Pod Insights a16z Podcast a16z Podcast | What’s Next for Marketplace Startups (Hint: Services) About this episode The past and future of marketplace startups -- where are we? Ever since eBay popularized an internet meeting place for buyers and sellers of, well, just about everything, we’ve been waiting for 100 other at-scale marketplaces for everything else, including services.
So in this hallway-style episode of the a16z Podcast (originally recorded as a video) Li Jin -- co-author with Andrew Chen of this post -- chats with a16z Deal & Research team operating partner about why there aren’t 100 thriving marketplaces for services yet... And what’s changing to make this next wave of marketplace startups super exciting.
Transcript 0:01 Speaker 2 Hi everyone, welcome to the A6 and Z podcast. Today's episode was originally recorded as a video, which you can find in our YouTube channel at YouTube.com/slash A16Z videos of a Hallway Sale conversation between Deal and Research Team operating partner Frank Chen and Lee Jin, who covers our consumer vertical and co-wrote with General Partner Andrew Chen a post on the topic of what's next for marketplace startups. So in this video episode, Frank and Lee discuss the past and future of marketplace startups. And why aren't there a hundred services marketplaces yet?
0:36 Speaker 1 Well, hello, YouTube. My name is Frank Chen, and we're gonna have a hallway conversation today. Well, what's a hallway conversation? It's exactly like it sounds. We have tons of conversations here in the office with each other as we're thinking about what kinds of companies should we invest in. And what we're gonna do is take you inside the sausage factory so that you can hear a conversation exactly like that. So today I'm here with Lee Jin, who works on our consumer investing team. Welcome, Lee.
1:05 Speaker 0 Hello, everyone.
1:06 Speaker 1 And today we're gonna talk about marketplaces. So everybody knows eBay. eBay is an awesome place to find anything because it connects buyers and sellers. And ever since eBay and how it's been 30 years, something like that, eBay was definitely an internet 1.0 company. We've been waiting for a hundred at-scale marketplaces. There's so many awesome benefits you get when you bring people, buyers and sellers together, but there haven't been. There haven't been a hundred at scale marketplaces, and Lee has been wondering why. And so maybe Lee, let's start talking about why aren't there a hundred at scale eBay size marketplaces for buying services?
1:49 Speaker 0 Yeah. Well, first I want to back that up with one statistic. So I was doing research about services and how they how important they are to the economy. So services represents two-thirds of private sector spend by consumers in the economy. But when I looked up the statistics online, I think the Bureau of Labor Statistics has a statistic that only 7% of the services that we purchase are digital, meaning that they have been mediated by the internet. So despite the huge amount of spend that we have on services, um, very few of them are actually occurring online. Um I think there's a four there's four reasons why that is the case. Um I think services are obviously really complex when we go to a service provider and we're looking for a haircut or a babysitter or child care or a hotel room, there's so many different attributes that we're looking for.
2:48 Speaker 0 There's so much complexity in our value function of things that we value. For instance, for a babysitter, are they good with children? Do they speak other languages? Do they know a CPR? Uh how kind are they? Um services are super heterogeneous and difficult to standardize into SKUs the way that goods are. And that complexity makes it really difficult to capture the relevant information in a marketplace.
3:17 Speaker 1 Hard to skewify. Exactly. So many different attributes.
3:20 Speaker 0 Yeah. I think another reason why services have sort of been underpenetrated is that there's no clear metric of success or what quality looks like. Um, so what a good service experience has been for me might not be viewed the same way by another consumer. Um what one consumer thinks of as a five-star experience could be three stars for someone else. Um, going back to my point about the very complex um value functions that consumers have.
3:52 Speaker 1 So sort of beauty is in the eye of the beholder.
3:55 Speaker 0 Exactly. Yeah. We think another reason why services have been slow to come online is because a lot of services are performed by small business owners and individual owner operators. Um, and these businesses typically don't have the tools or the resources to really invest in going online, setting up a web page, making sure that their availability is somehow listed online and that they're bookable. Um, so a lot of it still remains offline for that reason.
