Maybe we don't really want travel search to get any better
Why AI hasn't (yet?) changed the fundamentals on how we search and book travel
I had to stay overnight in New York City last week.
Not really a new thing for me; we have a big office in Manhattan and I’m probably in the city at least once or twice a month. Compared to most place I have to travel, I don’t really mind New York City. I’ve lived there before, and it is a quick flight from Boston with options on the hour, and if my schedule allows, I generally try and do a day trip (and you know how much I love one of those).
But some times I need to stay overnight, and last week week was one of those times. And while traveling to NYC is easy, one problem is always that the hotel rates are ridiculous. You’re routinely looking at >$400/night for a decent place, and if something is going on in the city those days, it could top $500/night.
That said, one of the benefits of working for a TMC is that I have access to travel agent rates. Most major chains have them, and require you to register with them and present your travel agent ID when you check-in. I have learned a lot about major chains through their dedicated programs, so it’s definitely worthwhile. The rates, however, can be hit or miss, and there is often nothing available. But it is always at least worth a look, especially in a place like NYC.
What happened next is likely the same thing that I’m sure has also happened to you: an hour hotel shopping and search extravaganza.
Logging into a site. Navigating their agent rates and codes. Popping in my dates and preferred part of the city. Zooming in on the map. Deciphering the star rating for the hotel. Exploring its amenities. Looking at the pictures. Mentally noting the cost plus add-on fees. Going to the next site, repeat. Then the next site, repeat. Then checking our internal tool (Melon). Then checking Hotels.com and Booking.com. Checking HotelsTonight. Checking any discounts on my Delta SkyMiles card. Then getting sidetracked with something else, only to remember the next day I still hadn’t booked the hotel, and having to repeat it all over again.
Is searching truly a waste of time?
If you just read my tale, I’m sure you’ve probably already thought that I just wasted a lot of time needlessly, and with good reason:
I work at a travel agency and could have had any number of (much better qualified) coworkers handle this relatively straightforward task for me. I’m literally surrounded by amazing travel agents everyday.
I’m a frequent traveler and was only in NYC for a night, so did the time spent really merit the effort I was putting into it? It’s not like I was going to take a swim in the hotel pool, or enjoy the selection of entertainment on the TV in the room. I was going to be in for a dinner and out the next morning.
As a tech person who often writes and plays with AI, you would think I could have had ChatGPT do this for me, right? Why was I spending my time when technology could have done it faster and better?
All great points, but at the end of the day, none of them stopped me from investing the time. Why?
"Our souls need proof of work”
I just so happened to also come across this thought provoking article from Julie Zhuo this week. Julie was a former product design VP at Facebook and most recently co-founded a data visualization startup named Sundial. (To be clear, I don’t know Julie; I just really liked the article).
In it, she makes the point that humans actually crave “hard work,” i.e. something that takes personal effort to achieve. She backs this up with some science about how dopamine (the pleasure chemical in our brain) works, and how it is released for things we find rewarding, often with the sheer the anticipation of completing it. My kids would tell you otherwise, but apparently doing work actually makes you happier.
Contrast that conclusion to the world we’re in, or slowly marching towards, and you can start to see the holes. I mean, if you read and believe all of the technological advancement that is happening or is coming soon with AI, robotics and whatever else, then you subscribe to the idea that it is all going to make our lives SO much simpler, we’ll have nothing to do all day, like in WALL-E:
If technology makes doing everything that much easier, then where will we derive our satisfaction from? It is a fantastic question to mull over.
Don’t you just love a bargain?
As it often happens, reading Julie’s article happened to coincide with my need to book a hotel in NYC, and 1 + 1 = 3.
What if I like spending time searching for hotels? What if finding that amazing deal actually creates a sense of satisfaction for me, and releases dopamine in my brain?
