Good customer service seems like common sense for businesses. But how valuable is it really? Until now, this has not been rigorously quantified across different companies. Businesses are understandably reluctant to share their CRM and sales data, and most research in this field has been based on surveys. But as more Americans seek customer service online, social media offers a better platform for analyzing interactions between service reps and customers. Using data from Twitter (where one of us works), we designed an experiment to study customer service interactions in two industries that generate a significant number of customer service complaints: airlines and wireless carriers. We found that prompt and personal customer service does indeed pay off — customers remember good and bad customer service experiences, and they’re willing to reward companies that treat them well. We identified more than 400,000 customer service-related tweets sent to the top five major airlines (American, Delta, JetBlue, Southwest, and United) and top four wireless carriers (AT&T, Sprint, T-Mobile, Verizon) in the U.S. from March 2015 to April 2016. Our sample of tweets was comprehensive, including complaints, questions, and comments. Since all tweets are public, we could review the entire conversation between the customer and the customer service agent (except direct messages) and code the interaction for attributes such as customer sentiment and tone (e.g., Is the interaction praise or scorn? Is the customer happy or angry?) We then contacted these customers on Twitter, up to six months after they tweeted at the companies, and invited them to take a brief survey. Without providing a reason for the survey, we asked them to participate in a common market-research exercise called conjoint analysis to see if their customer service experiences affected how they valued the brands. For example, for customers who had tweeted at airlines, the conjoint asked them to imagine buying a ticket for a two-hour non-stop flight. They had to choose between several combinations or “offers” that varied across dimensions such as airline, seat type, on-time arrival rate, and fare — similar to how customers would shop for fares on sites like Kayak or Expedia. We offered a similar exercise for wireless service customers. From the conjoint exercise, we could discern what value, in dollar terms, customers attributed to their preferred airline. On Twitter, 1,877 users completed the conjoint exercises – 673 of them had received responses from companies, 375 received no response, and 829 had no customer service interaction and served as our control group for baseline willingness to pay. We then tested our hypothesis: do customers who had a positive interaction with a brand’s customer service representative value that brand more? Or in management parlance, when a brand provides better customer service, will customers reward that brand with greater loyalty or pay a price premium? Good Customer Service Matters on All Platforms Customers who had interacted with a brand’s customer service representative on Twitter were significantly more likely to pay more for the brand, or choose the brand more often from a comparably-priced consideration set, compared to our control group of customers who had no such interaction. On average, across all tweets and regardless of whether the customer used a negative, neutral, or positive tone, we found customers who received any kind of response to their tweet were willing to pay almost $9 more for a ticket on that airline in the future. This extra $9 can be thought of as incremental brand value the airline has gained in the customer’s mind. In other words, all else being equal, a customer would be willing to buy a ticket from the airline even if the airline cost $9 more than its competitors. We found similar results for wireless carriers. Customers who received any kind of response to their tweet were willing to pay $8 more, on average, for a monthly wireless plan from that carrier in the future, compared to the control group. Unlike airline tickets, wireless plans are monthly and recurring, so an $8 per month higher premium can lead to a significant revenue boost. We also surveyed customers on their likelihood to recommend the brand to others, so we could derive a Net Promoter Score (NPS), a common measure of customer loyalty. We found that receiving a response improved NPS by 37 points for airlines and 59 points for wireless carriers, consistent with our findings from the conjoint exercise. (This bump is significant considering NPS scores range from -100 to 100.) In addition, these effects held up for at least six months after the interaction, suggesting some permanence to the positive impact of good service. Respond to Customers, Even If They’re Upset The connection between good customer service and brand loyalty may seem intuitive. What’s more surprising is that seeking to engage an angry or confrontational customer can also have a positive effect on brands. Handling angry customers is a daily task for any customer service rep. While most companies do earnestly try to solve customer problems, inevitably there are some problems that cannot ever be fixed — the canceled flight that causes you to miss your sister’s wedding, or the dropped calls during your critical business negotiation. In many cases, there is little a company can do to redress a customer’s specific grievance. But sometimes customers are just looking for a little empathy. When customers used a negative or even an angry tone in their initial tweet to a brand’s customer service team, we saw that the best approach was to respond to negative comments instead of ignoring them. In our study, simply receiving a