4:28 Speaker 1 I think of my favorite near neighborhood barber where I get my haircut, Belmont Hair Studio. It's this delightful family, their immigrant family, and they love cutting hair and they love talking to people. But their sign of the prices and services available, they haven't touched that for three decades.
4:44 Speaker 0 Yeah, and often when you call them, it's actually it's been updated and the prices have gone up versus what's online because that web page is probably 20 years old. Yeah,
4:52 Speaker 1 it's the absolute last thing they think about.
4:54 Speaker 0 Yeah, exactly. And then I think another key aspect that distinguishes goods from services is that goods, um, the stages of the value chain from manufacturing to production to shipping it to a warehouse to uh consumption and purchase by a consumer, all of those stages in the value chain can sort of be split apart and broken up and sort of and each tackled by a different company in a different marketplace. So, for instance, Amazon is responsible for keeping things in stock and making sure that your orders get fulfilled. Services are a bit more complex because they involve this very synchronous creation and receiving of value by the consumer.
5:41 Speaker 0 Oftentimes services happen at the same place at the same location in an in-person meeting between the service provider and the consumer. So there's difficulty in extracting pieces of the value chain that can actually be brought online. Um, like if you're going to go in for a haircut and meet with your barber and actually set up a time to do that, then it's probably likely that it might be easier today to just schedule that in person or to do the payment in person while you're getting your haircut rather than breaking up those stages of the value chain into a marketplace.
6:23 Speaker 1 So all of those are reasons that we don't have a hundred at-scale marketplaces, right? It's hard. There's lots of complexity in SKUIFI, right? Products and attributes. There's lots of complexities in delivering. But a couple companies have managed to emerge, and I think what you did in the blog post was you started looking at the history of the marketplaces that have been successful. So when you dug into the history, what did you find?
6:48 Speaker 0 Yeah, so we really zoomed out and thought about the last, say, 30 years of marketplaces and what's been built so far. Um, and when we took that survey of the history of service marketplaces, what we found was that there's um there's pretty clearly been various eras of marketplaces. Specific specifically, there's been four major paradigms or models for marketplaces that have been trying to tackle services. So in the beginning, in the early days of the internet in the 90s, um, the first attempt at bringing services online were what we call the listings era. Uh, basically web pages that were digitized versions of the yellow pages.
7:34 Speaker 0 So in the yellow pages, you just had, you know, an entire book full of service providers that you could go down and actually call. Um marketplaces from this era were pretty similar to that. So examples of this would be Craigslist or Yelp, which basically gave consumers a huge list of all of the available service providers nearby. Um, but it really put the onus on the consumer to have to do the legwork in uh clicking on all those links, uh, finding their contact information, reaching out, coordinating with the provider, um, setting up a time to meet offline and actually transacting offline.
8:17 Speaker 1 Yeah. And so there was definitely a set of people who love browsing through Yellow Pages' ads, right? The ones at the back, or they loved looking through classified ads and newspapers and sort of highlighting them and clipping them out. But like for most of the people, that's a miserable experience. Like that's the last thing that you want to find a great hairdresser or to find a great daycare. You don't want to go through lists and lists and lists, and you kind of are suspicious of the big ads because you're like, oh, maybe like you know, they're spending all their money on advertising and not on actual child care, right? And so for most, that was a pretty bad experience. Yeah, so browsing listings is a pretty miserable experience.
8:51 Speaker 0 Yeah. And I think listings, it's not only tedious and a lot of work for consumers. Um, there were also problems with trust. So for services that required some component of trust where you were potentially letting someone into your home or letting someone interact with your child, it was really difficult to go onto these listings platforms and have high confidence that the service provider that you were speaking to would actually be someone that you could trust.
