Markets by Grant | TravelTech Platforms

15 min readJul 6, 2022

Hello and welcome back to Markets by Grant.

Let me tell you about a recent trip I planned.

I like to think that my travel planning process is fairly orderly and sequential. It’s not easy, nor is it perfect, but if my mom were to read this, I’d like to think she’d be proud (hi mom).

My process roughly resembles a lonely, labor-intensive game of chutes and ladders. I start with where I want to go (duh) and book transport to get there. Then I slowly flesh things out, starting with the big things (lodging) and moving down to the small things (food and activities). Pretty typical stuff.

Within each of these categories, I like to think my process is also pretty typical. I find inspiration, figure out what’s actually out there, validate that it’s good, and then book it.

Mind you, this is a very leaky funnel. If I were to map out the average consumer journey, it might get abruptly cut off after any of these steps–as most folks aren’t typically inspired by somewhere to go and book a whole trip around it in the same year, let alone the same sitting. In this case, I didn’t permit myself to churn, because I had a woman to impress:

What I never realize, of course, is that all these categories are interrelated. If I’m going to the beach one day, I can’t book a nice lunch in the city that same day. And what if my ideal hotel is too far from that hiking trail I really want to try? Long story short, things never end up as sequential and orderly as I think, and I invariably run into the ‘versioning’ issue — of making updates and tracking their downstream effects through the whole trip. Now you see what I meant when I said chutes and ladders:

…but what I really forget is that not only am I constantly navigating between categories of activities and across different process steps of figuring them out, but I also have to do so over countless different people, places, and sources.

A real life example: My friend’s friend told me about this restaurant near the beach. This restaurant has a 3.9 on TripAdvisor, a 4.3 on Google, a 4.8 on Yelp, it has one Michelin star, and it was featured on the Infatuation 4 years ago. Some blogger said locals go there. Google Street View and user-submitted photos look a bit lackluster. I can’t seem to find their menu to make sure they have vegetarian options. Looks like I can book it on Resy, but only for the patio at 5 or 10pm, and it’s going to be 50 degrees and rainy according to my weather app. Maybe it’s not the best beach day after all.

You get the gist:

…and the only thing synthesizing all of this information, gathered from all these sources, is my highly unreliable brain, at 10pm on a Tuesday after a long day of work.

But this is all old news, right? I mean, any average traveler could make a fairly decent market map of today’s traveltech landscape:

So what’s my point? The process sucks, and everyone knows it. And to fix these pain points, I can see two alternate visions of the future:

I stop overthinking my travel planning (unlikely)
New travel platforms solve problems of social proof, transparency, and excessive verticalization

Before I continue, a quick interjection.

From here on out, I’ll be talking about traveltech platforms. By definition, these platforms operate a software layer which connects suppliers of products and services (supply side) with customers of these products and services (demand side (in this case, consumers)). Platforms provide the demand side with supplier information, and they provide the supply side with customer lead generation. As part of the deal, both sides get valuable infrastructure and tools as well. Sometimes suppliers opt-in and sometimes they don’t. And sometimes suppliers are the consumers themselves, adding information to a growing ecosystem.

Why platforms? Because they’re the heart of the traveltech industry, in my opinion, and also because I’ve got to draw the line somewhere. So, in my market map above, we’ll keep the platforms and OTAs, and ditch everything else.

Onward.

To explain what I mean by number two above, I’ll use the same visualization.

The problem with the X-axis (top) of my process flow/market map is the number of intervening steps. Most planners like me can’t get sufficient information or develop sufficient trust in one platform alone, so we feel the need to cross-validate across multiple platforms — taking a convoluted route from inspiration to booking.

If we either…

Had enough information in one place to make a decision, or
Trusted the source enough to feel more confidence and less need to analyze and validate

…we’d stay in the same platform more and convert into action/booking more.

Thus, trust and transparency are hugely important drivers for any travel platform hoping to resonate with, retain, and convert users.

