July 8 was not a good day for three major organizations. The New York Stock Exchange, United Airlines, and the Wall Street Journal each suffered major service disruptions. And while some were speculating that a cyberattack was the culprit behind the outages, the truth – that it was merely a three-pronged coincidence – was far less alarming, if far more mundane. The cause of the outage, a faulty router, was especially mundane for United, but the airline nonetheless was forced to cancel four “mainline” United flights and 55 flights operated by its United Express affiliates. In all, more than 1,200 delays were reported across United’s nationwide network.
To survive any service disruption, airlines and other service providers should communicate proactively and appropriately to their customers, diagnose what went wrong, and continuously learn and adapt in order to prevent future disruptions. These three building blocks become the foundation of an agile operation, which can withstand major disruptions and can regain customer trust. Big data analytics can enhance efforts in each of these areas to optimize an organization’s response and minimize the impact on the business and ensure that they survive the service outage.
For the purposes of this discussion, let’s use airlines to illustrate how organizations can use big data analytics in these three areas, although the principles apply across industry.
Effective communication at the right time can be the difference between regaining the loyalty of customers who recently suffered service failures and delivering them to the competition. For example, a proactive call to a loyal customer who recently suffered a flight cancellation offering an apology and a gift voucher can make her feel valued. A personalized message to a first-time customer confirming, for example, a refund of an incorrectly charged in-flight entertainment experience can make her feel less hassled and might even incentivize her to fly with the airline again.
However, getting the right message out to the customer who most needs to hear it can be challenging, as airlines serve thousands of customers every day, and each of those customers has a unique experience.
Big data and advanced analytics can provide an answer to this challenge. Analytics can help airlines sift through thousands of delay events on a daily or hourly basis and can empower airlines with a prioritized customer list complete with appropriate messaging. Those messages can be tailored based on the type, timing, and severity of service failures, as well as based on factors such as loyalty, customer influence, and the likelihood of an individual customer to complain.
Understanding what caused a failure can help airlines strengthen their operations to a greater extent against disruptions of various sizes and significance. Surfacing the causes that led to a specific disruption event, however, often is the most difficult part of any diagnosis.
For example, when air traffic controllers went on strike in a major European hub, delay and cancelation ripples were felt across various UK and European airports. In such situations, it becomes difficult for airports to narrow down which operations lever is being strained most – ground staff, crew, or another entity – leading to a proliferation of delays and cancellations.
In situations like this, big data and advanced analytics can demystify the root cause. The analytics engine can parse dynamic network patterns and separate delays caused by network ripple from those caused by internal airport factors. This could help airports understand which internal factors (ground staff or crew or any other) were strained most and help airports design a better resource allocation mix for future disruption days.
Learn and Adapt Continuously
In a rapidly changing environment, it often can be hazardous to use static service failure trends from the past as rules of thumb to inform the future – for example, putting more ground staff for flights from a given city all year round or buffering standby for a particular part of the day to improve punctuality for the day. Such static and generic rules may lead to higher investments, such as increasing standby staff, with continuously diminishing returns.
On the other hand, what would be very effective is a continuous feedback loop, which enables operations to learn from each passing day and to calibrate their actions at the most granular level – for example, adding ground staff only in specific scenarios. Using big data and advanced analytics, one can process terabytes of data to churn out predictive indicators and inform operations continuously based on the most recent history.
A “communicate, diagnose, learn and adapt” approach powered by big data analytics can be the answer for adaptive operations, which can withstand headwinds caused by severe disruptions while retaining the most important asset: customers.
Laks Srinivasan is Co-Chief Operating Officer at Opera Solutions. He jointly oversees Sales, Platform, Shared Services, R&D, IT, Country Management, and Human Capital Management.
Laks has more than 20 years of experience, having held numerous senior positions in marketing, credit risk management, and customer portfolio management, with a focus on analytics and decision automation technologies. He has worked with a number of customers in the retail banking, capital markets, mortgage, and retail industries addressing and solving various business problems for growing the top line and the bottom line.
Subscribe to Data Informed for the latest information and news on big data and analytics for the enterprise.