9:20 Speaker 1 Yeah. One of the fun facts from this era is if you can find yellow pages, go look. You might have to go to a museum to try to find one of these things. But if you look at locksmiths, you will see a lot of locksmiths named A and triple A locksmiths, and the reason for that is they would actually show up first in the listings, right? And so if there was something like that you could do to game traffic to your business, it's gonna create trust problems, right? Which is what makes triple A locksmith any better than ZZZ locksmith. Uh, but for for some reason, we do have a lot of triple A locksmiths and very few Z Z Z. So there was definitely some gaming going on.
9:58 Speaker 0 I think the modern day counterpart. To the AAA locksmith example is providers who are just listing their service over and over again on Craigslist. And you'll see those duplicate listings if you go back a few pages because it's chronologically ordered, and so the people who rise to the top are the ones that just post a ton of times.
10:19 Speaker 1 Yeah. And they'll post small variants, right? I'm a home theater installer, San Jose, home theater installer, Cooper Geno, right? So yeah. You're like, I know that's that's you. Same you. So that was the listings era. So pretty miserable from um end user experience point of view, right? Like I'm browsing, I don't know who's there, I don't know how to trust. Like what happened next? Like how did how did we get past the well, Crux list still exists, but it's not the marketplace for everything.
10:46 Speaker 0 Yeah, so what we started seeing next is um companies started addressing those issues of trust, of difficulty of finding providers and tediousness of actually reaching out to these folks by picking off various pieces of Craigslist and and creating what we call unbundled Craigslist marketplaces. And so there's that famous graphic that you can probably see online, which is like the Craigslist homepage, and basically every single category now has its own specialized marketplace that tackles just that particular category. So for instance, home services, which used to be, which is still its own Craigslist section. Um now we have Angie's list and ThumbTAC and Task Rabbit that are just doing home services.
11:36 Speaker 0 Um in the child care section, we have things like uh care.com, um and there's a bunch of other examples. So basically the next generation of marketplaces unbundled Craigslist and created verticalized marketplaces around a specific service category so that they could offer uh richer features, a better filtering um technology, better ways for users to be able to search and uncover the right provider for them.
12:05 Speaker 1 So by specializing, you could build features tailored to the service providers that you were trying to connect with consumers. So what's a good example of that?
12:14 Speaker 0 Yeah, so an example that I like is on Angie's list. Uh Angie's list actually has like a certification process that contractors and home service providers can actually go through so that they get a little badge on their listing, which is like Angie's list certified. Recommended recommended. And a lot of these vertical, verticalized uh marketplaces have some sort of similar credentialing or verification or recommended process because what they're trying to do is they're trying to help consumers figure out what's good quality versus what's poor quality, what who are the service providers that have a lot of good customer satisfaction versus the ones that you know are more average.
13:06 Speaker 0 And so this is this is trying to tackle that problem of search and discovery.
13:13 Speaker 1 Yeah. So the badge of awesomeness helps service providers stand out from each other. I remember this era well because I was actually working at a consumer company called Respond.com, which was a participant in this. We were trying to be a marketplace that would let you find local service providers. And the way it worked is you'd describe what you want in your own words and you'd fill out a form. And then we would take that and send it to a list of local service providers, and they would bid on your business. So kind of a reverse auction, you would say, I will change that lock for $50, or you get another one saying, no, I'll change the alloc for $75 or something like that. And one of the things we ended up having to do is that for every service provider, they wanted to know different things. The locksmith wants to know, do you need me now?
13:59 Speaker 1 Do you need to change locks? What do you need me to do? Whereas the window covering person needed to know how many windows. And do you want blinds or do you want curtains? And do you want short or fuzzy, right? Right. And so we ended up having to build these specialized forms for every service provider so that when they got the request, they could respond to it intelligently. So the specialization helped sort of streamline the communication
14:24 Speaker 0 between somebody who wants
14:25 Speaker 1 something and then somebody who was providing something.
14:28 Speaker 0 Right.