The problem with the Y-axis (left side) of my process flow/market map is the excessive verticalization. For travel planning use cases like mine, the specialization of things to do across platforms (ie: restaurants, activities, lodging, etc) is a hindrance, not an enabler. Platforms which integrate these siloed verticals to curate more personalized, all-encompassing experiences around the user also have a better chance of resonating, retaining, and converting.

I’ll admit that this is a more controversial point of view. Platforms have coalesced around specific service areas for a reason: because it’s easier to build concentration of technical and operational capabilities around subject matter than customers, and it’s easier to market as the one-stop-shop for verticals than the one-stop-shop for specific customer profiles. And so it makes sense that this has been the direction to date, but this direction has left us with today’s pain points. Improving AI-driven personalization capabilities, coupled with an increasingly crowded market, coupled with generational trends toward experiential rather than task-organized travel, speak to the promise of bringing more integrated experiences to the user.

In short, here’s my vision:

And how do we do all this? By understanding who your user is. To achieve trust and create transparency in the right places, you need to find out the user’s values and desires. And the same is true should you want to curate integrated experiences which will resonate with your user.

And what do you get when you throw together a bunch of buzzwords like ‘values’, ‘desires’, ‘trust’, ‘transparency’, ‘understand your user’, and ‘integration’ in one giant pot? You get a giant pot of ‘easy for you to say, Grant’. And so I’d like to anchor these positions with two proxies to understanding your user:

The first is learning that person’s non-negotiables — ie: not just what they like, or want, but what they require. Your needs for an experience are more important than wants, and they set the predictive baseline of your enjoyment of it (see references section). This information ranges from easy, filter-able stuff like vegan options or price point, to characteristics which require psychographic profiling, like sensitivity to slow service. If your platform doesn’t confirm that the restaurant you recommended is fast-casual with vegan options, I’ll have to go elsewhere to confirm.

The second is learning who that person admires. This is the social proof hack to build and exploit trust. The classic question: is a consumer more likely to trust a 4.7 rating on Google or a rave review from a trusted friend or authority? Intuitively the answer is the latter. First, today’s ‘out of five’ rating system doesn’t convey sufficient nuance to indicate true promise of an opportunity, and all detailed reviews aggregating to that 4.7 average are anecdotal stories, biased towards overly positive or negative, from people very unlikely to have similar tastes to you. The 5-star system is fine when products and services have objective characteristics (these pants are blue), but lose their applicability when faced with subjective, experiential categories. Thus, the value of one friend, favorite influencer, authority figure, or other trusted/admired individual is more powerful in generating conversion than thousands of aggregated reviews (see references section).

And, of course, combining ‘needs’ data and ‘admiration/trust’ data with historical data on your reactions to past experiences is table stakes.

So, that’s my thesis on the space.

Oh, you want me to summarize it concisely? Fine:

I like…travel platforms which leverage social proof and optimize information transparency to integrate experiences around the individual desires and values of today’s traveler.

How does that look manifested as a startup in the market today?

4.5 words: AI-enabled experience discovery platforms. And there are actually quite a few of them.

But before I introduce the landscape of startups which are tackling this space using what I view to be compelling strategies, I’d like to introduce a few paradoxes which have been on my mind as I’ve met with founders in this space. These will inform my evaluation of the startup landscape I analyze in my next piece.

The Paradoxes

Traveling Here vs There

What is travel anyway? How are we defining the boundaries? Do we have to journey a certain distance from home for our trip to count as travel? To another continent? Country? State? City? Neighborhood? Block?

In my opinion, it’s all of the above. The defining shared characteristic of TravelTech platforms isn’t the ‘where’, it’s the capability of surfacing the ‘what’ to do and ‘how’ to do it. As an NYC resident who uses, let’s say, Welcome, to explore things to do in my neighborhood, and an NYC resident who uses Welcome to explore things to do in Lisbon — the place data is different, my persona may be slightly different, but the fundamental use case is the same. It’s about inspiration, discovery, validation, and booking of things to do and how to do them.

