It’s hard to know where we’re at in the cycle of hype for predictive analytics in business travel. There’s enough breathless excitement surrounding the concept, though, that people are getting confused about what the term actually means.
NB: This is an article from Egencia
It gets blurred with the related, but different practices of big data, artificial intelligence (AI), and machine learning (ML). Predictive analytics is distinct. With so much noisy evangelism, it’s worth focusing on how predictive analytics promises to affect corporate travel.
The differences between predictive analytics, AI, ML, and big data
If you’re not a data expert, it’s forgivable to conflate big data, AI, ML, and predictive analytics. Each involves doing innovative things with previously unheard-of volumes and varieties of data. With the increase in the volume of data, the field of data analysis has expanded exponentially in recent years. Drivers of change include advances in computing and storage along with the simplification of data integration standards. But big data, AI, ML, and predictive analytics have distinct objectives and methods.
Big data is a term that describes the collection and analysis of large, highly diverse data sets. AI uses software to analyze data and perform human-like thinking, such as studying a medical record and determining if a patient needs a different medication. ML involves software that can learn from data. For example, an ML program could see a million digital images of plants and trees and teach itself to tell the difference between them.
Predictive analytics is the practice of using AI and ML to make predictions about future events. This typically means studying past data and using algorithms to detect meaningful patterns that can suggest future events — and then alerting someone about what’s coming. Imagine that your company operates oil pumps spread out across half of Texas. Remote sensors monitor mechanical functioning and feed huge streams of data into a central predictive analytics engine. The software can interpret the data and predict which oil rig needs maintenance, before it breaks down. This is predictive maintenance.
Practical scenarios for predictive analytics in business travel
What will the travel industry do with predictive analytics? It’s tempting to get fanciful. You could dream up scenarios where software analyzes your medical history and predicts that you’ll end up in the emergency room from an allergic reaction to a mint left on your hotel pillow. A more realistic idea might involve ML learning about your company’s booking habits and making predictions about ways you might save money or increase traveler satisfaction.
Consider that you have to take a business trip to Paris. If you look up hotel recommendations on a consumer site, you’ll see hundreds of suggestions. Digging through them will be a frustrating process that cuts into your work time.
With predictive analytics, the booking system knows you and your business. It can recommend hotels where your colleagues stay in Paris. It can present suggested hotels in order of distance from your company’s Paris office. The tool might even predict the best times of day to avoid an expensive taxi ride based on traffic analysis.
Predictive analytics and multinational enterprises
The use of predictive analytics in corporate travel is still in its early stages. Innovations are on the way, though, and the future looks exciting. This is true for multinational corporations and smaller businesses. What seems likely is that predictive analytics engines for big companies will build on the extensive, existing travel history of the firm’s employees where smaller businesses will probably work with predictive analytics based on peer travel data.
Here’s an example: Each year, a global tech company sends thousands of employees to major tech conferences around the world. If their travel management company (TMC) uses predictive analytics, they could get notifications about the best way for each business team or territory to save on conference-related fares through advance bookings. For instance, the system tells EU division heads that travel costs are predicted to go up for travel to conferences in the APAC region at a certain time of year. It could then recommend sending fewer people to certain shows.
Challenges using predictive analytics in travel
A number of obstacles might impede an easy rollout of predictive analytics in business travel. One issue is the data itself. Currently, the data required for effective predictive modeling resides within many separate online providers and offline services. Without assembling all the relevant data for analysis, predictions won’t be worthwhile. This is a problem Egencia is actively working on with its travel management platform
It’s also easy to underestimate the challenge of building predictive models for travel. We operate worldwide, like our global customers. What one subsidiary of your business might consider a good travel experience — or a good deal — might not be the same as a division in another geography. In our experience, for example, business travelers in Asia prefer luxury hotels over bargain bookings. This is due to a subtlety in Asian corporate culture that generally rewards people who project an aura of success.
In this case, the luxury hotel is viewed more as an investment in marketing and sales than a travel expense. On the other hand, American travelers might want a luxury hotel, but the company’s travel policy forbids it. Getting predictive analytics right means being aware of these types of nuances and getting better at learning what they mean.
The big challenge, though, revolves around making travel predictive analytics actionable. Insights are great. But until you can turn those insights into a direct, business action, it won’t help you much. Let’s say your predictive modeling tool recommends a certain flight. It will only work for your business if the tool integrates with your travel platform and its policy data. Then, you’ll know if the booking is within policy.
Bringing predictive analytics to life in business travel
TMCs are working on bringing predictive analytics to life as a value-added feature for multinational business’ travel programs. It’s a new field, with many fascinating ideas and potential innovations. The challenge for TMCs is to understand the benefits of predictive analytics from the perspective of a global organization. The task is to stay focused on use cases that help save money, improve policy compliance, and help travelers enjoy a better business travel experience.
