A travel manager who is interested in which tactics might support best practice strategy in travel management could do worse than track KDS's activity.
Long before it was purchased by American Express, KDS was demonstrating its value as an innovator in travel and expense management technology.
Its latest initiatives, which are based on usage analytics and machine learning or AI, as it's also known, continue this trend.
We hear a lot about artificial intelligence and other techniques but it's worth thinking about how their application can improve travel management outputs.
KDS will use it "to expand on languages and different types of data fields recognised by the application and exploring functionality that compares OCR [optical character recognition] data to expensed items to identify any manual changes and possibly indicate errors or fraud".
In practice it means that the system will "self-teach". For example, it will be able to determine whether an expensed meal receipt was breakfast or dinner — quite important for any travel manager wanting to know if it's more advantageous to target breakfast or the upgraded WiFi as what to ask hotels to include within the negotiated room rate.
But the KDS enhancements also make life easier for travellers. OCR, which is available on mobile as well as the desktop, allows digital expense reports to be updated from a photo of a receipt. The ability to use mobile increases the opportunity for doing expense claims at a time that suits. The ability to do this by only taking a photo certainly reduces the time travellers spend on expense forms and increases their ability to complete them properly.
The new enhancements also allow multiple taxi bookings with a trip booking. This sounds small but it's another way of encouraging compliance and gaining data. For years travel managers have been at pains to point out the benefits of booking a hotel at the same time as booking the flight. The ability to include "multi-leg taxi bookings" means it is now possible for the traveller to book the airport transfer and the transfers between the hotel and the conference centre or customer meeting the next morning in the air and accommodation booking.
The easier it gets for travellers to book preferred partners through preferred channels, the more compliance there will be to corporate travel policy. And more data will be harvested.
The opportunities for machine learning to transform the data to rich data are many. It could, for example, identify from the receipt whether savings could have been made by booking the rail tickets outside of peak travel time or booking further in advance of time of travel. It could identify from a hotel receipt whether parking fees are a significant expense that will need reviewing in the next round of hotel RFPs.
The days when the official travel data couldn't tell you what proportion of a traveller's hotel bill could be attributed to the room rate or whether the rail expense was for an expensive seat on a short journey or a cheaper seat on a long journey are history.
Making the travel booking and expense reporting process more user friendly can only improve compliance and these slicker tools also have the very nice by-product of yet richer and richer data.