Have you ever been stuck in an airport terminal, drinking an overpriced smoothie under a cloudless sky and wondering what delayed your flight by several hours? For airlines like Lufthansa, even a thirty second delay in one process can have a domino effect. That’s why it’s imperative to keep every tiny movement from boarding to safety checks as efficient as possible. Without access to real-time data analysis, no one at Lufthansa could possibly get an accurate read on the company’s global processes. Philipp Grindemann, Head of Business Development and Project Management, Lufthansa CityLine, says the company uploads 50 million activities regarding 2 million flights into Celonis each night. With a workload like that, it’s no wonder the company uses artificial intelligence to predict delays and identify opportunities to make up lost time. When a complex process impacts millions of customers, the ability to make intelligent decisions based on real-time insights is essential.
Watching things in real-time and being able to act on them as opposed to reviewing quarterly reports is a monumental improvement to a process. It essentially turns a company from a cruise liner into a high-speed ferry. You can make decisions on the fly, change direction, and innovate without having to wait for a monstrous structure to catch up. In companies that aren’t using Process Mining, teams simply can’t make improvements fast enough to respond in real time, due to the scale of variables. Process Mining turns raw data into an event log, and when applied, its analysis shortens processes by reducing the number of steps involved. It also shaves time off each step by identifying common snags and fixing them, allowing a company to reap benefits. When Process Mining is conducted using real-time data and analytics, companies can improve at the same speed their employees or even customers make decisions. As Uber’s Global Head of Process Excellence Martin Rowlson puts it, the goal in ensuring customer satisfaction is to reach a threshold of having no support contact at all. Because if a customer’s journey goes so smoothly that they never need to reach out to your company, it means their expectations were met or exceeded. But no company can figure out all those tiny micro-decisions and data points without immense speed.
When businesses moved at a slower pace, there wasn’t as high of a demand for real-time data analytics, but customers have become increasingly attached to speed. And companies must deliver on systems that are unthinkably complex. As an example, consider how many tiny processes it takes for a person to order an Uber and take a trip. Now, multiply the row of data that simple process would create by 100 million. According to Uber’s Rowlsen, that’s how much data the company contends with on a daily basis. If Uber analyzed information on its rides on a daily basis, or hour by hour, the company would immediately be overwhelmed with support tickets and snags in its processes. Without the ability to see under the proverbial hood—and act— at every moment, ridesharing simply couldn’t exist the way it does.
Rowlson says real-time analytics allowed Uber to incorporate Task Mining into their already successful Process Mining program, and that yields a more holistic and complete look at the company’s systems. While Process Mining identifies all of a company’s systems, people, and activities, it takes Task Mining to fill any blind spots that occur when employees move from app to app. Given the rise of business process outsourcing (BPO) and complexities of platforms like Uber, it’s become vital for companies to collect data from myriad sources, and that’s really only available through Task Mining. It’s simply a full picture of a company’s daily operations. And as that picture gets increasingly clear and detailed, it requires an input of data funneled in at faster speeds. Advanced changes to a company can’t happen until all viable data is on the table. And that’s the key to success, according to Edd Wilder-James of the Harvard Business Review. “The biggest obstacle to using advanced data analysis isn’t skill base or technology,” he says. “It’s plain old access to the data.” And access isn’t useful unless it’s immediate, which proves the value of real-time.