Originally published October 30, 2024. Updated August 2026.
Self-service was supposed to make corporate travel simpler, cheaper, and less stressful. Online booking tools have been the default for two decades. Has self-service actually delivered on that promise? Or has it just shifted the work onto travelers themselves?
Here’s the self-service efficiency myth in corporate travel, stated plainly. Automating a task is not the same as making it effortless for the person doing it. When I first wrote about this in 2024, self-service wasn’t new. What was new was how far it had spread.
The question I keep coming back to is simple. Does self-service actually deliver the efficiency it promises? Or does it just relocate the effort onto the traveler? AI agents are now entering the booking flow. That makes this question more relevant, not less.
Self-Service Has Spread Across the Entire Journey, Not Just Booking
In 2024, “self-service” mostly meant the booking tool. That’s no longer where it stops. Walk through a single trip today. Where does the traveler now do the work alone? Planning. Booking flights and hotels. Arranging ground transport. Checking in and tagging their own luggage. Finding support mid-trip when something goes wrong. Checking out. Resolving anything that comes up after.
Two years ago, a human still handled most of that list. Today, the traveler handles nearly all of it.
The Real Trade-Off
Self-service earns its popularity honestly. It’s fast. It gives travelers far more choice than waiting on an agent ever did. It creates a sense of control, often close to instant gratification.
GBTA survey data backs this up. 78% of U.S. business travelers say they prefer self-serve platforms when planning and managing a trip. The preference holds elsewhere too: 74% in Canada, 73% in Spain, 77% in Italy.
That preference is genuine. Cost-cutting travel programs didn’t manufacture it.
The tradeoff is real. Self-service assumes the traveler can act as their own expert. Simple, familiar bookings are fine. Complex bookings are not. Unfamiliar destinations are not. Anything with real ambiguity demands research, time, and patience many travelers don’t have in the moment. What looks efficient on the provider’s side can become a higher time-and-effort cost on the traveler’s side. That’s the opposite of what “efficiency” is supposed to mean.
When You Actually Need a Human, Self-Service Gets in the Way
This is where the myth does the most damage. The research backs that up directly.
A BCD Travel survey of 875 business travelers found something clear. Pre-trip stress comes far more from complex procedures and disruption than from booking itself. Understanding required documentation is the top stress factor, at 54%. Rebooking after a change follows, at 44%. Booking flights and accommodation rank as the least stressful parts of a trip, at 18% and 16%.
Self-service works fine for the calm, predictable part of a trip. It fails exactly at disruption. That’s exactly when the stakes are highest.
Travelers know what they want when disruption hits. A TravelPerk survey found 33% choose a phone call with a human agent as their top preference. Another 22% prefer speaking in person. That’s 55% wanting real human contact. Online self-service comes in at 23%. An AI chatbot trails far behind, at just 9%.
Yet the tools don’t support that preference. GBTA’s 2025 research found only 13% of travel buyers say their booking tool actually supports self-service changes and cancellations. That’s the single most cited pain point in managed travel. A separate GBTA poll found 30% of corporate travel buyers are reevaluating their TMC this year. Technology dissatisfaction is the top reason, cited by 39%.
Anyone who has tried to escalate out of a chatbot knows how difficult that path can be. AI-driven chat has made this worse, not better. It has gotten more polished at sounding helpful. It’s just as hard to escalate past.
Part of the problem is structural. A 2025 Sabre survey of nearly 500 travel agencies found 91% run four or more separate booking systems. More than half run seven or more. When a traveler can’t get a straight answer from self-service, the reason is often simple. The provider doesn’t have one system with the full picture either.
The same self-service creep shows up physically, too. Airport self-check-in. Self-tagging luggage. It’s efficient for a solo traveler with a backpack. It’s real friction for anyone traveling with family or significant luggage.
Who Actually Captures the Efficiency Gain?
Every point above shifts work that used to be done by supplier or TMC staff onto the traveler. That shift reduces the supplier’s own labor and operational cost, by definition. My read is that this is where the story gets uncomfortable. The traveler side of this ledger is hard to measure. The company side isn’t. It’s well documented.
Industry fee benchmarking puts a self-service online booking transaction at roughly $7.84. The same booking made with a live agent on the phone costs $25.20. That’s a company saving more than 3x per transaction, just by moving the work onto the traveler. Layer that across a program and the numbers get bigger. GBTA research shows a centralized self-service platform can save up to 20% on travel costs versus the open market. The average savings rate is 13.4%. Booking Holdings’ CFO reported a roughly 10% year-over-year drop in customer-service cost per booking in 2025. Generative AI gets part of the credit for that improvement.
But that transaction fee is only part of the picture. It’s also the transparent part. There’s a second, murkier layer underneath it. TMCs and suppliers also earn revenue from supplier commissions and incentive payments. That’s the money a hotel or airline pays a TMC for routing bookings its way. Industry reporting on TMC fee transparency is blunt about where that money usually ends up. Unless a contract spells out who keeps it, the TMC does. That means some of the labor-cost savings from self-service never shows up as a lower fee at all. It just becomes commission revenue the corporate buyer never sees.
So the fair version of this story has three parties, not two. The traveler absorbs the labor. The corporate buyer captures some of the savings, in the form of a lower visible transaction fee. The TMC or supplier likely keeps another slice, as commission revenue that rarely gets disclosed. For a travel manager, the real question isn’t just whether the OTA or TMC saved money. It’s whether that saving showed up anywhere the buyer could actually see it. And whether it came with better traveler outcomes, or just less support.
