Travel suppliers want to provide personalized offers to travelers. Some travel managers want to customize travel policies to better fit individual travelers and the specifics of a trip. The problem is, some online booking tools can’t or don’t easily handle the nuances.
OBTs connect with traveler profiles. Those often include basic data like an employee’s level in the organization, loyalty program info and airplane seat preferences. But they’re usually not dynamic and don’t cover all the attributes some travel managers wish would influence policy parameters. OBTs also incorporate rules engines. They can be pretty sophisticated, allowing for subgrouping and other kinds of granularity. For the most part, though, they use static rules.
Sources at booking tool providers said customers can add information and configure systems almost any way they want. But that takes some work and is difficult at scale. “If you have the time and inclination, you can really make a policy for everyone in the company,” said Deem chief commercial officer Tony D’Astolfo.
That concept has supporters but needs tech advances. OBT providers looking to make systems smarter may incorporate machine learning and artificial intelligence.
Doing so could help support more complex policies like a new one at Novartis. According to corporate travel director Kathy Kaden, travelers should consider a connecting itinerary rather than a nonstop flight if the journey is at least six hours, if the connection wouldn’t add more than two hours and if it would save at least $300.
“But it’s not easy,” Kaden said in September during the Business Meetings and Travel Technology Expo in New York. “There’s no technology that can put those three layers in place.”
The Novartis policy aims to balance cost savings and what’s reasonable from the employee perspective. Like many other programs these days, the company considers productivity, comfort and wellness. “The way people travel now and would like to travel now is very different than the way Baby Boomers traveled,” Kaden said. “Tools and policies don’t reflect that trend.”
There’s also interest among program managers in tiered policies based not on someone’s place in the corporate hierarchy but on how often they travel or the purpose of a trip.
Travel Tech Consulting’s Norm Rose has studied OBTs since they emerged. He said policy parameters are “pretty binary.” They indicate preferred or nonpreferred suppliers and identify lowest logical fares, but all that “may not be clearly defined and implemented through every channel,” he said. “Implementing subtle policy — such as the ability to fly business class after you fly two international trips in economy during the calendar or fiscal year — is difficult with current tools.”
To keep pace with e-commerce personalization, Rose said, providers may turn to new technologies to help with “so-called fuzzy logic where assumptions are being made.”
That’s what Amadeus is working on with Cytric. Jay Richmond, head of the Amadeus business travel group in North America, said during a July 2016 interview that the booking tool’s rules engine “handles fuzzy logic and scenario-based decision trees.”
In follow-up emailed comments, Richmond explained that Cytric supports static and dynamic rules. One example of the latter is a rule defining which hotel rates to display to a user, perhaps omitting rates that are a certain amount or percentage more expensive than the best available rate.
Asked about the Novartis scenario, Richmond replied that Cytric can manage three layers of policy, “however, this can have a dependency on the exact requirements.”
He boiled down the problem Amadeus is trying to fix: “As suppliers have gotten more complex in how they manage revenue and yield and how they manage content, ancillaries and fare families, the tools that travelers use to search and book haven’t kept pace. A lot of what we thought of as roguish behavior really is the tools’ inability to deal with complex scenarios of policy against complex supplier content. Policy is making travelers angry because the tool can’t support it.”
He noted that some companies need manual intervention on 20 percent to 30 percent of their online bookings because tools flag them as out of policy even though they weren’t. “That is untenable,” Richmond said.
Speaking last month during The Company Dime’s Teleconference, Carlson Wagonlit Travel Solutions Group vice president Christophe Renard agreed that “it is very complex to really personalize, as much as you would like, what is displayed to each traveler.”
The personalization trend in the travel industry “is at odds” with the concept of a travel policy, said Ben Newell, Sabre vice president of product management. “How do we create personalization inside the program parameters? If you are this kind of traveler, or at this level in the company, do we have a different rule set for you? We have some of that in place, but it needs to be richer.”
Branded fares have been a particular issue for users of the Sabre GetThere booking tool, according to Newell. “Previously there was one price that we had to evaluate policy against,” he said. “Now you have four different flavors of that price and the different criteria within that are really important for business travelers. It’s an area where we have seen a lot of challenge across the industry to keep up. It is something we are spending time on.”