14:28 Speaker 1 Got it. So that was sort of era two, which is we moved out of straight listings and we moved to an era where there were specialized feature sets for the marketplace. And then and then what happened? What was next?
14:40 Speaker 0 Yeah, so then starting in the late 2000s, um, early half of the 2010s, we saw the rise of a ton of on-demand marketplaces, um, or what we call in the blog post Uber for X marketplaces. Um these, I think the the why now for those that generation of marketplaces was really the advent of mobile and the rapid adoption of smartphones by consumers and service providers. All of a sudden people could know where various providers were located relative to them. Um you had location that you could do matching on. Um people could respond to requests immediately rather than waiting until they got home to their computer and could check their email.
15:29 Speaker 0 And so this was really conducive to an on-demand model where you could receive a service very quickly, almost immediately in real time, without having to wait and to have a back and forth messaging process over potentially multiple days. Um the on-demand model of marketplaces was really conducive to um simple services that were uh very atomic and simple transactions that could uh you could do matching for in a very quick algorithmic way without you know matching on various different complex dimensions.
16:09 Speaker 1 You didn't have to ask 27 questions, it's kind of push button and something useful happens, right? Car arrives, a somebody who will pick up your stuff at Costco arrives or whatever, right?
16:20 Speaker 0 Exactly. And I think um another thing that we saw happen with this generation of marketplaces is that because the matching happened in an algorithmic centralized way, what happened was that the supply side um sort of became more obscured and sort of abstracted away for consumers. So as a consumer, you would say something like my Uber driver picked me up, or I'm getting food delivered through DoorDash. You wouldn't think of the counterparty of the transaction as being that actual end provider, the way that you did in previous generations of marketplaces. Yeah. You
17:00 Speaker 1 didn't know their name, you just wanted to know that like I don't know, they had four stars or more or something like that, right? And that was good enough. Yeah.
17:06 Speaker 0 Because
17:06 Speaker 1 otherwise they were fungible, right? They're sort of replaceable.
17:09 Speaker 0 Yeah. And that's because these on-demand marketplaces really nailed in on a singular use case and took hold of that entire end-to-end transaction from discovery to matching uh to payment and actually a transaction. Um so because that entire service delivery process was captured by those marketplaces, um, people didn't really think of uh the the end provider wasn't top of mind for the consumer.
17:37 Speaker 1 They were sort of more or less replaceable, right? Exactly. As long as you're reputable and strong enough to carry my boxes, I don't really care who care who shows up to carry my boxes.
17:46 Speaker 0 Yeah. And we also saw, I think, uh a large reckoning with this generation of marketplaces as well, um, where a lot of the startups that were trying to become Uber for X didn't ultimately end up lasting and surviving as standalone companies. And we think that that's because the model got applied probably a bit too liberally to service categories that weren't as conducive to the on-demand model. So there's a lot of services out there that actually don't need to be delivered immediately that people are willing to wait for. Um there's other service categories that are more complex where consumers actually want to know who they're interacting with at the end of the day rather than just the platform itself.
18:34 Speaker 1 And that gave way to the next era of marketplaces, right? Exactly. So where the person actually mattered, they weren't fungible. And so what do you guys call this era?
18:43 Speaker 0 Yeah, so the most recent era of marketplaces is something we call the managed marketplace era. Um so managed marketplaces went a step beyond all of the previous eras of service marketplaces in really providing a ton of value add to the consumer as well as to the service provider in tackling even more complex services and doing a heavier operational lift. Um and so these managed marketplaces take on some additional value add or functionality that consumers and suppliers really value that they previously had probably been doing themselves. So examples of this would be Honor, um, which is one of our portfolio companies that uh provides elder care.
