Leveraging the Experience of Long-term Team members
Huntington Beach, March, 2023
Our kids were flying to California for Christmas in 2022 when Winter Storm Elliot, an extratropical cyclone, formed over their Minnesota home. The National Oceanic and Atmospheric Administration (NOAA) called it a “historic arctic outbreak,” while the Governor of New York called it the Blizzard of the Century.
They finally made it to our house a day after the holiday and were no worse for wear. Unfortunately, things continued to get worse for Southwest Airlines.
Skies were clear, but Skysolver, the company’s legacy scheduling system, couldn’t build a flight plan for our kids’ return trip to Minnesota. When they finally got off the ground two days late, they were diverted mid-flight to Kansas City, delayed overnight, then routed to Minneapolis via Chicago. During the system outage, 16,900 flights were canceled, leaving over 2 million passengers stranded.
It cost Southwest $850M in fines alone.
Skysolver (created in the 1990s) couldn’t handle the stress of rescheduling so many flights across Southwest’s complex, point-to-point network. For years, the company’s leadership prioritized adding new features over improving the capacity of the legacy system itself; as the layers of code grew, knowledge of how the overall system worked shrank.
The issue was so well known before the incident thatSouthwest’s pilots picketed a company party to protest. Today, it’s the classic cautionary tale of technical debt in software systems. People in the industry refer to it simply as “the holiday meltdown.”
When the storm hit, the company was forced to revert to manual phone calls—a slow, analog process that couldn’t possibly replace the lost digital intelligence. Newer staff members who weren’t involved in those system changes over the previous 30 years lacked the experience needed to “patch” the logic in real time.
Legacy People?
The Holiday Meltdown wasn’t an isolated case - technical debt isa universal problem in large corporations. I’ve got firsthand evidence.
I overheard someone I’d never met talking with people from my team. Her name was TJ. She was looking for data from the new enterprise software system that would replace all the legacy systems. After introducing myself, I discovered she had written all the code for the company’s custom purchasing system two decades earlier. I asked TJ who she reported to; she couldn’t remember because she’d been moved to different teams and managers so many times.
In corporate-speak, “legacy” is code for old and outdated software. Unfortunately, people also get tagged as “legacy.”
Right then and there, I invited her to join my team.
Everything Replaces Something
When you’re trying to get a company to adopt new business software, turning off the legacy system is usually the hardest part of the project. Technical knowledge accumulates in thousands of custom reports and in employees’ minds. If you don’t decommission a system that’s being replaced, you’re forced to create new reports to combine the data.
TJ joined my team and spent the next year decommissioning more than 1,000 reports in the old systems. Some of the work was technical, creating new versions of old reports in the new system. Most of her time, however, was spent convincing people that they could live without their old reports. She carefully and caringly showed them how to adapt their work to the new system.
That wasn’t the first time my data team used legacy knowledge to help the company move forward; during the five years the IT team was busy setting up the new ERP system, my data team decommissioned seven other major business systems.
“Legacy” systems have a bad reputation in corporate America; legacy people shouldn’t. When I met TJ, I knew I’d found our secret weapon.
To remind you of this week’s data concept, enjoy Secret Separation by The Fixx, from the Frictionless Data Spotify playlist.
For the full story about making data flow faster and better, check out Frictionless Data on Amazon.