Newell said Sabre also is working to enable GetThere accounts to define different policies based on the trip purpose identified by travelers during booking.
As it is, Sabre’s policy engine has “over 3,000” toggles related to policies, preferences, configuration and messaging. A rules engine offers 30 templates for air, car, hotel and rail. Those include “more than 200 different conditions and each of those has six different behaviors,” said Angela Redel-Gomez, Sabre product manager for traveler experience policy solutions. “All these things inside each rule make each evaluation unique.”
Concur’s booking tool lets users set policies by trip purpose and accommodates journeys including both business and leisure components.
When it added support for branded fares, Concur enabled clients to “choose for whom or when premium fares are allowed versus needing approval versus out of policy,” according to comments attributed to SVP Doug Anderson.
Concur clients also can pick which employees are covered by a “high-frequency traveler” policy option. “We don’t count user trips and then automatically move them between policy sets,” according to Anderson, “but they can be easily reclassified either directly in our tool or within ERP/HR tools that connect with our system.”
Regarding Rose’s specific example, Concur users can designate a booking reason code dictating that a traveler can book in business class after surpassing the economy segment threshold. Anderson also indicated that Concur, via lowest logical fare rules, can accommodate a multi-parameter policy like the one described by Kaden.
At NuTravel, chief strategy officer Rich Miller said the company sees its rules engine as a differentiator. “But does it provide everything the industry is evolving to? Probably not today,” he conceded. “We’re getting those requests. Travel managers are taking into consideration the wear and tear on the traveler. They are becoming more sophisticated on what they want to offer and allow their travelers to do.”
He said simple policies on connections and class of service have given way to gray shades. “Policy engines in most cases don’t allow that today,” Miller said. “We are looking to build those enhancements.”
The same goes for Deem. “We are not at the point where our system is modifying on the fly, based on what people do,” D’Astolfo said. Such intelligence is a goal, but “policy will not be the driver,” he added. “It is one element but the overarching driver really is information, all these data sources, to create a better search.” He referenced Deem’s Olset acquisition and the opportunity to apply Olset’s algorithms and machine learning capabilities.
Deem users already can preset different policies based on trip purpose, according to D’Astolfo. Like clients of competing tools, they can create subgroups within subgroups. Each can have its own policy parameters. D’Astolfo said he has seen some clients go as deep as four or five levels.
Is It Worth It?
Rose wondered if more sophistication in corporate booking tools would have the opposite of the intended effect. Conditional logic, policy messaging and content variability based on the traveler or the trip can serve to confuse or annoy travelers and bog down the system. These consequences could widen the perceived ease-of-use gap between consumer and corporate tools. It could lead to program leakage or travelers more often calling agents and incurring full-service fees. If it is frustrating, self-service becomes self-defeating.
According to Anderson, Concur has seen a trend toward more simplified policies. Those, he suggested, help with program adoption. “When policies are too complex or nuanced, it’s harder for travelers to make the right choices,” Anderson added. “A growing number of businesses are placing greater focus on managing exceptions – for example, air bookings exceeding cost thresholds or trips longer than X days.”
“It is a ridiculously complex area,” said Sabre’s Newell.
Sabre is redesigning the GetThere air shopping path (due out this quarter) and updating the mobile site. As it does, it’s trying to work out how to keep things simple for travelers while giving travel managers what they need behind the scenes.
“That balance is why [travel managers] struggle,” Newell said, “and why it’s harder for an OBT to make those kinds of improvements.”
For example, Newell explained, a client may wish to display a hotel property in a given market but gray it out. That way the traveler knows that content is in the system but not within policy. Maybe the client configures the system to show that property, allow the traveler to book it and present messaging. Or maybe the client does all that and also includes a field in which the traveler must explain his or her choice. “We tend to think policy is black and white — ‘you can or cannot book this’ — but that’s not the reality we have seen,” Newell said.
Additional info: A pilot of Amadeus Cytric in North America is underway at some corporate clients and TMCs. The official launch is scheduled for the second quarter. Officials at Egencia and KDS owner American Express Global Business Travel did not provide information.