19:31 Speaker 0 Um, Honor actually interviews vets, recruits caretakers who they employ as W-2 employees, um, so that they meet some sort of minimum standard bar and they can be trained and they can provide really high quality service. And previously, in you know, in the care.com era of care marketplaces, the end user, the end consumer would have had to reach out to a bunch of different. Professionals interview all of them and do all of this work. Now a platform like Honor is taking on that work and the operational heavier lift on their behalf.
20:10 Speaker 1 Yeah, Seth and Crewe are doing such a good job. And he was explaining to me the last time I was chatting with him about how complicated it is to find you the perfect person to do the perfect task and interact with just the right person, right? There's so many variables, like what language do they speak and can they lift heavy things, and are they punctual or not? And all of these things are very complicated and lead you to either have an okay or a great experience.
20:35 Speaker 0 Right. And
20:36 Speaker 1 so they actually have to use machine learning algorithms to match service providers to maximize the chances that you'll actually have a great experience because it is complicated.
20:45 Speaker 0 Yeah, exactly. I think of the managed marketplace era and this managed model as being a really good model for services that are complex that require higher trust, um, that maybe have higher stakes involved, um, or potentially like are involving higher value transactions. Um for instance, uh open door is another example of a managed marketplace for buying and selling real estate. This is another one of our recent investments. Um rather than um following the existing model of real estate in which most real estate agents actually are contractor or independent contractors and they work on commission, open door actually, I think employs people to go show these homes to help homeowners sell their homes, and they actually uh they're super managed because they actually purchase the home from you if it doesn't sell within a certain time frame.
21:47 Speaker 1 Yeah. One of the Alex, who's the general partner who led the investments theses is that 10 or 20 years from now we'll be buying homes from corporations rather than people, because corporations have processes, technology they can bring to bear to sort of make sure every experience is awesome, right? They're pricing it right, they're staging your home right, they're um uh showing you the right houses, uh, whereas individual operators, individual agents may or may not have the experience to do that. Right. So it makes total sense that like we would professionalize things like that. And for the most important things in our lives, right? What house do we live in? Who takes care of our parents when they can't take care of themselves, or who watches my kids while I'm at work, right? All of these things really need this era of management, right?
22:34 Speaker 1 You need to trust the service provider, you need to trust the company, bringing you just the right set of people.
22:41 Speaker 0 Yeah, exactly. I think of managed marketplaces as a model that works really well if the existing customer experience is really suboptimal in like a less managed version. Managed marketplaces can often create an extremely wow customer experience that you would have never been able to accomplish before had the customer still been looking for a service provider on, for instance, a listings platform.
23:08 Speaker 1 Right.
23:09 Speaker 0 And it's a lot easier for them,
23:10 Speaker 1 right? You don't get a list of six people that you then have to call and interview and schedule and figure out how to pay and all that. Yeah,
23:17 Speaker 0 exactly.
23:18 Speaker 1 Yeah. Awesome. So that was sort of the managed marketplace era. Well, that was a fun romp through history. We like to do these historical walks through all of the types of investments we made. So we talked about listings from the 90s, we talked about the unbundled Craigslist era, we talked about the Uber for X, and then we sort of have just been talking about managed marketplaces where you need higher trust for the service provider on the other end. What are you excited about now?
23:45 Speaker 0 Yeah, so we ask ourselves that every day and thinking about what investments we should make. Um zooming out of this entire historical picture. I think our one of our observations about the best marketplace investments and the best marketplace companies are that they tend to be marketplaces that have more constrained supply than demand. Um the the biggest marketplace outcomes that we've seen, or the biggest marketplace companies that we've seen are ones in which they've always struggled to keep up with the demand on the platform. So, really good examples of that would be Airbnb and Uber. So during the entire lifetimes of both of those companies, there's always been an excess of demand and supply has always struggled to keep up.
24:35 Speaker 0 There's never been enough homeowners who are willing to open up their homes because there's just so many consumers who want that kind of authentic travel experience. Similar with Uber, ever since that business model was pioneered, um, they've always had more trouble, relatively more trouble finding drivers than finding consumers who actually want to ride. Um so we think that this is actually a really key insight because we think that the best marketplace companies, the ones that can grow to be the biggest and really scale, are going to be the ones that tap into some sort of latent demand that already exists in consumers that consumers already really want but are unable to access today.