For context, I grew up in Upstate New York, the middle child of 3. Both my parents worked their butts off to provide for us; my father a lifelong GE employee, my mother often doing clerical or retail jobs (sometimes at night, when we were younger) to help make ends meet. We had a great upbringing with food on the table and clothes on our backs, but we certainly did not live a lavish lifestyle. That’s how I was raised.
So as you can imagine for a person with that mindset, bargain hunting is hardwired into my brain, regardless of my current financial position. If I’m buying something retail, I always check the clearance rack first. Food is bought in bulk, and is usually generic brands, or whatever is on sale. If we need something, we wait for promos. Of course there are exceptions, but those are the general rules. I’m not cheap; I’m just frugal. And I don’t think it is a bad thing, either.
Travel is no different. For family vacations, I’m using all those miles & points my work trips have helped me accumulate. On work trips, my motto has always been “don’t be precious” so I use the airline discounts afforded to us by our partners, even if it isn’t the quickest or best option. Sure, just like in my personal life there are exceptions, but I don’t think I’m frivolous with company money.
So going back to my hour-long rabbit hole of finding a hotel in NYC — maybe that is why I spent the time doing it, after all. Was part of my brain was looking for that
4.5-star hotel at a great price to trigger a dopamine hit? Did my soul find some kind of meaning in putting in the hard work of traversing a half dozen websites to find a bargain for over an hour? In a strange way, I kind of think so.
(PS: I eventually settled on a place by Grand Central I’ve stayed before that was about $250 for the night all in — success!)
Search modals still reign supreme
AI and travel have always been closely linked. When OpenAI released Operator a few months back, it listed companies like Priceline and Tripadvisor as launch partners. Expedia was in the game even earlier, with one of the first integrations into ChatGPT by a major travel retailer in 2023. But their travel search hasn’t really changed:
Neither has Priceline’s. Sure, there is “Penny” in the bottom right, with the mystical starry iconography that seems to be the visual hammer for AI, but the same old flight search modals we’ve seen for decades take up the majority of the screen real estate.
I’ve looked at dozens of sites and it is all the same. (Try it yourself — proving me wrong may get you that dopamine you crave. 😉) No matter where you look, search modals still reign supreme, with some AI sprinkled in here or there. Chat options abound, but they are all optional, not the center of the show. That said, there are plenty of startups banking on a new paradigm — like Airial, who I met with a few weeks ago in San Francisco — but the main players are still keeping AI as a secondary channel.
On the Corporate OBT front, Concur Travel, still the largest global OBT in use today, announced their Joule product at their Fusion event last week. In its current iteration it is still very much billed as an AI assistant focused on expense, and wasn’t yet augmenting search. Other players from the usual TMCs to Navan all list some kind of AI capability, but none of it really sits in the search space, nor are they abandoning the traditional search modals to start workflows. I’ve seen news of Otto, part of Steve Singh’s Madrona group, but it is still in development/early beta.
Furthermore, any stats I can find on the topic point to AI being much more widely used in trip planning, not the actual booking process. Which makes sense, because vacationers are often going to new places and gathering inspiration from multiple sources, tasks ideally suited for generative AI. But the guy just looking for a hotel in NYC on his 100th trip? Not so much.
Most people still want to explore
Perhaps it is a controversial position, but I think people like searching for travel. And I’m not sure that will dramatically change, at least for a while.
As a TMC, one of the top pieces of traveler feedback we get is, “I found it cheaper at xyz.com”. We hear it all the time, and regardless of whether the traveler was looking at a real apple-to-apples comparison or not (not going down that rabbit hole for this article) the fact remains they searched somewhere else to begin with.
Sure, with AI we’ll add better personalization capabilities, we’ll summarize pages of content into a few bullets, drive better support workflows and, for some people, will give them a chat interface they may use from time to time. But I think most people will still want to explore to find what they want.
Why? I think people secretly enjoy it! It gives us all a bit of satisfaction in doing some real work — work that makes us feel like we’re getting the best option for the money — which gives the whole thing a bit more purpose.
Am I off-base? What do you think?