response — any response at all — increased the customer’s willingness to pay later, even in cases where customers were aggrieved. While successfully resolving an issue created more brand value (about $6 for our airline sample), responding without providing a resolution was still worth about $2 in added brand value for airlines. We found even larger effects for wireless carriers. For customers who received no response, we found no statistically significant change in their willingness to pay. But, customers who got any response to their negative tweet were on average willing to pay $7 per month more for a wireless plan from that company than customers who got no response. For cases where the issue was resolved, they were willing to pay $8 more; if the agent was unable to resolve the issue, they were still willing to pay $6 more. The lesson for managers is to reply to every customer service comment online, even the proverbial “I’ll never fly your airline again!” A mere acknowledgement of the customer’s problem can defuse initial frustration and put the customer back on the road to loyalty. Instead of the customer seeing the company as the enemy, a sympathetic response can reorient the situation so that the customer now feels that the company is on his or her side. That being said, don’t ignore your happiest customers. We found the highest increases in willingness to pay actually came when businesses responded to customers who tweeted a positive comment at the company. Receiving a response to a positive comment generated $28 more for a future airline ticket and $12 more per month for wireless plans. Customers who say good things about your business are your advocates and your brand loyalists. You can demonstrate that you value them by acknowledging them and thanking them for their loyalty. Good Service Happens Fast As important as it is to respond to every customer issue, it is even more important to respond quickly. We observed that a brand can capture substantially more value by replying right away. When an airline responded to a customer’s tweet in five minutes or less, that customer was willing to pay almost $20 more for a ticket on that airline in future months. Similarly, wireless customers were willing to pay a whopping $17 more per month for a phone plan when they received a reply within five minutes. Customer service representatives need to move fast to capitalize on these opportunities. For airlines, we found that after 20 minutes had elapsed, customers were only willing to pay $3 more, a decrease of 85% in value compared to customers who received responses in five minutes or less. After an hour, customers were only willing to pay $2 more. We found that the median time airlines took to respond to the tweets in our sample was about 20 minutes, meaning that at least half of all airlines were leaving significant money on the table. While we were only able to measure the response time for interactions on Twitter, we believe fast responses can generate goodwill in all customer service channels. In our study, a response time of five minutes or less meant the airline ranked in the fastest 20% of response times in our data. We expect to see similar effects, regardless of channel, as long as the company is responding faster than customer expectations. (The average customer expects companies to help them within 5 minutes by phone, within 1 hour by social media, and between 1-24 hours for email.) These patterns also held even if the customer’s complaint was unresolved, meaning that even a short acknowledgement of the customer’s issue and reassurance that the agent is looking into it can pay off. This is consistent with other psychology research showing that we dislike the uncertainty of making a request to someone and hearing nothing back. Customers Are People, So Be Personal Another insight from our research is the value of making a personal connection with a customer requesting support. Personalizing a message by typing a few extra characters can make a huge difference. Customers who received an unsigned response showed no detectable increase in willingness to pay compared to the general population. But, when a customer service agent added their name or initials in their first reply to a customer, we observed that their willingness to pay increased by $14 for a future flight on that airline compared to those who received an unsigned response. Similarly, in the wireless industry, customers were willing to pay $3 more for a monthly plan if the agent signed their name compared to those who received an unsigned response. When agents sign their name in their tweets or posts, it humanizes them and helps customers feel that the company, or at least someone within the company, is on their side. Customers are also likely to feel more comfortable following up about an issue later if they have the name of the employee who helped them. As consumers turn to a wider array of channels for help and expect faster responses, it has become more challenging to provide customer service. Bottom-line pressure restricts what companies are able to provide without breaking the bank. Our research shows that customer service that shows empathy can drive a lot of value, and there are some simple best practices to turn aggrieved customers into loyal advocates. First, surprise customers by responding quickly, so that they feel someone is watching out for them. Even a simple acknowledgement to buy time to diagnose the customer’s issue can drive future revenue. Second, don’t shy away from responding to unhappy customers, even if you can’t immediately resolve their issue. Finally, even small gestures such as having agents sign their names or initials creates immediate value for your business.