25:22 Speaker 0 And so in other words, we're looking for categories where there's a lot of demand that is being unfulfilled today. Um that leads us to what we think is next, which is regulated services and marketplaces that address regulated services. So we the reason why we zeroed in on regulated services is because by definition, there is an artificial constraint on supply in those service categories. Um there's regulation that exists to basically be able to say who is able and who is not able to provide the service to consumers. And so we think those are really promising categories for people to build companies in.
26:04 Speaker 1 What are some examples of services, the service industries that are regulated?
26:09 Speaker 0 Yeah, so you can go on to the Bureau of Labor Statistics and it'll actually give you like a huge long list of all of these professions. Um the major ones would be law, accounting, medicine, and healthcare. Um those are the major professions where a huge majority of those service workers are actually regulated and licensed. Exactly, or exactly, has passed the medical board of California, et cetera. And then you get down to actually like professions where there's not as much risk to public health and and wellness if the provider is not certified. So, for instance, in some markets and some geographies, florists actually need to be certified in order to operate a flower shop and to practice as a florist.
27:04 Speaker 1 I didn't know that.
27:05 Speaker 0 Um, and people have been driven out of business because they can't afford to go through that licensing process. Um, another example of this is um in many geographies, interior designers actually need to be licensed in some way in order to decorate other people's homes. Um there's there's categories where we think licensing makes a lot of sense to protect the public health. And then there's categories where we think that there could actually potentially be an unlicensed supply pool that meets a lot of the lit the latent demand that exists out there.
27:42 Speaker 1 So a little like Lyft when they thought about, gee, do you need a taxi medallion in order to drive your car around? Their argument was no, you have a car, you can have proven track record of safety because you haven't gotten into any accidents, so why don't we let you drive your car around?
27:56 Speaker 0 Exactly. And without that kind of unlock, without expanding it to the unlicensed pool of potential drivers, um, there's no way that Uber or Lyft could be anywhere near the size they are today. Yeah.
28:10 Speaker 1 So it sounds like the first thing that we're looking for is we don't necessarily have to have consumers learn a new habit, right? It's something that they do already, get your haircut or whatever, uh, or go get a uh medical procedure, right? So it's something that you're doing already, but we think that the a crop of startups will make this experience better, and why don't you uh provide some examples? How will regulated service marketplaces make your life better first as a consumer of these services?
28:41 Speaker 0 Yeah, I think we've all had experiences where we have tried to get access to a service provider in a regulated industry, and we come upon challenges in accessing that service. So, for instance, um it might take months to get an appointment with a particular specialist in a medical field. Or perhaps it's cost prohibitive for you to receive interior design services because those people need to be licensed. So this shows up in the consumers' lives in the form of like decrease in access, higher prices that you have to pay, longer waiting lists, or just generally having difficulty accessing that type of service.
29:25 Speaker 1 Yeah, so that definitely sounds better, right? Uh ready supply. I can meet with somebody sooner, I can have the same service at a lower price. Yep. So, what are the challenges startups have in making all that magic true, right? Because they need to unlock the supply side for that to be true.
29:41 Speaker 0 Right. So there's a few strategies that uh we walk through in the blog post about how can marketplaces actually unlock the supply side in these regulated industries. Um of them is in some industries, it's enough to just make discovery easier. Of these licensed providers. If today there's no resource for people to turn to in finding licensed providers, potentially just having a place where you can go and be connected to them is back to listings.
30:14 Speaker 1 Exactly.
30:15 Speaker 0 Right. Good search. So an example of this is like Zoc Doc, which helps consumers find doctors previously where consumers had to call and leave voicemails and try to make appointments, they can now do that online. So that's that's one strategy. Another strategy for unlocking supply is hiring and managing the providers themselves as a platform.
