The Data Exchange collects, tracks and analyzes data on $60 billion in business travel spend annually. It provides data and analysis to its subscribers, which may be financial institutions, business and marketing concerns, governments and, yes, some corporate travel buyers. The data it collects, reports on and analyzes can come directly from travel suppliers, but also from a variety of government and industry sources as well as from individual corporations that contribute data to the company’s “data lake”. As such, founder and CEO Susan Hopley has a bird’s eye view of advancement in data security, data sharing and data interrogation and analysis. BTN editorial director Elizabeth West caught up with Hopley to talk about how key technologies like blockchain and artificial intelligence are changing the data world. In BTN’s extended online piece, she talks about how those developments could fundamentally change travel management.
How long have you been working with blockchain, and how has it changed your business?
Susan Hopley: The pillars of data are really privacy, security and access. The challenge has been sharing that data while working within those parameters. [The data industry] hasn’t really been able to do that. We began working with blockchain technology about two years ago, assigning digital signatures or tokens to the data sources, such that each piece of data is ostensibly de-identified from its source but also permanently appended with that signature. On the blockchain, each piece of certified, verified and de-identified data can be shared with verified subscribers to that [individual] blockchain. We know everywhere each piece of data goes. That’s an exciting place to be.
The Data Exchange operates on the idea that if you contribute data, you should be compensated for it. Tell me more about that. What that could enable for travel buyers?
Hopley: Data isn’t a monetary commodity, but understanding its value and creating a fairer sharing of that value is important. If I want to understand my carbon footprint, for example, I probably have to pay somebody to do that. Fair enough. But I also want something for the data that I’m sharing that contributes to that calculation. It’s possible you don’t want to share your own data and you just want to know what’s going on “out there,” and that’s fine. But if you contribute data to the pie, you want a piece of the revenue that comes from it. It’s easy to value the data, but you need to know your data is appropriately controlled, distributed and compensated for—and it would be compensated each time it is shared and consumed, not just in one transaction.
That is also now facilitated by a blockchain because that digital signature associated with the data can be tracked?
Hopley: Absolutely. It’s the only way to do it.
Let’s talk about AI and how it’s contributing to better business travel data. What are AI-powered processes doing now that they weren’t doing a few years ago?
Hopley: It may be horrible to think that machines are brighter than we are, but in certain ways, they can be. Travel data is very rich data, and there’s so much valuable information embedded in it. Our financial house clients correlate it to the economy, to stocks. Governments use it for development decisions. But it has to be good quality data, and that’s where AI can really come into play. It obviously needs to be taught, and we have written many scripts for it. But at this point, the algorithms display this kind of hexagonal “thinking” that is able to ping out on so many different dimensions and return to us the possible data aberrations that we need to address. It does this in ways that humans really can’t do because we have to look at it in a linear fashion—first, this, and then this and then this. But the machine doesn’t have to do that. So the capability is kind of exploding and showing us ways to improve our industry data.
“It’s easy to value the data, but you need to know your data is appropriately controlled, distributed and compensated for.”
What will that improved data and AI layers begin to bring to travel programs?
Hopley: There are so many ways you can look at data and analyze it, but people aren’t doing it because they’re in their little silos. And even if you are benchmarking with a TMC, maybe you are looking at five “like” companies or maybe 10. But why not look at a much bigger data universe and see what’s really going on in the market—especially if you can apply AI layers to it? AI will take all the data and information we’ve got, and it will be able to tell you—probably accurately, but we’ll have to make sure the models are right—which carriers you should be doing business with and what rates, or how your contracts should be different. It’s not going to be comparing one company’s negotiated rate against another. Rather, it will understand the market and the individual program and be able to deliver insights. That’s where we’re going with it. And not just for airline contracts, but that’s an example, and I find it absolutely fascinating.
You will need the data from corporations to do that.
Hopley: Yes, but why not contribute the data if 1) it’s completely de-identified and 2) the corporate stands to be compensated for it each time another company pings against it?
Is there a point at which all this just becomes available in a booking tool in real time? Or, for that matter, it's able to project a future rate to inform bookings at the point of sale?
Hopley: I think decision-making in corporate travel has been relatively myopic based on too little information at the time of purchase. So currently it may be better to have a rule in place and a negotiated rate to control spend to a known level, rather than making the best decision with the information you have in the moment.
So what happens to the concept of negotiated rates if there’s real-time data visibility? Are you saying there won’t be a need for them?
Hopley: I think that’s the right question, but I think it’s going to evolve because we don’t have validated proof that it will work. We assume it will work, so we move into this process, and we assume it will be better, but it’s too early to say. However, I’m going to go down that path [of trying to provide decision-making data at the point of sale] because I think it will, and I think it will make life a lot easier.
Is it all about rate, or how would you facilitate other decision-making factors?
Hopley: Take a hotel, for example. Does it have a restaurant? Does it have a swimming pool? Does it have a club floor or a meeting space? There are all types of decision-making factors when you are booking a hotel room. Does it have the amenities you need for the trip, and what is the price? Travel managers may want to think that a hotel is booked because it’s company policy and, therefore, it’s the sensible thing to do. It might be that it’s not the sensible thing to do. What if an adjacent hotel has better amenities, a meeting room and costs less? In an AI system, they could collect all the individual pieces of data quickly and assemble them in a way in which someone, instead of making a decision based on policy or—sometimes—on points, they can make the decision based on the value to them. Then, looking at that data, again aided by AI, the buyer could learn what specifically their travelers value and, potentially, drive better and more effective deals if they do continue to negotiate.
Thank you so much for spending the time to speak with me, Susan.
Hopley: It’s been a pleasure. Bye from the U.K. It’s time for tea.
Built with Shorthand