As an experienced investor in early-stage travel tech
companies, I am as excited as anyone that the travel industry value chain is on
the verge of a truly seismic reinvention, driven by generative AI (genAI). The
intelligence revolution we’re going through is happening at a pace that makes
the industrial and digital revolutions seem lethargic, with an unprecedented
rate of user adoption happening at scale across what is becoming a global
platform of platforms.
The investment community will tell you that there are use
cases - and pitch decks - raining down from different geographies, looking at
different verticals, targeting different business cases, all “powered by genAI.”
Any founders or investors getting it right could be in for a home run.
But my enthusiasm as an investor is kept in check by my
experience as an investor. False
positives bloom during the early days of revolutionary times and while
generational companies are germinating as we speak, the landscape can be tricky
for investors and entrepreneurs.
With my exclusive focus on travel technology, I can offer
some perspective regarding specific B2B and B2C business cases and investment
potential for genAI startups and scaleups in travel and hospitality.
Here’s some advice based on my observations thus far.
Pump the brakes
When the latest pitch for, say, a gen-AI-powered trip
planning startup for off-the-beaten track multi-day tours lands in my inbox, I know
I’ve seen this movie before. I can fast-forward to the set-piece scenes where
the pitch talks about the addressable market, how the product is a perfect
match with Gen Z trends and shows a mocked-up customer journey ending in a seamlessly
purchased high-ticket sale.
But I know that in many cases there won’t be a scene where
the founders discuss in detail customer acquisition costs or one where the
go-to-market strategy is timetabled month-by-month or a detailed proposal for a
post-seed pre-Series A organizational structure for the business.
The oversupply of startups when a new technology emerges is
something we’ve seen before.
So, the first piece of advice I’d like to share is: pump the
brakes. Slow down, step outside the generative AI bubble if you are in it, and,
if anything, get back to basics.
For consumer-facing startups, most of the questions the
investor wants answered are not about the genAI angle. The pitch needs to be
focused on a novel solution to a clear pain point, unit economics, go-to-market
strategies and domain expertise of the founding team, rather than architecture,
latency and data flywheel improvement rates.
The B2C startups in the Thayer portfolio have passed the
screen test. Point.me is tackling head-on a significant traveler headache –
how can I use the points I’ve earned on my credit card to buy the flight I want,
now. Yes, AI is optimizing its core search, find, book and pay features, from
where it can upsell higher-margin subscriptions and concierge services, but the
story is not about AI. It’s about a
complex solution to a real problem.
Cruisebound offers another example of a team focused on
solving a real problem while leveraging AI to drive efficiencies and, in time,
scalable micro-personalization. Other
good ones are coming, and they all share the common thread of creative solutions
to real problems and strategies that speak to the almost insurmountable
challenge of economically acquiring customers at scale. AI is almost always part of the story but
rarely ever the story itself.
B2C startups in travel and hospitality are trying to secure
funding to succeed in one of the most competitive e-commerce verticals, with
some well-funded incumbents. Investors understand the risk/reward profile to
these types of investment is different.
Having said that, I believe that genAI and the intelligence
revolution have the potential to fundamentally change every touchpoint on the
travel value chain, within which there will be some B2C winners. My take is
that these winners will be highly targeted on a specific niche within travel
and hospitality, sensitive to the ways in which genAI can deliver added value
to its customer base and focused on leveraging AI to drive down acquisition
costs and pump up customer lifetime value.
How they accomplish that is the essence of the revolution and something
we look for in every pitch.
Re-tune the engine
B2B has always been a different proposition from B2C, and it
is where I am finding the most interesting genAI-enabled investment
opportunities. B2B has been the Thayer Ventures sweet spot over the years and
it’s where our domain expertise lies. Our entry in the Phocuswright Travel
StartUps Interactive Database shows that 78% of our investments are in B2B.
AI has informed many of these B2B pitches over the years,
but there’s a difference emerging with genAI – the C-suite is more interested
now. Senior execs are starting to realize that the promise and potential of AI,
such as micro-personalization at scale, is being hampered by technical debt,
and these execs are now looking at ways to address this barrier to growth.
Technical debt is a headwind for many sellers and suppliers,
but it is also responsible for many consumer-facing traveler pain points. For
investors, backing companies that are addressing the tech debt problem is an
opportunity. The debt exists because travel went through heavy phases of
digitization back in the early and mid 90s – online travel agency extranets, global
distribution systems, loyalty schemes and frequent flyer programs, on-premise property
management systems and central reservation systems. We’ve been updating and
fixing and working around ever since without rewriting the source code.