30:42 Speaker 1 Go full stack. Exactly.
30:44 Speaker 0 Right. So the managed model. And under the managed model, the the outcome of that is that companies can really raise the level of service and quality and accessibility and make these service providers a lot easier to access and have a consistent standards bar for them to really elevate the consumer experience.
31:08 Speaker 1 Got it. So down this path, you're likely raising more money because having them on your books is more expensive rather than having them as contractors. But your argument is that in some of these marketplaces, you have to have that extra increment of quality for the service to work.
31:25 Speaker 0 Right. Yeah, potentially there is just huge variance among the licensed supply pool, such that a managed marketplace can really come in and relieve consumers of that problem of finding the high quality supply. The next strategy that we identified in unlocking supply is companies that are augmenting or expanding the licensed supply pool. So a perfect example of this is Lyft or Uber. So expanding beyond the current pool of professional livery drivers, who are other people in the population that could provide a similar service, perhaps not the same one that consumers had experienced before, but something that met the same the same consumer need.
32:11 Speaker 0 And you can apply this model to other industries as well. So a startup that we recently met with building a marketplace for buying a dog as a pet. So I don't have any dogs, but what I've heard is that it's just really difficult for consumers to find a good responsible dog breeder and a healthy dog. Exactly. And a lot of the dog breeders live in other geographies or other states, and it's not easy to go and visit them. And so what Good Dog is doing is it's actually finding these breeders in the middle of the country that otherwise you couldn't get connected to and creating its own set of standards as an alternative to alternative to the licensing that exists today, which doesn't necessarily map to what consumers actually value and expanding the supply pool there.
33:06 Speaker 1 Got it. So rather than the traditional license, you trust good dog to find a good breeder. And you trust that because they're doing all the vetting and all of the standards and all the training that the breeder you get connected to will be first rate.
33:20 Speaker 0 Yeah. And then the last strategy to unlock supply in these regulated industries is using AI and automation. Yay,
33:30 Speaker 1 machine learning. So
33:31 Speaker 0 probably your favorite category.
33:33 Speaker 1 My favorite.
33:35 Speaker 0 So today, in order to be a truck driver in the US, you have to be licensed, and this is the most common profession in most states in the in America. And yet we're still extremely supply-constrained when it comes to the truck to driving industry. There just needs to be a lot more drivers on the road than there are today. And so there's a number of companies working on self-driving trucks to relieve the industry of that supply challenge. I've also seen startups that are applying AI to dermatology and skincare. So there was a company I met with recently called MG Acne, in which you take a selfie of your face and it uses AI to detect pimples or discoloration. And instead of sending that to a dermatologist to review, they use AI to review it and to determine the correct product, the correct ingredients that you should be using on your skin.
34:29 Speaker 1 That's a clever way to get around the constrained supply of board-certified dermatologists. Right. Exactly. It's not like an advanced thing. It's sort of the very basics, what creams and what should you do with your skin?
34:46 Speaker 0 Precisely. And for companies that are, you know, not issuing prescription ingredients and asking consumers to go get a prescription, I think that kind of solution is perfect because otherwise there's a lot of people out there who would never have the time or the resources to go to an actual dermatologist for help.
35:08 Speaker 1 Got it. So of all of the companies that you've sort of met in this regulated services arena, what are some of your favorite categories? Like what are you personally excited about?
35:19 Speaker 0 Yeah, so personally I am really excited about the mental health and wellness space. So therapy today is a product that is really what I think of as a luxury product. So mental health is something that people can really only access if they are really wealthy and have extra time and money to pay for it. Most psychiatrists are out of network, and so they can charge prices of you know even $600 an hour. And so this is really a product that most people can't access. And so I'm excited about the potential for a marketplace that perhaps leverages AI and automation, perhaps make some of the service delivery asynchronous rather than synchronous for a marketplace to really come in and improve access to that kind of service.
