skip navigation

connect

March 2020

CWT

Predictive modeling: A crystal ball for business trips

“Life can only be understood backwards; but it must be lived forwards”, the Danish philosopher Søren Kierkegaard once said. That is true. But it would be much better if we could understand life forwards – which is why we use predictive modeling. It may sound like a pipe dream, but it is already part of our everyday lives in many areas: for example, when Amazon tells us what to buy or Spotify suggests suitable songs, and Facebook fills our timeline with articles that meet our interests. What they generally want us to do is to buy, read or listen!

On business trips, having this foreknowledge has a much greater impact, which is why predictive modeling is particularly worthwhile. It can be used to anticipate travelers’ preferences, for example, or forecast prices for different transportation means.

We use more than 85 indicators to make our predictions. This means collecting, computing and analyzing large amounts of data, including existing travel data, commodity prices and macro-economic indicators. We take factors into account that are already known – such as holidays – as well as more uncertain ones – such as the weather.

Using algorithms, we try to identify patterns and correlations. These in turn help us to anticipate companies’ future travel spend, which we break down to the number of trips and costs per trip.

These data and findings allow companies to take measures or at least adjust their course before the travel budget is spent. For example, at times of peak demand, they can point travelers to alternative airlines or hotels. At the same time, companies have a better negotiating position vis-à-vis suppliers if they already have a clear picture of their targets and maximum rates. Or they can limit the number of suppliers if the prediction recommends it.

Predictive modeling can also boost program compliance, for example when companies develop personalized itineraries based on travelers’ past behavior. It also makes it easier to anticipate which travelers are likely to be non-compliant.

The CWT Solutions Group has set itself the goal of demonstrating the advantages of predictive modeling. We will continue to show you case studies and examples in the future.

Source: Christophe Renard, Vice President, Solutions Group, CWT

connect

Our monthly online newsletter connect delivers the latest information and offers from the business travel industry.

Archive