None of those savings are traveler-side metrics. They’re measured in dollars, not in traveler time, stress, or satisfaction. This post has been tracking that other ledger the whole way through. Pre-trip stress comes from documentation and rebooking, not from booking itself. A majority of travelers want a human the moment a disruption hits. They often can’t reach one easily.
Self-service in corporate travel has been sold as an efficiency story for the traveler. The actual, measurable efficiency gain has landed almost entirely on the company’s side of the ledger, with a piece of it going somewhere the company itself can’t fully see either.
The traveler absorbed the labor. The corporate buyer captured part of the savings. Some of it went somewhere else entirely, unseen by either side. That’s not a scandal. But it’s worth naming plainly, instead of calling this a win for the traveler when the data says otherwise.
The Human Connection Was Part of the Product
There’s a piece of this that no survey about tool preference captures. It’s the familiarity of dealing with the same person, someone who knows your account and your preferences, instead of an impersonal system every time. Losing that relationship isn’t just nostalgia. It’s a real loss of support and advocacy for travelers. It’s also the first thing to vanish as self-service expands.
Where This Is Heading
This same pattern is now repeating with AI agents. They’re often pitched as the next leap in self-service efficiency. Book It, Agent and The AI Illusion both make a version of this same point. An AI agent that can’t hand off cleanly to a human has exactly the same failure mode as a chatbot with no escalation path. It’s self-service with a friendlier interface. It’s not a different model.
My Take
Let’s call this what it is. Self-service in corporate travel was never really an efficiency gift to the traveler. It was a cost transfer, dressed up as convenience. Companies can point to real numbers. 3x cheaper transactions. Double-digit program savings.
Travelers can point to real numbers too. Stress runs highest exactly where self-service is weakest. More than half want a human voice when something goes wrong, and can’t easily get one. My take is simple. The industry hasn’t been honest about which side of the ledger the “efficiency” actually landed on. Parts of the industry may not even know the full answer themselves, since some of that value never gets tracked past a supplier’s own commission statement.
That doesn’t mean rolling self-service back. It means building the one thing it was never designed to do well. Know when to get out of the way. The right answer isn’t self-service maximized. It isn’t autonomous AI agents maximized either. It’s AI-driven automation with a human genuinely in the loop, placed at the exact points where self-service breaks down. Disruption. Complexity. Real ambiguity.
Three things need to happen, concretely. Build the escalation trigger into the workflow itself. Don’t leave it as a support-ticket afterthought. Make the human handoff measurable. A vendor that reports booking-time savings to the penny can report escalation-response time to the minute. And put some of the savings back into the traveler’s experience. Some of that money already sits with the buyer, as the transaction-fee gain. Some of it likely sits with the TMC or supplier, as commission revenue nobody talks about. Either way, it should show up as a faster human handoff and genuinely usable support. Not just a lower invoice.
That’s the actual efficiency story worth telling in year two of this shift. It’s not “self-service versus assisted.” It’s about who’s accountable for making sure the savings don’t come entirely at the traveler’s expense.
Sources
- GBTA, Direct Travel, Spotnana, TROOP (2025). Achieving the Perfect Business Trip. https://gbta.org/achieving-the-perfect-business-trip-new-study-reveals-top-challenges-and-technology-solutions-for-success/
- GBTA (2025). Business Travel Industry Kicks Off 2025 with Optimism Amid Evolving Challenges (Q1 2025 Business Travel Outlook Poll). https://gbta.org/business-travel-industry-kicks-off-2025-with-optimism-amid-evolving-challenges-according-to-latest-gbta-poll/
- The Business Travel Magazine (2025). Just Under a Third of Buyers Considering TMC Switch in 2025. https://thebusinesstravelmag.com/just-under-a-third-of-buyers-considering-tmc-switch-in-2025/
- Booking.com for Business. The Rise of Self-Service Booking in Business Travel. https://business.booking.com/en-us/business-travel-resources/blog/rise-of-self-service-booking-business-travel/
- Sabre (2025). Information Overload: Sabre’s Global Travel Agency Survey Finds Content Fragmentation Is Driving Up Agency Costs. https://www.sabre.com/releases/information-overload-sabres-global-travel-agency-survey-finds-content-fragmentation-is-driving-up-agency-costs-and-undermining-customer-experience/
- BCD Travel (2025). Traveler Survey Reveals Stress Factors of Business Travel. https://news.bcdtravel.com/traveler-survey-reveals-stress-factors-of-business-travel/
- TravelPerk (2025). The State of Travel Disruption. https://www.perk.com/blog/business-travel-chaos-survey/
- Perk (fka TravelPerk) (2025). Corporate Travel Management Company Fees (& How to Reduce Them). https://www.perk.com/blog/travel-management-company-fees/ — source for the supplier-commission transparency claim: “unless the contract spells out who keeps that money, the TMC often does.”
- PYMNTS (2026). AI Cuts Booking Holdings’ Customer Service Costs 10% as Volumes Rise. https://www.pymnts.com/earnings/2026/ai-cuts-booking-holdings-customer-service-costs-10-as-volumes-rise/
- Book It, Agent: Why Agentic AI in Corporate Travel Is Still Missing the Point — rajeevgoswami.me
- The AI Illusion: Why ChatGPT Won’t Kill the OBT — rajeevgoswami.me
Rajeev Goswami is CEO of WWStay and a member of the GBTA Technology Committee.
Leave a Reply