Oracle's January acquisition of Sun Microsystems has presented numerous travel management challenges related to data collection and preferred airline relationships. Noting the company's entry into the manufacturing and shipping business has brought new travel patterns and new markets, Oracle senior manager and global air procurement lead Rita Visser said: "We needed to understand how our footprint was going to change."
Visser during a recent Business Travel Media Group webinar sponsored by MasterCard said that combining the two companies meant new travel patterns, new markets and a different mix of travelers. Based in Redwood City, Calif., Oracle now operates offices in more than 100 countries and serves customers in 145 countries. Its annual air spend is about $270 million. According the latest Business Travel NewsCorporate Travel 100 report, Oracle has had pre-existing preferred airline deals with American Airlines, Delta Air Lines, Lufthansa and United Airlines.
As part of the Sun consolidation process, Visser contacted regional directors within Oracle's new locations to understand how travel was supported, how travel spend was analyzed and how data was gathered. "It really was talking with the people that we were bringing on from a spend perspective," she said, "then looking at it and asking, 'Is there a carrier that we have that can support that business?' and if not, then start a conversation with a new supplier."
Some of the "big challenges that you run into [when formulating new preferred airline agreements] is understanding what is going to be measured," Visser said. "You have those conversations right up front. In an established relationship, we hope that we have already closed that gap. We are talking to our major carriers on a quarterly basis."
History Lesson
In the past, Oracle faced data gaps whereby the total spend with a preferred airline was significantly greater than what the airline claimed during negotiations. "We went in with a carrier probably about two years ago, showing $10 million in spend and they were showing $5 million," Visser explained. "When we show twice what they do, we need to close that gap. Are they looking at undiscounted spend? Are they looking at tickets that aren't flown? You as the buyer want to have the upper hand with the data as opposed to the supplier. Imagine how the negotiations would have looked had we used $5 million instead of $10 million."
By using internally aggregated data, travel agency figures, credit card information and airline data processed through the Prism system, Visser proved the higher number, which indeed had resulted partly from seats not purchased with the corporate discount, unused tickets and ancillary fees. Visser acknowledged that the process was manual, requiring data gathering "on a ticket-by-ticket basis," but noted that the airline relied solely on Prism data.
She added that new airline partners must understand "that this is the number that we look at, this is how we get it from the agency and this is how we pull it internally. If you can get close to that number, great; if you can't, let's have some conversations."
Meanwhile, in working with its preferred airlines, Oracle "can't lose sight of" the priority to reconcile ancillary fee data and calculate total trip costs, Visser said. "Average ticket price is an anomaly at this point."
Visser pulls ancillary data from Oracle's corporate card and separates it by cost. For instance, "if it's $150, it's possibly a change fee," she said. "If it's $6, it's possibly a drink." The company is "trying to make sure it can normalize [the data] to understand what those ancillary costs are.
"I will continue to say to all of my suppliers that I just want reporting on [ancillary fees]," she continued. "I want to know and understand from [their] side. From our side, based on some calculations, we are able to present to [airlines] a number."
Trying to keep her "fingers on that pulse to make sure that data is correct" allows Visser to present a full picture to senior management, ensuring that internal data reporting mechanisms and agency data are accurate, and "as close as possible to one another."