Na-Na-Na-Na-Na-Na-Na-Na … Botman!
Chatbots have flooded the corporate travel space in the past two years. There’s artificial intelligence under the hood of
each, taking in traveler preferences, travel policy parameters and historic booking data and combining it with aggregated
content from traditional sources like global distribution systems and sometimes nontraditional sources like Expedia,
Priceline subsidiaries and Booking.com. Crunching it all together and delivering personalized itinerary choices in a simple
interface is just one piece of the puzzle for Mezi, and the technology is sharp. Shinde claims Mezi can process 60 percent
of requests without a failure, and should be at 80 percent within six months. When the technology does fail, Mezi has
people behind the scenes ready to take over. That human assist is where Shinde breaks with tools like HelloGbye or
Hello Hipmunk, which strive for 100 percent AI capability. Instead, it sides with Kayak co-founder Paul English, who has
launched Lola with similar human underpinnings but hasn’t fully entered the corporate market.
Where Shinde has succeeded and competitors trail is in the partnership business. Mezi hasn’t ruled out corporate-direct relationships, but Shinde’s strategy favors travel management companies. Mezi has signed about a dozen TMC
“pocket travel assistant” but as the next-generation TMC platform. “Our partners need to share that vision, ” he said.
Casto Travel CEO Marc Casto does. He partnered with Mezi to launch a white-label chatbot, plus an alternative AI-supported service workflow with specialized
agents who are unfettered by GDS codes and archaic interfaces. He sees a future in which AI solves an industry labor crisis for qualified agents. “Our AI-assisted
agents are at least five times more productive than a traditional agent, ” he told BTN in September. “And the excitement among [them] has been amazing.”
As managed travel takes its first steps into the brave
new world of New Distribution Capability-enabled direct connects between airlines and corporate clients,
one lesson has become clear: Actually, there is no
such thing as a direct connect. At least one intermediary is almost always involved. Building a distribution pipe from an airline, even a Web-based one, “is
not simple to achieve, ” Moshe Rafiah said with some
understatement. And that is why the name Travelfusion popped up in story
after story in 2017, emerging as the first major aggregator of the alternative
air distribution era.
Rafiah founded London-based Travelfusion in 2000 to bundle application programming interfaces from low-cost carriers and feed their content to online
travel agents, travel management companies and other service providers. That
experience positioned Travelfusion nicely when full-service carriers decided
they wanted to start distributing the same way, via APIs built to NDC standards.
Travelfusion now offers APIs from more than 300 suppliers worldwide. The
vast majority remain budget carriers, but the number of full-service airlines
on the list shot up from two to 14 in 2017, and Rafiah said another dozen are
in the pipeline.
Perhaps just as significantly, he said, Travelfusion APIs are now taken by
nearly all online booking tools. This summer, BTN sister publication The Beat
revealed Concur has largely stopped building airline APIs itself. “All we do is
call Travelfusion, ” said a Concur executive. Another milestone was Travelfusion’s involvement in a customised connection of Web fares from the U.K.
airline Flybe into the KDS booking tool used by PwC.
Yet, in spite of the progress, Rafiah is “not going to hide the difficulties”
of managed travel making a success of NDC. While Travelfusion and other
aggregators can join airlines to TMCs, the TMCs are struggling to integrate
those bookings into the back-office systems where they perform essential
client functions like feeding traveller tracking tools and creating management
information. Although Rafiah said it is “not an area we are interested in strategically, ” Travelfusion is actively investigating building technology to solve this
problem for TMCs in 2018.
U.S. DISTRICT JUDGE FOR THE SOUTHERN
DISTRICT OF NEW YORK
Following an eight-week trial last year, an 11-member jury concluded that
Sabre’s distribution contract provisions—as applied to US Airways’ 2011 contract, at least—”unreasonably restrained trade.” The jury also concluded the
airline plaintiff did not bring sufficient evidence to show Sabre colluded with
competing global distribution systems.
Judges are inherently influential in such proceedings: They approve or deny motions that
advance or toss claims; they decide which evidence is admissible; and they dictate the rules
of the road on which plaintiffs, defendants and
juries must drive.
This year, U.S. District Judge Lorna Schofield
added several key decisions to the years-long
antitrust case. In March, she affirmed the jury’s
verdict, denying Sabre’s bid to overturn it or retry the contract claim it lost. Sabre has insisted
it has operated within the bounds of law. In an
appeal filed this year, it further contested the
judgment. American, too, has appealed aspects
of the case. That higher court review continues.
This year, Schofield also denied a bid by American Airlines, which merged with US Air ways and
continued the case after it was filed in 2011, to
elaborate on the scope of the verdict so that it
would further clarify the rights of airlines when
faced with contract terms similar or identical to those the jury found unlawful. Airlines joining that failed request included Air Canada, Alaska Airlines,
JetBlue Airways, Lufthansa Group and United.
To several industry watchers, the resolution of the case in the district court
augured incremental, though not big bang, changes in distribution. At the
very least, it provided new leverage for airlines negotiating GDS deals. As one
airline lawyer commented after the trial: “Having a jury determine that the
provisions in that contract were illegal obviously raises very serious implications for Sabre if they want to continue to insist on them in contracts.”