So, hearkening back to my diagrams in the section above, I’d wager that the process is pretty much the same regardless of whether I’m planning a trip to Lisbon or a long weekend with friends here in town. Long story short, I don’t really buy the huge distinction between ‘here’ travel and ‘there’ travel in terms of the customer journey.

With that said, I am ‘here’ much more often than I’m ‘there’. I could have been toying with the idea of this trip to Lisbon for years, and, for whatever reason, haven’t pulled the trigger. That restaurant I’ve been wanting to try in Lisbon has had a seat ‘open’ for me for years — and I’m not making any friends with TravelTech platforms by starting and stopping my trip planning to Lisbon every year. Meanwhile, here in NYC, if there’s a restaurant I’ve been meaning to try, and for whatever reason, tonight isn’t the right night, I’m going to have way more ‘at bats’. If not today, maybe tomorrow, or next week, and very likely, within the year. For Lisbon, it’ll have to be 2023. On the other hand, I may be more open to balling out during my trip to Lisbon, given the expense and commitment of getting there.

Put another way, the ‘there’ travel model clearly works, but its efficiency metrics, from a conversion standpoint, aren’t quite as great. Ie: if I tell 100 New Yorkers about 10 restaurants to try in New York (or 100 Lisboners about 10 restaurants in Lisbon), chances are, we’re going to get a fair number of conversions (which these restaurants might be willing to pay for). If I tell 100 New Yorkers about 10 restaurants to try in Lisbon (or vice versa), the conversion rate is going to be way lower. So, the cost of reaching these 100 people better be very cheap, or the restaurant better be willing to pay me a load of money for the low conversion rate. And the restaurant will likely only be willing to shell out the major dough if the customer is willing to do so as well. Long story short, ‘there’ travel lends itself to higher ticket sizes for lower conversion rates, and ‘here’ travel lends itself to higher conversion rates at lower price points.

Last argument, and then I’ll let this one go. Leisure travel, in its many forms, is a highly discretionary spend category which is tied closely to overall consumer confidence and economic health. As one founder told me, it’s very easy to cancel a trip. In fact, it often feels like ‘the right thing to do, given everything going on’. Relative to your personal wealth, this typically starts with the bigger ticket items, and trickles downstream to that long weekend trip upstate, and that fancy dinner downtown, and tickets to that comedy show. Higher income travelers are much more resilient and loyal, but all travelers cut spending in lean times. Regardless of who you are, this rules strongly in favor of ‘here’ travel being the less cyclical, more stable category.

Traveling to Travel Platforms

Think about all the travel websites, apps, publications, and agencies out there. In order to travel, we also have to travel to reach travel platforms, scattered across the internet, our phones, and the world around us. Do we really have the appetite for another? Do we have the mindshare to keep one of them top of mind? Can it reasonably beat all the others to become our ‘go-to’?

Generally speaking, I’m more bullish on content which ‘comes to you’ — ie: the ‘90–10 rule’ from the movie Hitch. These apps should come 90% of the way to you, with a compelling, personalized value proposition inspired by your needs and pre-validated by social proof — and you should only have to convert with 10% of the effort. How tech ‘comes to you’ has evolved significantly over the past few years, and I view the latest, most direct evolution to be messaging-native apps, which integrate with your texts, group chats, etc, and provide their most important features there, natively in your messages.

The tradeoffs, of course, are that 1) this is hard to do, and 2) silver platters tend to be less interactive.

If a messaging-native app which comes 90% of the way to you and gives you a personalized recommendation gets it wrong, it misses out on that opportunity to convert. If a travel agency gets you to sit down and sink 90% of the work into researching using its platform, through its wealth of features and capabilities, it might not convert you right away, but it’s less likely to lose you outright.

Smooth vs Lumpy Platforms

To what extent should the providers on the supply side of a travel platform be empowered to differentiate themselves? I’ll admit that I made up the terms ‘Smooth’ and ‘Lumpy’ in this context, and I’ll be the first to acknowledge I probably could have come up with better words. But whatever, too late to press backspace now. And just incase your mind went elsewhere, smooth and lumpy are meant to describe the appearance of a platform to its demand side:

A smooth platform is one place where customers can easily navigate across a standardized set of supplier options. In a smooth platform, suppliers’ products and services are represented as commodities, with little uniqueness. These platforms tend to cater towards making life easier for the demand side, and helping them convert. In a smooth platform, the supply side needs the business, and is okay with sacrificing some of their uniqueness (and margin) for the sake of leads. Tripadvisor is a good example of a smooth platform.

