Many of the specific challenges that the travel industry faces in big data result from its long-term usage of information systems for key processes. One consequence of this is that key data is often fragmented across multiple functions and units.

　　Many of the specific challenges that the travel industry faces in big data result from its long-term usage of information systems for key processes. One consequence of this is that key data is often fragmented across multiple functions and units. For example, airline data on the passenger experience is spread across flight operations, baggage, loyalty programmes, complaint databases, and external sources like social media.

　　In order to make effective decisions about how to promote offers to customers and deal with unexpected scenarios, airlines need to combine all of this information into one data warehouse and one set of algorithms, which would require considerable investment. Creating an integrated source of customer information is not only expensive, but difficult no matter how large the available budget.

　　Another result of the long-term use of information systems in large, established travel companies is that big data technology architectures will have to coexist with existing hardware, software, and databases. Those “legacy” tools and the data they contain are still necessary, and will still be useful in analyzing and improving travel operations and passenger relationships.

　　Big data technologies may be the only technologies for startup and purely online travel firms, but large companies will have a hybrid environment for the foreseeable future. This will lead to challenges of IT architectural cohesion and efficient functioning of all these new and old systems.

　　In addition, the real-time IT architectures used by many travel industry companies can’t run on Hadoop or other open-source environments. Known as TPF or Transaction Processing Facility, they were developed by IBM in the 1960s and 70s, and have been refined ever since.

　　We are currently decommissioning TPF in favor of open source systems, but many airlines and some hotels still use TPF. These systems could not be ported over to a big data-style platform, but would have to be completely replaced.

　　One final challenge for travel firms, as well as businesses in general, is the difficulty of maintaining a sustained competitive advantage from big data. Some U.S. based airlines, for example, developed early competitive leads in such areas as revenue management and customer loyalty analytics. Today, however, such programmes are widely distributed throughout the airline industry, and are common in hotels and passenger rail as well. Maintaining competitive advantage requires continual innovation, unique data, or experimentation with new technologies.

　　What do you think? What are the most pressing big data challenges for you?