’s go back to Pittsburgh; maybe that’s the best, because once I get to Chicago and Houston, I have no connection again, so that doesn’t help me. And we landed in the same solution that I had initially said not good enough.

So, I think that again is the power of the frontline agents giving options to customers, even though they may choose an outcome that on the face of it doesn’t seem to be the most effective. And it turned out that I didn’t go to Pittsburgh. Now, that’s a whole different story.

SARAH GREEN CARMICHAEL: Oh, no! Wait. What?

JAGDIP SINGH: So, I land in Denver to catch my plane to go to Pittsburgh as I was scheduled to go. And as I’m going from one gate to the other I pass a gate, gate 41, where there is a flight leaving for Cleveland. Well, it should not be. It should have left two hours ago, but it’s delayed. And it’s sitting there, and I said, Hey, maybe I can get on that flight. And I say, Can I get on this flight? And yes, I could.

SARAH GREEN CARMICHAEL: Amazing.

JAGDIP SINGH: Amazing, don’t you think.

SARAH GREEN CARMICHAEL: You have gone to heroic lengths to be on this podcast. So, thank you.

JAGDIP SINGH: So, I think that shows that solutions are so important. And I don’t want to leave a sense that it’s all agents. I think agents are a piece of a puzzle. Today many companies are using artificial intelligence using analytics to improve the efficiency of their operations. I mean look at all the airlines: they’re using these data engines and AI to fill every seat they can by using real-time and dynamic analysis. The agents also need information to solve customers’ problems effectively. So, if somebody in the chain of the frontline work had looked at my record and figured out that I will be arriving in Denver in time for this flight that will be leaving to Cleveland and it has seats on it, I think I would have been most delighted passenger if when I landed they gave me another boarding pass and said, Guess what? Our analytical and artificial intelligence analysis showed that you could catch this plane to Cleveland, and here a boarding pass for you. The organizations have that kind of data today, but their deployment to give them to the frontline agents to make them smarter and better in solving problems, that’s the chain that’s broken.

SARAH GREEN CARMICHAEL: This sort of organizational problem, is that getting the data to the right people the most pressing thing? Is it hiring different kinds of customer service people? Is it revamping organizational policies and practices? What could organizations do to really make the biggest difference?

JAGDIP SINGH: I think there are three main differences organizations need to make if they really need to mobilize these frontline interactions to build customer loyalty. No. 1, make these agents more effective in the work they need to do. And effectiveness of agents is not just personality. It’s also training; it’s also culture. It’s also giving them autonomy to be empowered, to listen to the customer and to respond with effective solutions, not a scripted solution.

No. 2 is to provide the agents with the resources and data and inputs that are needed to make good decisions. So, an agent sitting—Sara, sitting in Reno— cannot access the big body of database that United has to figure out when this flight reached Denver that’s heading to Cleveland. But the data warehousing, the analytics, and the artificial intelligence today is up to scale to do that; and to make that data flow to the front lines so that it can be used by the agents effectively in order to solve customer problems I think is a very important area that organizations really need to pay attention to and be more effective in making that connection. They are doing it for a seat allocation. They are doing it for pricing. They need to do it for a customer problem solving.

The third area, the frontline people are doing things that are making that data are more useful. The big data organizations currently work with and analyze is looking at broad patterns, whereas agents are responding to individual, specific, contextualized problems. And they do something very useful, like Sarah did for me at Reno airport: they try to figure out, how can I modify, adapt, and creatively generate a solution that does not exist in the database but is very relevant for this context. I think the key lies not in just moving data to the front lines but to learn from the frontline creativity and make the big data better, smarter, and more effective.

SARAH GREEN CARMICHAEL: Well, Jagdip, this has been really a fun conversation. Thank you for spending some time with us—and for getting back to where you needed to be to have this chat.

JAGDIP SINGH: Thank you, Sarah. I’m delighted I was able to make it in time—and to Sara in Reno, who made it happen.

SARAH GREEN CARMICHAEL: Yes. Thank you, Sara in Reno.

That’s Jagdip Singh. He’s a professor of marketing at the Weatherhead School of Management at Case Western Reserve University. Read more about his customer service research in the article “’Sorry’ Is Not Enough” in the January–February issue of Harvard Business Review or at HBR.org.