In the lumpiest platforms, you might not even know you’re using a platform. It’s there, hosting and enabling its supply side to run their businesses, but it may ‘white-label’ itself out of the demand side’s view. Shopify is an example of a lumpy platform, and Thatch is a good example of a Shopify-equivalent for travel creators. Lumpy platforms’ customers are really its supply side, who have ample demand and don’t need leads. Thus, they focus more on supply side infrastructure and less on demand generation.

Broad vs Narrow Platforms

As a platform, you can either create narrowness around your customer or your product/service area. As I mentioned earlier, in the case of traveltech, we’ve largely seen startups narrow in on service area to focus on building concentrated technical and operational competencies, but in this piece, I’m suggesting that startups instead build concentrated competencies around their customer. So when I say ‘narrow’ vs ‘broad’, I’m referring to the demand side of the platform — ie: are we serving a specific customer segment, or anyone?

The tradeoffs here are clear. By serving specific people, I’m immediately missing out on the opportunity to serve everybody else — but if I serve everybody, I may not be serving specific people well enough. Narrower platforms are often able to extract more loyalty and value from their target customers, which is important when markets are competitive or acquiring customers is tough. Broader platforms may struggle with less revenue per user, but that doesn’t matter too much if customer acquisition costs are low and its products apply to everyone.

Content Quality vs Scalability

One last question before I leave you to tend to the headache I’ve caused. Most travel platforms, in some way, are content companies. They know that in order for you to convert, you’ll need to see pictures, videos, itineraries, menus, reviews, etc. The sexier the content looks, the more likely you are to be inspired by the experience, and engage with the content or convert to booking. So every travel tech company both wants their content to be ample in volume and as attractive as possible.

In order to populate their platforms with content at scale, many players crowdsource from their users. Google Maps is an example of this. It starts with a barebones template for a place, which over time is filled with user-generated photos, reviews, tips, and other metadata. Looking through a Google Maps page for a restaurant, you’ll see a smattering of incorrect opening hours, misspelled complaints, nondescript praise, and smudgy photos of half-eaten meals.

On the other side of the spectrum, you may have content companies like Conde Nast Traveler. They cover very few places within very select locations, but do so with a high degree of production value, richness, and polish. Their in-house content is certainly more likely to get a reader to convert than a Google review, but their content covers a fraction of the places, and is seen by a fraction of the eyes. And if Conde Nast Traveler tried to cover every place Google Maps does, its content would be woefully insufficient or horrifically expensive.

This sums it up. Everyone wants great content, but in-house, well-produced, fully-audited content is not scalable. Is DIY-ing it worth it? In my opinion, the answer is yes if one side of your platform is much narrower than the other.

Broad on Both Sides: If your platform hosts a million products/services for a million people, in-house content is a lost cause.
Narrow on Demand Side: If your platform caters to a specific audience, you may need to curate more focused content to resonate with them.
Narrow on Supply Side: If your supply side is consolidated, creating in-house content isn’t such a herculean effort.
Narrow on Both Sides: If both your supply side and your demand side are narrow, you might not need to have a platform in the first place.

Phew, I’ll leave it there for now. In my upcoming piece, I’ll dive into the numbers and the startup landscape, analyzed with the lenses of the paradoxes and theses raised here. Until then, feel free to roast me in the comments section.

Questions/feedback/ideas? HMU at grant.demeter@av.vc

Shout-Outs

Thanks to…
Ron Levin of the Yard Ventures (Prev. TravelPerk)
John Peterson of The Nudge
Mike Rosenthal of It’s Good
…for your help and thought partnership

References

Customer experience, needs, and wants:

5-star reviews and social proof:

Travel trends and interesting reads: