Predictive Analytics Can Help Save Costs
Jan 16, 2020
Utilizing predictive analytics allows organizations to use the known to anticipate the unknown. There are patterns in an organization's data on past transactions and business operations that can be uncovered using machine learning algorithms in order to predict the future. Used well, predictive analytics can help an organization become more efficient by predicting resource needs, flagging risks, and identify suspicious transactions, thereby saving costs.
Below are several examples of the cost-saving utility of predictive analytics in healthcare, manufacturing, and other industries. These examples demonstrate the potential of predictive analytics to streamline operations and eliminate losses.
Having the right amount of resources available at the right time – without excesses or deficiencies – can be difficult, and surpluses and shortages are both costly. One way that predictive analytics can save costs is by streamlining resource allocation. For example, Vanderbilt University Medical Center uses the previous year’s Operating Room traffic data to predict staffing needs. Previously, the hospital’s operating room traffic estimates could be off by as many as 30 surgeries. An unforeseen uptick in traffic would mean calling in additional staff last-minute or keeping staff on for additional shifts at overtime pay, while a downtick would result in excess, underutilized staff. Now, the hospital makes staffing decisions according to a predictive model. They report that model estimates are usually within 8 surgeries of the actual number, and the hospital saves a tremendous amount of money on staffing.
Analogous methods have been applied in supply chain management to maximize the value of materials and reduce waste. Magnitogorsk, an iron and steel works company, cut down on three million Euros of raw material costs by using a predictive model to determine precise amounts of ferroalloys necessary to produce desired steel grades at the lowest costs in an individual smelting. Predictive analytics allowed them to do more with less.
Another way that predictive analytics can be used to save costs is by identifying risks, enabling more effective targeting of resources towards managing those risks. For example, casino operators such as Macau are using predictive analytics to identify high-value players who are likely to throw in the towel. Casinos have an interest in keeping such players happy during the inevitable ups and downs of gambling so that they’ll be willing to continue playing. If a player suffers too many painful losses, he or she may lose interest and stop coming back. By analyzing the size or frequency of losses and other factors to determine an individual player’s risk of stopping, the casino can deploy staff to offer vouchers or other incentives to ease the pain of loss and incentivize continued play. Casinos already use such incentives to court players, and predictive analytics helps them do it effectively.
Similarly, predictive analytics has been shown to be effective in identifying hospital patients’ risk of readmission within 30 days of discharge after a treatment. When staff time is limited, targeting follow-up efforts towards high risk patients can maximize their impact. Under Medicare’s Hospital Readmissions Reduction program, hospitals are subject to fines when patients are readmitted within 30 days of discharge, so identifying patients likely to experience relapses and complications can help hospitals improve patient outcomes while avoiding costly fines.
A third cost-saving application of predictive analytics is detecting and preventing fraud. Banks and credit card companies have been using predictive analytics to identify fraudulent transactions for years, and other industries are beginning to benefit from similar methods. For example, Amazon uses predictive analytics to detect fraudulent return requests in its online marketplace, rooting out bad actors before wrongful payments are disbursed.
Fraud comes in many forms, and these forms are constantly changing as fraudsters devise new tactics to get around an organization’s existing checks. With predictive analytics, an organization can detect patterns associated with known fraud schemes in order to stop fraudsters from using them. Predictive analytics can also be used to detect anomalies in the data that may be evidence of a new scheme.
These are just a few examples of the many cost-saving applications of predictive analytics. Predictive analytics can help maximize an organization’s resources by predicting staffing and supply needs, targeting staff efforts effectively, managing risk, as well as detecting and protecting themselves against fraudulent activity.
Contact Us here to learn how ASR can help your organization harness the power of predictive analytics.