Travel Rockstar Interview - Steve Clagg

Travel Rockstar: Steve Clagg on Data, Gaps, and the Hunt for "One App to Rule Them All"

In this episode of Travel Rockstar, Magnus Kunhardt sits down with Steve Clagg, the technologist who spent a decade building out Microsoft's corporate travel technology portfolio after stumbling into the industry almost by accident. From paper tickets to transformers, from GDS rigidity to AI governance, Steve shares two decades of hard-won perspective on where travel technology has been — and where the gaps still are.

Let's start with something about you. How did you start your professional career, and how did you get into the travel industry?

My professional career kicked off when I got hired straight out of a PhD program in educational technology. A software security company brought me on to develop their computer-based training materials, and that's when I realized I didn't actually need the degree to get a job — so I never finished it. From there I moved into software development, did some field work in architecture, and then spent several years in software engineering at Microsoft.

I discovered travel firsthand a bit before that, actually, while working as a consultant implementing software security solutions for clients around the world. I had boots on the ground experience long before I ever thought about managing travel professionally. That came later, when I owned a platform in Office 365 for extensibility — letting third parties build solutions that plugged into Word and Outlook. A couple of those partners were travel companies, and that's when the bug bit me. Eventually I crossed over from engineering into operations and joined Microsoft's corporate travel team, where I spent about a decade building and managing their technology portfolio. So the quick version is: academia, technology, consulting, development, then travel operations.

Is there something that particularly fascinates you about this industry?

The people, honestly. Collaborating with partners, at both light scale and heavy scale, has been the most rewarding part of it for me. There's also a huge appetite for innovation in this space, and that combination — good people plus strong appetite — is what keeps me going.

At the same time, what first pulled me in technically was noticing a gap. I'd been looking at data structures in the Microsoft Graph — calendar, mail, tasks — and realized travel wasn't represented as a data object at all. Travel just wasn't getting enough attention in the data infrastructure and productivity stack for corporates, and that bothered me in a good way. People complain that corporate travel is slow and behind, and that's true, but it also means there's enormous opportunity for digital transformation. If corporate travel were already the bleeding edge of everything, I probably wouldn't be nearly as interested. Looking for gaps and figuring out how to close them has basically been my modus operandi throughout my career.

Looking back at how the industry handled the shift from a single, consolidated data source toward today's more fragmented technology landscape, do you see history repeating itself?

There's definitely a parallel. Decades ago, everything ran through the GDS, and that rigidity eventually caused stagnation — the GDSs tried to evolve, but introducing radical change was never quite in their own best interest, so you got a natural conservatism that protected the status quo. Now we're seeing direct connects, aggregation, and NDC reshaping distribution all over again, and it's putting real pressure on the GDSs to change while they're simultaneously trying to protect what they already have.

A pessimist would say they're doomed to a slow, painful decline. An optimist — which I lean toward — would say there's still a functional need for content aggregation, so that capability will simply evolve and live somewhere else in the stack, closer to the supplier or closer to the client. It's genuinely an open architectural question we haven't fully answered yet.

Who or what has inspired you, and continues to do so today?

The people I met early in my journey into travel are still my biggest inspiration. I've learned a tremendous amount from technologists like Johnny Thorson and Norm Rose, who's now retired, and from conversations with Steve Singh — I consider these people giants in how they perceive patterns and envision where the industry is heading. I also have to give credit to the media organizations doing deep thinking on industry trends — outlets like BTN, Skift, and PhocusWire have been genuinely provoking in the best sense.

And I owe a real debt to my first manager, Eric Bailey, who's the one who actually pulled me over the fence from engineering into travel and supported the direction I wanted to take. That kind of mentorship matters enormously.

What has been your favourite experience during your professional career to date?

The moments that have stuck with me most are workshops and summits — gatherings of people with like minds and unlike minds, where the dialogue itself produces a kind of synthesis you can't get any other way. One that stands out is a summit Steve Singh held in San Diego a few years back, which led to a published paper on responsible travel trends and direction. I think we need more of those think-tank style sessions, because the output — an actual product you can publish from the conversation — is incredibly valuable. It's people talking, not machines talking, and that's exactly what gets things working. It's a big part of why, in my own consulting practice now, I want to bring more client workshops and shared thinking into how I engage.

In your opinion, what is the greatest achievement in travel technology since the turn of the millennium?