36:16 Speaker 0 And I've seen a few attempts at this so far that I think are pretty clever and interesting. So one startup based in San Francisco is called Basis, and they're using unlicensed but trained people in the community to basically serve as therapists. And so the concept is that you have an hour-long long conversation with someone who isn't actually a therapist that who knows the techniques to help you identify problems or actually tackle issues in your life that can really improve your mental health and your happiness.
36:57 Speaker 1 And so let me let me play a board-certified psychotherapist and give you the emoji, which is oh my goodness, we're really gonna let untrained, unboard certified psychotherapists out into the wild and like coach people, like what could possibly go wrong? So how do we answer that question?
37:17 Speaker 0 Yeah, I think that's going to be something that they need to tackle and really figure out how to address, because I think there is definitely a limit of the types of cases and the types of patients that the unlicensed but trained model can be applied to.
37:35 Speaker 2 I
37:36 Speaker 0 think for a large portion of the population that's just going through everyday stresses and anxiety, having someone to talk to is just a great resource in and of itself. Right. Um,
37:48 Speaker 1 maybe we pick patients that aren't most at risk.
37:50 Speaker 0 Exactly. Yeah, and perhaps for the patients that are most at risk, they can be then routed to or sent to you know an actual licensed therapist.
38:04 Speaker 1 So mental health is one category you're interested in. What else?
38:07 Speaker 0 Yeah. Um another vertical that I think this kind of uh regulated service marketplace could be really applicable to, one that is very near and dear to my heart, is um something in the beauty space. So a lot of makeup artists today have to be licensed as cosmetologists, and they have to go and receive that licensing and pay the fees associated with that. Um makeup artistry is actually also a luxury good today.
38:37 Speaker 2 So
38:37 Speaker 0 we think of celebrities as people with you know makeup artists that they have following them around, doing their makeup every day, maybe for special events, women will splurge for a makeup artist, but on a day-to-day basis uh we have to do our own makeup and we have to figure it, figure it out ourselves. And so, how can we give every woman that experience of having a personal makeup artist who helps them look their best and feel confident every day? Um I think that would be a really interesting application of AI and a lot of the themes that we talked about in this conversation to really improve access to that kind of service. Yeah,
39:19 Speaker 1 I love the idea that you're gonna democratize the service by bringing the prices down. And I love the idea of sort of maybe makeup artists building their brand by YouTube, and so like you're kind of following them anyway, and then now they can actually do your makeup. I think that's a great idea. And then if you could offer encouragement to entrepreneurs who are starting companies in this space, what what would you say to them?
39:42 Speaker 0 Yeah, I would just say that there's so many opportunities that are still left. Uh there's a table in our blog posts with the various regulated professions and examples of startups that are tackling them. And you'll probably notice that many of them are actually blank and don't have any names listed now. Next to them. And that's because we couldn't find any startups that are doing it. And so I would just say to entrepreneurs that there's so many service categories that are still left to build marketplaces in. Marketplaces for services are a huge opportunity because services represent 80% of the US GDP. And so there's a ton of opportunities left. And if you're working on any of them, we'd love to hear from you.
40:32 Speaker 1 All right, Startup Land. If one of those empty spaces in the blog post has your name on it because you can't stop thinking about it. You're at the breakfast table and you're like, oh, I can't stop thinking about this, then think about starting a company and come see Lee. Well, thanks, Lee, for joining us in this hallway conversation.
40:49 Speaker 0 Thank you.
40:50 Speaker 1 We hope to do lots more of these. And so if you liked this conversation, leave us comments, subscribe to the channel. And if you have thoughts about what other hallway conversations you'd like to hear about, let us know. So see you next time.
Previous What's Next for Education Startups Next Immigration Shifts 2019: Prep & Resources for Startups Page updated August 25, 2026 · transcript and summary are machine-generated