Travel Stress Quantified Using Big DataTravel Stress Quantified Using Big Data

5 Big Wishes For Big Data Deployments

Millions of workers are on the roads, railways and tarmacs every day. But what's the impact on productivity when those trips go wrong, when business travelers experience stress and suffer indignities large and small?

CWT, a $28 billion global travel, meetings and events management company, wanted to find out.

Through its CWT Solutions Group consulting arm, the company embarked on an initiative last June to examine some 15.3 million trips in its transaction database and combine this information with several other internal and external data sets.

The result, announced in April, is an algorithm-based tool called the CWT Travel Stress Index (TSI). The TSI measures the financial impact of lost productivity incurred through trip-related stress. The tool is now being used to inform CWT's recommendations to its clients.

[ State and local governments have only a fraction of the resources they'll need to handle future data volumes. Read more at Big Data Will Stymie Local Government. ]

"We had the transactional data and we had some traveler profile data, but it was scattered and we had to bring them together," Catalin Ciobanu, director of innovation, big data analytics at CWT, told information during a Skype call. A physicist formerly with Fermi National Accelerator Laboratory (Fermilab), Ciobanu was profiled by information last December.

Along with the transactional data, Ciobanu gathered survey data on more than 7,000 CWT travelers. "We asked them to rank, on a stress scale of one to ten, 33 activities," Ciobanu said. (In the first iteration, just 22 of the activities were used; Ciobanu hopes to add in the other 11 in future versions of the tool.)

This part of the research revealed three types of stressors: lost time (such as the inability to work on a plane or in a hotel without an Internet connection); surprise (such as lost luggage) and a third category Ciobanu describes as "routine breakers" (such as having to wake up early or the inability to eat healthy foods).

This data also highlighted demographic differences. For instance, female travelers were more stressed than their male counterparts in virtually every category, and more senior -- presumably older -- workers experienced more travel stress than junior employees. Interestingly, senior workers were less likely to be stressed by surprises.

More details from the stress trigger analysis, with variations by demographic category, can be found here.

Calculating Costs

With the transactional and perception data sets in place, CWT collected compensation benchmark data from across all industries, countries and job titles, as well as information about geopolitical risks. The goal was to calculate the productivity hit caused by stressful travel for different types of travelers.

"When we looked at it all, we saw there is an irreducible component of stress that is close to 70%," Ciobanu said. However, this still leaves 32% that can be acted upon with different travel policies. For example, CWT can use to the model to advise its clients about connectivity options for each stage of a trip or to recommend a specific carrier based on on-time or lost-luggage performance.

While the stress index is an interesting use of data, travel management companies like CWT and American Express Travel have been struggling to prove their value in an increasingly self-service world, Andreas Weigend told information in a phone call. Weigend, the former chief scientist of Amazon.com, now teaches at Stanford University and directs the school's Social Data Lab.

In the 1980s, travel agent "experts" accessed arcane back-end systems; in the 1990s, Web-based systems like Expedia and Orbitz arrived; in the 2000s, social media-centric travel review sites appeared. "The next phase [of air travel booking] is the connection to the social graph, the connection between people," Weigend said, noting that air carrier KLM now allows travelers to add their Twitter handle on their flight's seating chart.

CWT is not alone in using data to help sort out stress of potential trip, according to Weigend. The Web travel booking site Hipmunk includes, along with the standard ways to search flights by price and time, a tab called "Agony" that sorts flights through a combination of best price, shortest length and fewest layovers.

Meanwhile, CWT's Ciobanu does not plan to stop with the stress index. Hinting at future plans, he said he wants move from "travel policy" to "traveler policy" -- that is, making the model more and more individualized. Without offering details, he recommended watching for announcements in the next nine months to a year.

The Travel Stress Index, which runs under Linux and uses about 2,000 to 3,000 lines of C++ code, looks at around one million trips per minute. While the TSI is proprietary and uses transactional data of CWT clients for its benchmarking, Ciobanu and his team have released details about the framework, called "Trip Reconstruction," on which it was built.

E2 is the only event of its kind, bringing together business and technology leaders across IT, marketing, and other lines of business looking for new ways to evolve their enterprise applications strategy and transform their organizations to achieve business value. Join us June 17-19 for three days of 40+ conference sessions and workshops across eight tracks and discover the latest insights in enterprise social software, big data and analytics, mobility, cloud, SaaS and APIs, UI/UX and more. Register for E2 Conference Boston today and save $200 off Full Event Passes, $100 off Conference, or get a FREE Keynote + Expo Pass!