I'd point to two things. The first was the shift from physical to digital — moving from paper tickets to electronic tickets eliminated a whole layer of friction almost overnight. The second, and the one I think we're only beginning to feel the full weight of, is the application of analytics to language: the rise of AI language models and the transformer architecture underneath them. That's what made it possible to predict language patterns at massive scale, and it's the next leap after physical-to-digital — something like the move from farm to industry to information to, now, intelligence.

Language is the foundation of how we communicate and how we know things — and it's not only words, it's images too. Bringing that into digital computation is genuinely transformative, and it cuts both ways: it makes the interface more natural, because I'm not forced into a rigid form to express myself, and it changes the back end, because language is also data, and predicting outcomes from that data goes far beyond the booking interface. So much of travel innovation has focused on the front-end booking experience. I think the bigger impact is still coming on the back end.

Do you have a favourite among the technologies that have been developed for the travel industry?

What excites me most is the idea of unifying fragmented experiences into common applications. We're still in a world with far too many apps for travel, and I'd genuinely like to see one app to rule them all — whether that ends up being our existing chat platforms or some new combination. WeChat was cutting-edge on this years ago, bundling so many tools into one messaging surface, and I think interoperability is only going to matter more as AI-driven and rules-based automation keeps expanding. Building a platform with a strong moat — where this application is the only place you can do A, B, and C — simply isn't going to survive going forward. Functions are going to have to open up. As enterprises connect their HR, finance, and engineering systems into unified productivity stacks, travel needs a seat at that same table, whether the connection runs through a GDS, a booking tool, a TMC, or direct with a supplier.

What would you recommend to newcomers to the industry in terms of technology?

It's genuinely a hard moment to be a travel buyer — budgets are tighter while responsibilities keep growing. I gave a talk a while back about how the role is shifting from travel program operator to what I'd call a technology steward. My advice is to build your own internal technology strategy: understand where your capability and knowledge gaps are, and if it's not realistic for you to own every domain personally, make sure someone on your team owns each one and stays current.

For resources, there's no shortage. Industry media will get you a lay of the land. Organizations like GBTA and BTN run education programs, and there are consultancies offering structured learning too — some free, some paid. But don't stop at travel-specific content. A lot of the innovation reshaping travel is actually coming from outside it, so keep an eye on FinTech and on the broader AI, data, and BI landscape. LinkedIn Learning and Google both have solid programs, and the AI providers themselves — Anthropic, OpenAI — offer free education too. YouTube is useful but messy; it works once you already have a framework for what you're looking for. And don't underestimate certification — a lot of programs are free, and having a tangible outcome you can point to in your next review with leadership is genuinely worth the time investment.

Where do you currently see the biggest and most important challenges in the industry? Do you have any thoughts on how you would tackle them?

Everything traces back to data. The most important foundation for AI readiness is clean, normalized, unified, codified data — and the governance, control, and management around it is where I think the real risk sits. Leadership tells teams to "do something with AI," and teams scramble without stopping to define purpose, North Star alignment, or how the underlying data is actually managed. What concerns me most is how little attention goes to what happens when things go wrong. Every demo shows the best-case outcome, because that's what sells — nobody demos the failure mode. That's exactly the question buyers need to be asking.

Beyond that, you have to understand what any system was trained on. If your travel application was trained on healthcare data and you're operating in finance, does it actually understand your scenarios, or does it miss things unique to your business? And your own corporate data needs clear boundaries too — you want it used to improve your own outcomes, not to train a model that ends up benefiting your competitors. If I boil it down to three things in priority order: first, define your baselines and how you'll measure progress, so suppliers can't hide behind vague promises when something underperforms; second, get your data unified and structured; third, lock down your data rights and governance. Only with those three in place can you responsibly move into the AI integration work everyone actually wants — spend forecasting, disruption handling, leakage reduction, and so on.

The last piece that's genuinely unclear to me too is total cost of ownership. Travel has grown up as a transaction-fee industry, but AI systems run on a token-cost model — essentially a CPU-cycles model — that's optimized for outcomes, not for completed bookings. I wouldn't be surprised to see pricing shift toward outcome-based models. I remember a company called Juno that experimented with exactly that — charging based on trip success and traveler satisfaction rather than on the booking transaction itself. It's an interesting signal, even if travel's economics are still far more tangled than AI's cost structure currently accounts for.

Quick-Fire Questions