Markets by Grant | AI Experience Discovery

17 min readAug 3, 2022

Hello and Welcome back to markets by Grant.

In my last piece, I raised some problems and paradoxes about traveltech platforms today, and came up with a thesis of…

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

…And I said that that thesis met its match in a startup category called: AI-enabled experience discovery platforms.

But I didn’t bother to explain what those were or how they work. That’s what this piece will be about. I’ll top-down it, starting with the market and moving into the landscape. Buckle up, it’s a long one.

First, let me define an AI experience discovery platform, so we have a rough idea of what we’re talking about here. These are platforms which use AI to learn their customer’s desires and values to recommend experiences for them to embark upon.

Ie: a platform knows it’s a rainy Sunday, I’m in NYC this weekend, I’m vegan, an influencer I like visited restaurant X, I like cocktails, and I like modern art.
So it recommends I visit this restaurant, which happens to be near MoMA, and then visit the museum, and then hit a bar for afternoon cocktails. They’re good rainy Sunday activities, which fit my specific interests, my current geography, and which are pre-validated by trusted friends and influencers. And note that the platform recommends integrated experiences — not just restaurants, or just museums. The content is organized around me, not around itself.

If that explanation just made you more confused, sit tight, I’ll have tangible examples of startups doing this down below. But first, the market.

The Market

How do we size a market which is so boundless? It applies to restaurants, bars, venues, transportation, lodging, and anything else a consumer might do for fun. And it could be for any place on earth they happen to find themselves in.

Fortunately, I’ve had the luxury of looking at a bunch of pitch decks in this space. Unfortunately, none of them have really attempted to size their market. The closest I have is a deck which says “the experience economy is worth trillions” — so I guess I’ll start there.

The Experience Economy

What is the experience economy? How big is it?

Looks like the term made its debut in 1998 on everyone’s favorite periodical: the Harvard Business Review. Back then, it was the idea that a highly differentiated customer experience could generate disproportionate willingness-to-pay for a business, in addition to the product or service provided — to the point where the experience could be considered a product offering itself.

Straight from the journal article, here’s an exciting screenshot which seeks to define “experiences” as an economic offering, compared with the traditional products and services a business may offer:

Key words which stand out to me here are “memorable”, “personal”, and “sensations”. The article goes on to suggest that true customer lifetime value is created not at the utility level, but at the emotional level.

Not to toot my own horn or anything, but it’s that key consumer insight which has been driving most of my thesis development to date (I’m just ~24 years late). And, in a way, this analysis of products which enable exploration of meaningful personal experiences is a natural progression from my previous pieces. For more on this theme, take a look at my four-part series on the “Journey to the Center of the Consumer”.

Anyway,

Since 1998, the term “experience economy” has been run through a game of telephone, and now seems to be used primarily to describe businesses whose primary good/service happens to be experiential in nature — ie: concerts, museums, fancy restaurants, etc. It’s somewhat subtle, but it’s not the same thing. At least in part, the slow warp of meaning is due to what everyone has hit us over the head with for years: Millennials value experiences over products.

I’ll belabor the point with an example, and then we can move on with our lives.

Somewhere, on a Monday morning, in a boardroom, in the 2010’s, someone barges in and says “guys, we’re not a bank anymore, we’re a cafe which offers banking services”. “We’re going to sell coffee with oat milk AND checking accounts”. “This is what the millennials are crying out for”. “I saw this in a dream last night. Promote me to managing director immediately”. And thus, the confluence of “experience-first” banks/cafes/community hubs/whatevers — where you stroll in, looking for a good time, and stumble out with a cappuccino, a debit card, and a refinanced loan.

Like it or not, this is an example of the experience economy. But the degree to which it’s successful as an offering is determined by a leading or lagging indicator:

The leading indicator is “admission”. If the Capital One Cafe is able to charge admission for its cafes, you’ll know people consider the experience of being there a valuable experiential offering. This could either be literal admission fees at the door, which the authors of the article suggest ad nauseam for every industry and use case possible (I don’t buy it) — or it could be more of a subtle admission, such as the length of travel required to arrive at the location, or purchase of a coffee when the cafe is not the most convenient or best coffee joint around.
The lagging indicator is “memorabilia”. If someone picks up or creates some form of memorabilia to try to immortalize or promote their experience, then you’ve delivered a successful experiential offering. This could be something super literal, like buying a t-shirt, or it could be picking up a business card on your way out, or — perhaps most relevantly here — it could be the act of photographing, posting, or otherwise sharing about the experience.

