by Kagan McLeod
abstract
Analyze e-mail behaviors and share best practices.
When athletes want to improve, they typically spend hours reviewing video of their performance. In the white-collar workplace, it’s hard to get such vivid feedback. But in recent years researchers have learned to mine a unique set of data that serves as a slow-motion replay of how an organization and its people function: the company’s e-mail, which shows who talked with whom, why, how, and how often.
Academics call this kind of investigation social network analysis. It has largely focused on internal communications aimed at learning how colleagues can collaborate most effectively. A new study uses e-mail analysis for a different purpose: to examine how employees interact with clients. Organizations can learn what patterns and behaviors affect client satisfaction and use the results to coach employees on more-effective communications. The researchers call this work virtual mirroring, because it helps people reflect on their style and compare it with others’. “This is one of the highlights of 15 years of research in which we show people their e-mail networks, determine what variables drive performance, and then show how people can improve their collaboration,” says Peter Gloor, the MIT research scientist who led the study.
The researchers identified 176 teams working with key client accounts at Genpact, a global professional services firm spun off from General Electric in Teams ranged in size from a few dozen people to several hundred. Twenty- six teams were designated as the experimental group, with the rest serving as a control; the two groups contained similar types of client companies (mostly from the Fortune 500) in a variety of industries. At four points during the study, researchers compiled and analyzed two months’ worth of e-mail between employees and clients; over the course of two years, they retrieved and examined more than 4.5 million messages in all. The periods during which e-mail was gathered coincided with the firm’s semiannual customer satisfaction survey, which yielded a Net Promoter Score indicating how happy each client was with the service provided by its team.
Each month during the study, team leaders attended a virtual mirroring session lasting 30 minutes to an hour. During it they were shown a scorecard containing key metrics from the e-mails. These tracked the directness of communication (meaning how frequently employees answered client questions on their own as opposed to how often they needed to loop in a supervisor), the simplicity of language in the subject line, the speed with which employees responded to clients’ messages, and the extent to which clients consistently dealt with a single employee rather than a rotating cast. The researchers hypothesized that teams with direct communication, simple language, fast response times, and consistent points of contact would receive higher Net Promoter Scores than their counterparts, and results showed that this was true. (A caveat: The researchers did not access the actual text of the e-mails; their semantic analysis was limited to the subject lines.)
The study’s most important finding involved how the feedback from the virtual mirroring sessions led to positive changes in behavior. “Employees reduced the complexity of their language and made their communication with clients much simpler,” says Andrea Fronzetti Colladon, a professor at the University of Perugia and a coauthor of the study. They also communicated more directly, responded more quickly, and tried to give clients a single, consistent point of contact. These changes had a significant effect: Over the course of the study, teams that participated in the e-mail analysis and mirroring sessions saw client satisfaction rise by 17% more than teams in the control group.
Although the results suggest that certain e-mail behaviors can improve client satisfaction, the researchers note that effective behaviors will vary according to context. “In postsales assistance [the context of the study], you want stable leaders, and you don’t want too much creativity,” Gloor says. “Clients are asking for an answer to a problem, and they almost always want an answer from the same person. But in a different context, such as new-product development, you’d need to have a more creative and more dynamic discussion, and you’d want to have people rotate more.”
In other studies the researchers broadened their work on e-mail analysis. In one, they analyzed the e-mail, phone calls, and web conference calls of top- performing salespeople and found, among other things, that high performers were more likely than their lower-performing counterparts to turn on the video camera during web calls and that they engaged in more back-and-forth with clients during those calls. And in another study at Genpact, the researchers found that e-mail analysis could help them accurately predict (up to five months in advance) whether an employee would quit—in some cases, they say, identifying that likelihood before the employee had recognized that he or she might be leaving.
It’s no surprise that the ways in which an employee communicates affect client satisfaction. Managers can benefit by regarding e-mail as a resource that leaves behind “digital breadcrumbs” that can be systematically analyzed. Like watching a video of your golf swing, sometimes looking in the mirror and studying your flaws provides valuable lessons in how to improve.