The result is that there is 30 years of industrial junk
knocking around parts of our industry, which is compromising the data that we
have.
GenAI lives on top of an historic tech stack that is loaded
with technical debt and which might be feeding the AI with data that is
inconsistent, outdated, biased. When the data is poor, so it the outcome.
And for B2B startups, there’s a massive opportunity for data
cleansing or the even more unglamorous term, data scrubbing. These businesses
are the picks, axes and shovels of the genAI goldrush.
Thayer’s B2B portfolio includes startups who have come up
with a solid business plan in light of the industry’s tech debt. Mews is one of
our highest profile investments, and its cloud-native PMS, launched in 2012, has
no technical debt as such. Its 5,000 hotel users are genAI-ready, and Mews is
also helping hotels with a legacy PMS tech debt to switch over and take
advantage of genAI. Deal Engine is
providing API-delivered, AI-enabled services that extract, transcribe and
rationalize data so airline customers can improve services.
Elsewhere, Lighthouse (formerly OTA Insight) is buying
businesses to deliver what it calls “the next generation commercial platform,”
using AI to analyze data pulled in from internal and third-party systems and
present it to hoteliers as actionable business intelligence across many
functions.
Hospitality, perhaps the travel sector with the highest
degree of tech debt, needs data cleansers more than most. Without clean data most of the B2C business
cases fall over.
Assess the passing opportunity
Having said earlier
on that startups need to pump the brakes, I’m now adding in time sensitivity to
the equation.
While entrepreneurs
pitch and investors ponder the theory and practice of genAI, modern travelers
are decoupling themselves from what we are currently offering.
Instead, travelers
expect hyper-personalization, location-based services, seamless interactions at
every touchpoint. GenAI is the delivery mechanism for these features (once the tech
debt has been addressed)
Meanwhile, GenZ in
particular is being trained by electronic and software giants such as Apple,
Samsung and Google to expect genAI-enabled features in their smartphones and
tablets. If they visit an OTA for advice or inspiration, they will engage with
intelligent chatbots powered by AI and read AI-generated summaries of reviews.
NLX, one of our
recent investments, taps into this zeitgeist for intelligent, conversational
chatbots. It leverages genAI to enhance its no-code platform approach, building
on the products it has launched using traditional AI. The platform integrates search,
customer service and commerce in one conversational experience, and with no
code, implementation for enterprises is straightforward.
The demand for
intelligent, conversational chatbots is a sign that the industry should rethink
“digital natives” and face up to a generation of genAI native travelers.
Remembering of course that genAI is not necessarily a Gen Z thing – there are
business and use cases across all demographics, including the higher-spending millennials,
Xers and boomers.
The advice here to
entrepreneurs is be aware that while this is genAI’s time, speed to market is
not necessarily what investors want. A structured go-to-market approach is a
better bet than promising to deliver something to an artificially tight
deadline which ends up not working.
It’s also worth
noting that the usual phases of investment – pre-seed to seed to Series A and
beyond - is still the investor community’s way of working. No-one is going to
dive straight in and offer you Series A amounts of money just because genAI is
the dominant tech of our time and your pitch-deck has genAI on every slide.
(Not generated by AI) Summary
- It’s tempting for entrepreneurs to get carried away by the
genAI frenzy, but a solid business plan and development roadmap is still going
to get you noticed as much as what you claim your startup can do.
- Data, data, data. Because without decent data, the promise
of genAI will be unfulfilled. B2B entrepreneurs have a window of opportunity to
develop solutions to help large-scale enterprises clean up their data, so that
these large-scale enterprises can leverage AI at scale.
- Even though its time has come, I do not think it’s
now-or-never for genAI startups. Entrepreneurs need to watch what’s happening
across genAI market and how the investment landscape is developing and think seriously
about the timing of their approach to investors.
- Did I mention data?
Update
Just as I was about the push the “send” button on this
article, Amadeus
released some research into generative AI in travel from the industry’s
perspective. It quantifies the scale of the disconnect I’ve been talking about –
tech debt is holding the industry back because genAI needs better data – and
hints at the potential addressable market for data cleansers.
The study found that more than half (56%) of the 300+ senior
travel tech execs think their technology infrastructure needs work before the
deployment of Generative AI.
At the same time four-in-five of them see generative AI as a
priority over the next year.
You do the math.