This may feel a little off topic, but I think these are actually meaningful performance indicators for AI-enabled experience discovery startups. Experiences are their product inventory, and the degree to which an experience is predictably salient and meaningful is a function of quantifiable “admission” and “memorabilia” transactions. Who is willing to pay admission for this experience? How many people and how much each?

The same goes for memorabilia. When the Capital One/Santander Cafe shows up in one of these AI-enabled experience discovery platforms with user-submitted photos, described as “super cute neighborhood coffee shop with the BEST matcha lattes, and a cool speakeasy in the back which leads to a cubicled office where you can talk about interest rates with nametagged barista-bankers” — then you’ll know they made it as experience-first.

So let’s keep these performance indicators back of mind as we evaluate startups in this space.

Back to the Market

Let’s get this over with. For our purposes, my definition of the experience economy is today’s version — it’s leisure activities people engage in primarily for the experience of enjoying them. And this closely mirrors modern definitions of the global leisure travel market, which encapsulates transportation, lodging, dining, events, and a small category called “other” (who knows). Long story short, I took you on a bit of a round trip there, and we’re now back to the top-down market sizing I promised:

I created the below chart based on a triangulation of 4 or 5 market size estimates per market slice, some of which varied quite a bit (see references section). Some quick insights behind the numbers:

The total leisure travel market fell off a cliff following COVID, and its underlying industries were hit as well, although not quite as badly.
Since, things have almost bounced back, but a bit differently. Domestic travel has been taking off (for obvious reasons), and higher-income travelers have led the return to the traditional travel industry.
It’s not easily visible, but the “online” portion of the Travel Industry comprises an impressive ~⅔ of the total industry in terms of transaction volume, and this amount is projected to reach ~75% by the end of the projection period.
As you absorb the info, note that it is not stacked, but each category is a segment of the one above. And here’s the chart. Wildest of all, these numbers are in billions of dollars:

So the total quantification of all these things is: very big.

But, spoiler: I don’t think a top-down sizing of this market really makes sense for the startups I’ll discuss today, because they don’t really monetize on the transaction volume in this “experience economy”.

A better way to size the market would be based on number of users and actual monetization strategy. I’ve seen 3 used in this market:

Classic freemium app pricing — ie: pay for special extra features/usage.
Content-style advertising/promotion revenues — ie: the “promoted” box on TripAdvisor, or the ads lining the right side of your screen, which monetize on clicks and views.
Take rate-based booking revenues — ie: OTAs or event promoters taking a commission percentage of the sales driven through the platform.

As you might imagine, each of these three monetization strategies depend on the maturity of the market and of your product. In nascent spaces like this, most start off with freemium until they reach a bulk of users and approach product-market fit, after which they turn on advertising/promotions (now that they have plenty of users and know them fairly well), and then start to diversify toward take rate-based relationships with suppliers (once they’re big and stable enough to negotiate with them).

So, given the nascency here, most of the startups I’ve met with are still on step 1, the freemium pricing, and are starting to experiment with 2 and 3. Most have sized their initial SAMs with this model in the high tens of billions.

An Argument With Myself

Before I dive into the landscape, I’ll introduce and respond to a few of the perceived challenges in this space:

Complaint 1: Network Effects: Network effects are against you until they’re for you. If you have an experience discovery platform with any social aspect or level of crowdsourced content, nobody is going to want to use it until everybody does. This means that CAC might be pretty high for a while, and product-market fit is a function of reaching critical mass.

I think this is true if a platform’s customer base is as broad as Google Maps, TripAdvisor, Airbnb etc. But I find platforms with some degree of narrowness on the supply or demand side to be most compelling in this space. These have less of a hurdle, in terms of quantity of participants, to reaching scale–and more importantly, they add value in product-led ways, other than in network effects/information symmetry alone. The last thing I’ll say is that these startups are just as much about local experiences as traditional travel experiences, which also makes community-building easier, CAC lower, and overall market resiliency higher.

Complaint 2: Funding Environment: I said “nascency” above, but traveltech OGs know that this experience discovery model has been tried before. This wave of startups is only the latest, and while I view them to be the best equipped, they’ve got to succeed where others have failed.

The relatively high CAC and failure of past generations of this business model means that whoever funds startups like this has to be in it for the long haul, with plenty of cash. And it also means you have to see this startup as potentially valuable enough to be worth the well-known risks. And funding is tough enough today as it is.