提升客户满意度有新招译者：Coral961
当运动员想提升自己的时候，他们典型的做法是花几小时回看自己平时表现的视频。在白领工作地方，则很难有这么生动的反馈。但是在最近几年研究者已经学会挖掘一套独特的资料作为一个公司及人员运作的慢动作回放：公司的邮件，可以展示谁跟谁对话，为什么，怎么样，和多频繁。
学者称这种调查为社交网络分析。它主要关注旨在了解同事间如何最有效地合作的人际沟通。一个新研究分析邮件用于不同的用途：检查员工如何与顾客互动。组织者可以了解什么样的模式和行为影响客户的满意度，并且用这些结果来训练员工实现更有效的沟通。研究者称这种研究为“虚拟镜像”，因为它帮助人们反省他们的工作方式并且与其他人进行对比。“这是15年研究中的一个亮点，我们向人们展示他们的邮件网络，确定了哪些因素可以驱动业绩，然后说明人们怎样可以改进他们的合作。”负责这项研究的麻省理工学院研究所的科学家皮特·格格尔说。
研究人员在简柏特，一个2005年从通用电气分离出来的全球专业服务公司，
根据关键客户账户分成176个工作组，规模上从几十个到几百个人不等。二十六个组被指定为实验队，剩下的作为管理队，这两队中有来自不同行业相似类型的公司客户（大多来自世界500强）。研究过程中，在四个地方，研究人员收集整理并分析了两个月内员工和客户之间的邮件；在两年中，他们读取和检查了总共超过450万条信息。在这期间一并收取了邮件和公司的半年一次的客户满意度调查表，调查表中会产生一个净推荐值，说明每个客户对他所在这个组提供的服务有多满意。 “这不是为难人” 在研究过程中，每个月（实验队的）小组组长都要参加一个持续30分钟到一个小时的虚拟镜像会议。在会议上他们会被展示一个记分卡，记分卡中有从邮件得出的关键度量。 这些关键度量追踪了沟通的直接度（意思是员工有多频繁地自己回复客户的问题，而不是他们有多频繁地需要主管介入），邮件标题的语言简练性，员工回复客户信息的速度，以及客户持续与一个员工而不是一连串人交涉的程度。 研究人员假设通过直接沟通，简单的语言，快速的答复时间以及一致的接触对象，这些组可以获得比对应组（管理队）更高的净推荐值，而且结果证明这是对的。 （提示：研究人员并没有接触到实际的邮件内容，他们的语义分析仅限于邮件标题） 研究最重要的发现包括来自虚拟镜像会期的反馈怎样积极改变了行为。“员工减少了他们语言中的复杂度，并且使他们与客户的沟通更简单，”佩鲁贾大学的一名教授同时也是该研究的合著者，安德鲁·弗朗泽蒂·科拉顿说。 他们的沟通更加直接，回复得更快，并且努力为客户提供一个单一、持续的接触对象。 这些改变产生了显著的效果：在研究的过程中，参与邮件分析和镜像会议的小组看见客户满意度比管理队的小组多提升了17%。 虽然研究结果表明某些邮件行为可以提升客户满意度，研究人员指出有效的行为会根据情境发生改变。 “在售后协助中（研究的情境），你想要稳定的负责人，你不想要太多的创造力”格格尔说。“客户正在征询一个问题的答案，他们几乎总是希望从同一个人那里得到回复。但是在另一个不同的情境中，比如新产品开发，你会需要由一个更加有创造力和动态的讨论，这样你就会希望有人轮番上阵。” 研究人员在其他研究中拓展了他们的邮件分析研究，在其中一项研究中，他们分析了业绩最好的销售员的邮件，电话通话，以及网络电话会议，并且发现，在（除了邮件以外）其他事情上，这些高业绩的人比更低业绩的人更倾向于在与客户的网络会议电话中打开视频摄像，并且在电话中与客户有更多的来回互动。 在简柏特的另一项研究中，研究人员发现邮件分析可以帮助他们准确预计（最多提前五个月）一个员工是否会离职—他们说，在某些情况下，在员工认识到他/她可能要离职前就已经识别了（离职的）可能性。 毫不奇怪一个员工的沟通方式会影响客户的满意度。把邮件当成一种留下“电子蛛丝马迹”并且可以系统分析的资源，管理人可以从中获益。就像观看你挥高尔夫球杆的视频，有时照照镜子并且研究你的缺点，可以学到如何改进的宝贵经验。
,
