Closing Chapter Case: The Emerging Field of Crisis Analytics - Part2 As referenced in part 1 at the beginning of the chapter, crisis analytics merges big data with organizational crisis management. In the discussion that follows, we offer recent applications of crisis analytics. Crisis Analytics Applications Applications of crisis analytics are relatively new. The discussion below shows examples from crisis mapping, accident prevention, crowdsourcing, the opioid crisis, and fiver level modeling- Crisis mapping. From a historical perspective, crisis mapping may be the firs documented use of big data in a humanitarian situation (Qadir, et al., 2016). The yeat was 1854 in London, England, and a severe cholera outbreak had overtaken the coty. Physician John Snow published a map of the epidemic using data that included the number of deaths by geographic location. These findings were mapped, and he argued that the deaths were caused by a water pump on Broad Street in the Sohc area of London. While these findings may seem logical today, Snow challenged an entrenched theory of the time, the idea that cholera was spread through the air and not by water (Koch \& Denike, 2010). From these humble beginnings, crisis mapping has advanced to the use of sophisticated software to track events on a geographical plane. For example, in 200sin. post-election violence in Kenya occurred, causing more than a thousand fatalitiesKenyan activlsts produced a "live-map" showing geographically where human rights abuses were taking place. Anyone with an Internet or mobile phone connection could report these abuses in real-time and, thus, provide support documentation of the atrocities which would have gone unreported In the past (Meier, 2012). In 2010, the earthquake in Haiti produced more applications of crisin mappine Uainl data from Twitter, text messages, Facebook, and mainstream medta, five opthsourch? maps were constructed using Ushahidi software to lacate survivors who nended lond, Water, and healthcare (Twarog, 2017). The amount of data was so overwheiming that over 100 volunteers through Tufts University in Boston, Massachusetts were traned to input the data. The resulting map provided the most up-to-date informaton avalabe for use by the humanitarian community. The US Marine Corps and the us coart Gyand During the COVID-19 pandemic, crisis mapping helped governmints and mediod saved hundreds of lives by using the map (Meier, 2012). grafessionals ldentify areas where infections were expanding Resources ghertid tis these hot spots saved fives. Accident prevention. Big data safety analytics has yielded impressive results, including reducing injury rates, lost workday rates, insurance fees, and worker's compensation fees (Schultz, 2015). Accident prevention data can originate from other sources, including digitally generated photographs from video cameras and smartphones. In the railway industry. video cameras have become popular to monitor train rails (ti & Ren, 2012). The aspiration is that images from the video cameras can help predict future problems. For example, Jamshidi and colleagues (2017) proposed a framework to predict rail defectis by analyzing the physical length of squats (Le., physical cracks on the rail) that can lengthen and eventually lead to structural failure. Crowdsourcing. New product ideas can be evaluated more thoroughly when both internal experts and diverse crowd members are involved in the process. This type of vetting can lessen the risk of myopia and groupthink that engulfs many firms when decisions are made without external input. Coca-Cola's disastrous launch of reformulated "New Coke" in 1985 illustrates this problem. While Coca-Cola spent. millions of dollars on research and taste tests that included over 200,000 people, executives-concerned with the firm's attrition in market share-overlooked the intangibles. Building pressure for change prompted Coca-Cola to test the new formula in such secrecy. Unfortunately, issues such as loyalty and emotional attachment to a brand were discounted (Carfagno \& Parnell, 2016). Crowdsourcing could have provided independent opinions and diversity from a wide range of participants to balance internal specialists (Chan, 2013; Surowiecki, 2005). The Opioid crisis. Drug overdoses are killing thousands of people in the United States annually. In 2016, 64,000 people died from this epidemic (Kelly, 2017). Big data is now being used to combat the crisis. With a robust quantity of both objective and subjective data available for analysis, companies can use this information to target areas of high and severe usage of the drug. In one example, the state of Missouri used data solutions from Xerox to eliminate prescription monitoring gaps that led to significant intervention and reduced the loss of life. The aggregation of data sources (i.e., big data) can help health professionals analyze patterns of impact and develop intervention plans to amellorate or eliminate problems (Kelley, 2017). Big data to analyze and solve the problem of opiold abuse in the US. The challenge has been to gather data from various; disjointed data silos across the state and mergethem into something useful. The state of Indiana has upgraded its information gathering and coding capabilities and created a new Management Performance Hub. Initially, the hub was designed with reducing traffic crashes in mind, but now, efforts are underway, to use it to create crisis mapping tools to show whete drug treatment centers should be. placed (Russell, 2017). Modeling river level behavior. Floods from france's longest fiver, the Loire, have been common over the years. In France, 74 percent of the cities are at risk from floodive. each year, and over 80 percent of the destruction from natural disasters are caused by flooding (Fertier et al, 2016). Modeling the river behavior can be beneficial to stakeholders along the river who could be affected by flooding. Fartier et al. (2016) proposed a project that utilizes stakeholders and crisis cells that provide critical information for decision-making before. and during the fiooding. The cells centralize and facilitate collaboration among the stakehalders in the regionalareas along the river. Managing the information sharing and collaboration from a geographical. perspective requires sophisticated data manafgernent. More essential data becomes available over time. The need to understand and transfer data from stakeholders becomes crucial to make timely decisions. Automation of big data helps filter and aggregate the data for effective decision-making. Because more data sources are constantly added, the situation models can become more accurate! reducing the damage; sutfering, and loss of human life. Assessing Big Data and Crisis Anaytics Big data presents numerous opportunities for crisis managers. Indeed, the potentigains from crisis analytics are immense. Data is powerfal. When analyzed propery, it can help decision-makers visualize crisis events before they occur. As such, it offess the most significant potential for crisis prevention. There are many challenges. First, data must be analyzed before it can generate Insight. Managers often have more data avallable than they have time to analyze Projects such as the Global Database of Events, Language, and Tone (GoELT) offer substantial promise but are still emerging Second, big data lends itself to paralysis by analysis, a state whercby decision-maten cannot navigate the wealth of avallable data efficiently and hence, fall to act crigs managers seek to satisfice. They are trained to make quick, workable lbut cok optinull decisions under uncertainty (Parnell \&cCrandall, 2017). Satisficing depisions may be required. The first satisfactory course of action is often selected (Fex, 2015). Criss analytics is primarily useful in the pre-crisis stage-when time is nat at a premun-tut is only useful duringa crisis when data can be analyred quickly and efificiently. Finally, crisks analytics requires extensive training and a new approoch to decigath. making. Practitioner models based on intultion and experience do not hargits alte power of the data. Also, the acadernic model of hypothesis testing aumented by perl review is also impractical. Statistical tools such as SmartpLs provide an intinthe inat fiexible, prediction-oriented approach to data analytics and modelint (Bairat al 202 ) Applying wech tools, along with a more practical approsech to data analpis, an iequiph Case Discussion Questions 1. What applications of crisis management analytics have you seen where you work? 2. What is a specific problem or crisis scenario that could benefit from crisis analytics