This all makes for a challenging spot to be in as a founder. The VCs I’ve spoken with in the space are all behind the AI-enabled experience discovery thesis, stating “this is definitely a problem which needs to be solved–but people have been trying it for a long time and I don’t think we’ve landed on a model which works yet, and I don’t want to be the one to put my neck out there for another high-risk consumer bet, especially in this funding environment”. Longwinded, but you get the point.

To each his/her own. It’s not B2B SaaS, that’s for sure, but I’m not sure that thoughtful stabs at this model have to be so capital-intensive and risky for their return. These days, AI/ML is more consumerized and buildable and time to value is shorter. Consumer preferences have continued to crystallize toward these solutions, and the “staycation” model has taken off following COVID. And as mentioned before, most startups I’m looking at aren’t trying to be “Everything Everywhere All at Once” (Haven’t seen it yet, heard it’s very good), but are more thoughtful about initial scope and areas of focus. So, while it’s still too early to tell whether these startups are seaworthy, I do believe the winds have changed.

Complaint 3: Incumbents: The classic question, “if we’ve known about this idea for a long time, and if it is so promising, and if there are big dogs with enough money and scale to go after it — why haven’t they?” “…And if, for whatever reason, the big incumbents miss the first boat, they’ll just build/buy competitively if a new entrant has any success.”

Grant’s response: In general, I don’t buy the “if it’s so good, why haven’t others done this before” question. Companies which set out with a business model, and built giant momentum around it, shouldn’t be expected to pursue every adjacent area of promise, at the expense of focus and scale.
And a more specific point is that the large traveltech incumbents, like Expedia Group, have evolved from big tech players with a strategic innovation focus, into large holding companies with a strategic acquisition focus. They’ve built acquisition engines and organizational muscles around them, and each purchase becomes more value-accretive with the increasing cross-sell and operational synergy opportunities of the portfolio approach. Long story short, I view this as a very favorable market to play in, as the big incumbents are there, they’re not aggressive downmarket, and not only do they acquire often, they need to acquire often to satisfy their shareholders. They aren’t the only potential acquirer persona here (see: Big Tech and Social), but I do think they are one of the more probable ones.

The Landscape

Ok, now without further ado, I’ll break down my view of some of the startups in this landscape, and I’ll do so with the lens of the “paradoxes” I raised in my last piece. For a quick refresher on these dimensions, give it a quick read.

And for a quick-er refresher on these dimensions, see below:

In short, I see each as a continuum which describes a relative characteristic of a startup playing in this space. Perhaps you see where I’m going with this…but rather than plotting every startup I’ve seen in this space on these continua, I’ll summarize them into four easy categories. Each represents a distinct business model/angle I’ve seen approaching this space:

Admittedly, these are pretty cryptic descriptions of what they are, but I figured I would go step-wise and give folks time for this to sink in prior to hitting them with the full eye chart. So, you know, let it sink in.

If you can read this at 75% zoom, you definitely have 20–20 vision:

I’ve added examples under each category to help contextualize the distinct approach of each business model. There’s plenty of variation and uniqueness within each category (and some are certainly more crowded than others), but I’ve found over time that most consumer-facing AI-enabled experience discovery platforms fit one of these profiles. And I guess it’s ok to have just four categories, given the specificity of the space I’m talking about. Anyway, let’s move into a study of each.

Note that many of the startups playing here are still in stealth, so out of respect, I’ve omitted a bunch of information. And, recognizing that this is becoming one of my longer pieces, I’ll just highlight a few key advantages and risks for each.

Content “Nudgers”

I haven’t quite seen anyone else doing what the Nudge does, and so this category is kind of a namesake. The Nudge is part native SMS texting app, part content company. It has a relatively narrow audience, so it doesn’t need to do a huge amount of AI-driven personalization, but it could.

What I like:

The product is for millennial, female, city-dwellers, and that’s who all the content and recommendations are tailored for. A large enough market, and a targeted enough starting place to easily curate more applicable, resonant experiences. Next, the Nudge is going to expand into the young parents market segment, for which it might have a totally different product shell. I’m a big fan of this segment-by-segment approach.
The product comes all the way to the user via SMS, and not aggressively often — which is a recipe for high engagement and very low churn, which are true for the business.
The content is sourced from users, then curated/audited/produced in-house — a relatively scalable play given the immersive richness of the content delivered.

What scares me:

Commitment
And also, it’s a very “consumer-y” market. If the content or brand falls out of favor/becomes too mainstream, there’s a risk of massive exodus. But hey, that’s consumer.

