The great thing about disruptive and
dislocating times is that they create the space for entrepreneurs to do their
thing. It’s very hard to predict exactly what that thing will be, but you can
predict there's going to be a lot of activity and remarkable creativity.
And we are absolutely in that time.
There’s something going on in travel, a
confluence of factors that are going to define the next phase of innovation, entrepreneurship
and investment. Base-level, there is and will continue to be an addressable
market of trillions of dollars—travel’s growing
faster than global GDP and there’s no sign of this slacking off. Yes destinations
will fall in and out of favor, source markets will shift and we can’t ignore
the climate impact, but the bottom line is travelers will keep on traveling and
the great interconnectedness of global relations will not be put back in the
bottle.
I’m also seeing that travel’s historic
resistance to technology adoption is crumbling and that’s the change I’m most
excited about. The pandemic left the industry flat-footed, exposing a lack of
agility in the tech stacks of sellers, suppliers and intermediaries. Now the
C-suite understands the downside of having a tech stack that falls short of
what the circumstances demand. They are more willing to listen to technical
voices (including startups) and are taking a more future-minded, ready-to-invest
posture.
At the intersection of these two sets—a
massive and growing addressable market crossed with an appetite to invest in innovation—is
a world of opportunity for entrepreneurs to rethink the way things are being
done and to come forward with inventions that we can't yet imagine, systems to
alleviate the problems that have plagued the industry for decades.
I deliberately haven’t mentioned “generative
AI,” because we need clarity on the market context in which genAI has planted
its flag. That context involves more travelers and more technology. What genAI
is doing is unlocking a pace of experimentation, testing and trial that allows
good ideas to percolate, some of which will settle in a landscape large enough
to support billion-dollar businesses.
Buying in to new traveler paradigms
I’ve talked before about legacy technical debt
in travel and how we need better data because it doesn't matter how slick the
AI concept if the AI sits on a pile of horseshit, it's going to return some
pretty bad suggestions. I’ve talked up the opportunities for B2B entrepreneurs in response to this, doing the dull stuff, the data cleansers and
orchestrators, the builders of cloud-based PMSs and CRSs and middleware layers that
support new data formats optimized for genAI integration and interoperability.
I’ve argued that genAI is a massive threat to traditional search, because travelers are starting to use options other than legacy
keyword search at the top of their travel planning funnel.
So how does all this play out in a B2B2C
context? How far away are we from genAI agents being more capable than human
and online travel agents? My take is that it will come, sooner than we might
think, but we need the picks and shovels—infrastructure and data—to build a
solid foundation first. Here are a few examples of companies we’re excited to
work with in these emerging areas of the genAI transition:
- Bonafide, one of Thayer’s most recent
investments, is working on a new infrastructure layer using technology that
enables brands to manage the exchange of information needed to drive commerce
within the genAI ecosystem. It will enable brands to distribute accurate and
validated content to genAI apps, from a single source, at not only the
discovery but also the purchase stage. Brands will have ownership of and
control over the data and content, optimizing their visibility within the LLMs
powering the genAI apps and to be a proactive participant in purchase decisions.
- Outfit Labs is one of our early-stage B2B2C
investments. It’s looking at how it can get to know a traveler’s preferences
through proactive natural language interactions and permission-based access from
partners. This speaks directly to another emerging use case where agentic AI can
be better than traditional search for travelers by using natural language or
conversational queries. In understanding that a traveler is a serious swimmer, prefers
luxury hotels from her past travel history, and needs a hotel in LA for an
upcoming business trip, it can apply that context to the recommendations it returns and suggest hotels
with lap pools instead of family-friendly splash pools.
- Cruisebound is another B2C business we’ve
invested in, relatively recently. Cruise and genAI aren’t as disconnected as
some might think. The offline-to-online shift in cruise still has some way to
go, and for considered, high-ticket purchases like cruise, content plays a critical
role in purchase decisions. We see genAI as a foundational element in that
digital transformation for the cruise sector. By ensuring LLMs are able to
understand the rich cruise content on its site, Cruisebound is already on its
way to becoming a go-to source for agentic AI queries for cruise
recommendations.
Adaptation over personalization
Personalization has become a bit of a jargon-y
concept because traditional search is not set up to understand the granularity
and nuances of what a traveler wants and how that changes with context. Over
the past six months, I’ve seen across my portfolio that businesses are starting
to crack the code—the models are getting better because consumers are playing
with them, telling them what they like and what they don’t like. That’s helping
train the models to predict what customers want in a given context.
For start-ups and innovators playing in the
B2C space, it’s not going to be easy, but it’s not impossible. Yes, the
incumbents have had thousands of engineers working on AI for a while, they sit
on megatons of data, have supplier relationships in place, and they’re still
putting a third of net revenue back into acquiring more travelers via the
traditional channels of search and discovery. That flywheel is going to take a
while to exhaust itself, but it will begin to transform eventually.
In the meantime, the disruption and
dislocation of the incumbents’ search and discovery sweet spot is well under
way. The market is big enough for new niche and scale players across both B2C
and B2B channels, so don’t be surprised to see some of them take hold in new
and exciting ways. No one really knows where these changes are going to land,
but everyone should know by now that it’s going to take a shift in data and
infrastructure for these changes to happen in the first place.