“嗯，为了证明，人们不惜编造数据。14%的人都知道这一点。”——《辛普森一家》霍默•辛普森语

    电脑天才们宣称，大数据时代已经来临。

    电脑已经强大到可以收集、汇总数以兆兆字节计的信息来回答各种问题，从如何安排员工薪酬待遇，到某支抵押贷款债券的风险有多大，无所不包。

    虽然数据不会说谎，但人们使用数据的方式却极为主观。量化分析在2007年的金融危机中起到了推波助澜的作用，但是如果企业只是简单地认为，光靠一屋子摆弄数据的分析师就可以解决问题，那么结果不仅会对他们的资产损益造成损害，同时也会损害他们的企业文化和员工的福利。

    企业执行委员会（Corporate Executive Board）的执行理事施维坦克•沙表示：“数据可以帮助人们做出决策，但如果觉得所有重要的决策都可以交给电脑那就错了。”企业执行委员会最近出版了一份调查报告，名为《超越洞察力赤字：大数据时代的大判断》（Overcoming the Insight Deficit: Big Judgment in an Era of Big Data.）。报告指出，认为只要有10个分析师，就能解决公司所有数据问题的想法是错误的。

    未来那些最擅于利用数据分析来引导决策的企业将获得许多竞争优势，对于这一点没什么人会表示怀疑。不过施维坦克•沙表示，仅仅拥有数据是不够的。根据企业执行委员会，在一份针对4,941人进行的调查中，只有38%的员工称得上是“消息灵通的怀疑主义者”，他们依赖数据，但并不盲从，既不害怕置疑数据分析的结果，也敢于从他人那里索要反馈。43%的员工对数据坚信不疑。还有19%的员工很少相信数据分析，喜欢凭直觉做事。

    施维坦克•沙认为：“我们必须面对这样一个事实：我们的教育系统没有教会我们如何有效地分析数据。我们向受访者展示了一些图表，问他们这些图表代表什么意思。结果就连那些从常青藤名校毕业的学生也很难搞清这些数据究竟代表了什么。”这种教育的缺失并非没有办法弥补。该研究显示，仅仅向学生提供培训，教给他们分析工具和软件的使用方法还不够，更重要的是教会他们如何与数据互动。换句话说，就是如何批判性地思考。

    要想让员工成为“消息灵通的怀疑主义者”，企业可以建立一种数据导向型的文化（但并不是成为数据的奴隶），从CEO开始自上而下地推广。如果CEO身体力行，其他员工也很可能参与进来。

    施坦维克•沙表示，另一种有效的方法就是在雇佣数据分析师的时候，不仅要考虑他们的分析能力，还要考虑他们是否有能力和意愿去向其他人解释这些数据代表了什么意思。“招聘分析师时还得考虑指导技能，”他说。“每一个分析师都能改善几十、甚至几百个人的决策能力。”

    葛兰素史克制药公司（GlaxoSmithKline）正在用数据分析来挑战一些人们认为无可辩驳的常理。该公司北美制药部的IT高级副总裁乔•托伊介绍道，葛兰素史克正在利用数据分析来重新设计销售运作，把这项过去完全依赖人脉的业务转变成一项依赖数据的业务。

    "Oh, people can come up with statistics to prove anything. Fourteen percent of people know that." – Homer Simpson

    The era of big data is here, the nerds proclaim.

    Computers are powerful enough to gather and synthesize terabytes of information to answer questions ranging from how best to compensate employees to how risky is that mortgage-backed security.

    But while the numbers don't lie, how people use them is extremely subjective. Quantitative analysis played a part in the financial crisis of 2007, after all, and companies that think a room full of analysts crunching numbers can solve their problems can do damage to not only their profits and losses but also to their corporate culture and employee well-being.

    "Making the decision at the end of the day can be aided by data, but the thought that computers will make all the important decisions is just not true," says Shvetank Shah, executive director of the Corporate Executive Board (CEB), which recently published a study titled Overcoming the Insight Deficit: Big Judgment in an Era of Big Data. "Saying that I've got 10 quant jocks who are going to solve all my data problems is the wrong way to go about it."

    There is little argument that many competitive advantages in the future will go to those who most effectively use analytics to guide decisions. Having the data, though, is not enough, Shah says. According to the CEB, only 38% of employees in a 4,941-person study were considered "informed skeptics" who rely on data but not so much that they are afraid to question the results and solicit feedback from others. The rest of the workforce either trust data without question (43% of the study participants) or rarely trust analysis and prefer to go with their gut (19%).

    "We have to face up to the fact that our education system isn't preparing us to analyze data effectively," Shah says. "We showed people graphs and asked them what they mean, and even folks with Ivy League educations struggled to make sense of the data."

    There are ways to bridge the education gap, though. Shah's research has shown that training is ineffective when students are told to focus on the analytic tool or software being used. Instead, they should be taught how to interact with the data or, in other words, how to think critically.

    One way for a company to ensure that they turn their employees into "informed skeptics" is to work on creating a data-driven (but not data-enslaved) culture, starting with the CEO on down. If the chief executive is on board, others will likely join.

    Another effective method, according to Shah, is to hire quants not only for their analytic abilities, but also for their ability and willingness to explain what they do to others. "You've got to hire quants for their coaching skills," he says. "Each quant can improve the decision making capabilities of 10s or 100s of others."

    GlaxoSmithKline (GSK) is using analytics to roast several sacred cows, says Joe Touey, senior vice president in IT at its North America Pharmaceuticals division. The company has embraced data analysis to redesign its sales operations, moving what was once a field based solely on relationships to one that's now based on data.

    托伊称：“决定销售人员的薪酬时，我们在使用哪些数据、不使用哪些数据的问题上做出了艰难的决定，由此树立了良好的声誉。”

    自从20世纪80年代初开始，制药公司就习惯使用从厂商那里获得的医生处方数据来决定销售人员的薪酬。医生开出的处方越多，销售人员的提成就越高。因此，医生吃回扣的现象也屡见不鲜，许多医生都被请去吃免费的饭，拿免费球票。因此制药公司的这种提成制度造成了一种医生和医疗代表之间暗藏猫腻的关系，也使一些病人怀疑为什么医生要开这种药，而不是那种药。

    从去年七月开始，葛兰素史克公司废止了这种提成模式。现在公司主要依赖对医生和管理人员的调查，以及对产品知识和业务敏锐度的考核来决定销售人员的奖金 。

    最高管理层的支持、透明的激励系统，以及消费者的正面反应降低了在销售部门中推行这种新机制的难度。托伊表示：“现在销售人员和医生的关系变得不那么重要了，更重要的是事实和科学。”

    葛兰素史克公司在与政府和保险公司的谈判中也奉行了以数据为指导的策略。葛兰素史克与默克（Merck）和辉瑞（Pfizer）等公司竞争着一块非常有限的市场蛋糕，不过托伊表示，数据分析正在为葛兰素史克带来某种优势。

    北卡罗来纳州蓝十字与蓝盾协会（Blue Cross and Blue Shield of North Carolina）信息管理与分析服务副总裁苏珊•海默•穆塔夫指出：医疗保健市场瞬息万变，有效的数据分析如果得到正确使用，就可以节约许多渠道成本。

    “节约成本的唯一方法就是检验信息，进行分析，”穆塔夫说。“成本是从什么地方产生的？哪些疗法是有效的？它是在哪交付的？”

    去年，北卡罗来纳蓝十字与蓝盾协会把许多来自不同部门的分析师集中在同一个部门里，以方便他们进行协作。同时该公司也希望雇佣一些不仅能够分析数据、还能用其它人能够理解和接受的方式提交数据的分析师。

    后来该公司的管理层意识到，让一群统计学博士花时间获取数据是很荒唐的，于是他们投入大量资金建立了一个卓越中心（Center of Excellence），卓越中心的分析师以发现问题、解答问题为主，而不是四处寻找数据并检验它们的准确性。

    穆塔夫表示，这些变革导致公司的成本出现显著下降，并且提高了客户的满意度水平。卓越中心的建立已使公司节省的医疗开支达到了初始开支的将近19倍。

    尽管以数据为中心的做法已经使北卡罗来纳州蓝十字与蓝盾协会获得了不小的成功，但该公司仍有一些员工至今不相信数据分析是公司未来成功的关键。要说服这些员工也是个不小的挑战。

    穆塔夫说道：“我们需要让人们理解，今天的工作环境已经和三四年前不同了。”

    译者：朴成奎

    "We are building a reputation by making hard decisions about what data we use to compensate our sales people and what we don't use," Touey says.

    Since the early 1980s, pharmaceutical companies have used physician prescription data acquired from vendors to determine how they pay their sales force -- sales reps would make more when doctors wrote more scripts. As any doctor who has scored a catered buffet or free sports tickets can attest, this strategy led to some questionable relationships and caused some patients to wonder why doctors were prescribing one treatment over another.

    Beginning last July, GSK scrapped this model. The company now relies on surveys of doctors and managers, as well as tests of product knowledge and business acumen, to determine sales bonuses.

    Selling this new model to the sales force became easier thanks to support from top management, the transparency of the incentive system, and the favorable reaction from customers, Touey says. "It has become less about relationships with doctors and more about facts and science," he says.

    GSK also is letting its data guide how it negotiates with governments and insurers. They compete with other giants like Merck (MRK) and Pfizer (PFE) for a very limited pie, and Touey says that analytics are providing an advantage to the firm.

    In the ever-changing world of healthcare, effective data analysis can provide cost savings in multiple channels when applied correctly, says Susan Helm-Murtagh, vice president of information management and analytic services for Blue Cross and Blue Shield of North Carolina.

    "The only way to do this is by examining information and analytics," Helm-Murtagh says. "Where are costs coming from? What cares are effective? Where is it delivered?"

    Last year, BCBSNC brought many of its analysts from different units into a centralized group where they could collaborate. The company also looks to hire analysts who are capable not only of crunching numbers but also of serving them up in ways that others can understand and accept.

    When executives realized the absurdity of having Ph.D. statisticians wasting time hunting for data, they made a large investment in a Center of Excellence (COE), which allows analysts to focus on asking and answering questions instead of looking for data and verifying its accuracy.

    These changes have resulted in significant cost savings and increasing levels of customer satisfaction, Helm-Murtagh says. The investment in a COE has already resulted in medical expense savings worth 19 times the initial cost.

    While the focus on data has been a resounding success for BCBSNC, the company still faces the challenge of convincing some employees that this is where the company's future success resides.

    "We have a job to do in helping people understand that this is a different place to work today than it was three or four years ago," Helm-Murtagh says.