Messaging “Buddies”

Subtitle: Another day, another startup

Remember when I raised my second complaint, about how many of these models have been tried and failed before? This category is a good example of that. Everyone tried it: Google, Facebook, and even Microsoft all launched and then sunsetted messaging-native, social-enabled experience discovery and booking assistants. Both Ethan and Magic made it to number one on Product Hunt, and experienced explosive hype and usage. And Cloe, of note, tried to do exactly the same thing which Ask Alex is trying to do now.

In horror movies, the bad guy picks off the group one by one, until there’s just one person left. And in these movies, that last person usually escapes the bad guy and lives to tell the tale. The question is: is Ask Alex that guy?

It’s early, but I think they have a reasonable chance:

What I like:

Unlike the big tech versions, Ask Alex isn’t just a thinly veiled funnel or customer engagement mechanism. And it’s not a toy project like Ethan or Magic’s initial product. And Ask Alex is much more vertical and focused (right now, on restaurant discovery) than its predecessors.
Similar to The Nudge, why churn on a contact which lives in your phone? A bonus here is that there’s currently no cost to the customer. Ask Alex is scalable. It’s a simple, all-software, no-UI product whose AI improves with use.
Ask Alex is both a first and a last line of defense. Before embarking on the effort of looking up restaurant options, you’re likely to give Alex a try via text. And Alex also books for you, so rather than logging into OpenTable, etc, searching times, and booking, you can just have Alex do it for you. Either way, at the top or bottom of the funnel, you’re engaging with the same product.

What scares me:

Clowns
And also, the shuttering of Cloe, a very similar play which came out not too long ago. There’s some post-mortem to be done there which might surface vital learnings.
Monetization is a big question mark. The “promotion” model seems to be the best bet, and that’s not the most scalable sales strategy out there, especially in a fragmented industry like dining. Freemium is more scalable, but willingness to pay is also a question mark.
Having no real UI, brand aesthetic, or interactive “home space” is kind of radical. It’s a utility play, which eschews all that money that UX designers are paid to make product aesthetic immersive and addictive. There’s no real freeform product exploration here–you can’t get lost in the product and forget about the fact that you’re supposed to use it for something. This is both a potential pro and con. If it works, it’s a big win for channel over product in software. If it doesn’t, a bunch of people will be using conventional wisdom to tell us why.

“I’m the Map!”

This is the hottest and most competitive space here right now–and this is also the prototypical product strategy in the space. If I wanted to introduce what an AI-enabled experience discovery platform is and does, I’d point someone toward one of these.

What I like:

I feel that these platforms most directly address the consumer pain points in this space. They encapsulate the broadest set of potential use cases, have the broadest supply side of experience content/inventory, and are the most socially-enabled. If one of these works, it might really work.

What scares me:

Loud thunderstorms
And also, these are the most ambitious plays in this space — and I see the potential outcomes as kind of binary. These are the platforms which rely on network effects and mass adoption to work, and that’s a big hurdle to overcome to demonstrate product-market fit.
In addition to user risk, there’s competitive risk. These platforms are, in a way, stepping on the toes of Google Maps, TripAdvisor, TikTok, FourSquare, Yelp, etc. Before and if they become successful, there might be competitive responses — and recall that these products are also substitutes for users, who have developed ingrained behaviors around many of them.

One more business model

I’m a sucker for SMB enablement plays, and Shopify-style “lumpy” platforms. Thus, I’m also super interested in more white label, infrastructure business models which enable travel creators/brands/businesses to set up and monetize their own online stores. Bonus: with their combined scale, they’re able to negotiate commission rates that any individual travel creator couldn’t alone. Thatch is a great example here. It’s insulated from the perilous task of consumer demand gen. And it’s a market bet on a growing category, with a classic SaaS business model in an underaddressed space, and with commissions arbitrage.

…Ok. That’s it for today. Thanks for sticking with me and feel free to ardently (dis)agree in the comments section or via email.

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
Matt Rosenberg of Welcome
Jordon Durst of Ask Alex
Betsy Mule of F Prime Capital
Alex Attard-Manche of the Yard Ventures
…for your help and thought partnership

Markets by Grant | AI Experience Discovery

The Market

The Experience Economy

Back to the Market

An Argument With Myself

The Landscape

Content “Nudgers”

What I like:

What scares me:

Messaging “Buddies”

What I like:

What scares me:

“I’m the Map!”

What I like:

What scares me:

One more business model

Shout-Outs

References

Written by Grant Demeter