00:01hi everyone welcome today 6mz podcast
00:04today's episode was originally recorded
00:06as a video which you can find in our
00:08youtube channel at youtube.com slash a
00:111600 e cell conversation between deal
00:14and research team operating partner
00:16Frank Chen and Li Jin who covers our
00:19consumer vertical and co-wrote with
00:21general partner Andrew Chen a post on
00:23the topic of what's next for marketplace
00:25startups so in this video episode
00:27Franken Li discussed the past and future
00:30of marketplace startups and why aren't
00:33there a hundred services marketplaces
00:35well hello youtube my name is Frank Chen
00:38and we're gonna have a hallway
00:40conversation today well what's a hallway
00:41conversation it's exactly like it sounds
00:44we have tons of conversations here in
00:47the office with each other as we're
00:49thinking about what kinds of companies
00:51should we invest in and what we're gonna
00:53do is take you inside the sausage
00:55factory so that you can hear a
00:57conversation exactly like that so today
01:01I'm here with Legion who works on our
01:03consumer investing team welcome Lee and
01:06today we're going to talk about
01:08marketplaces so everybody knows eBay
01:12eBay is an awesome place to find
01:14anything because it connects buyers and
01:16sellers and ever since I Bey and how
01:19it's been 30 years something like that
01:21anyway was definitely an internet 1.0
01:23company we've been waiting for a hundred
01:27at scale marketplaces there's so many
01:29awesome benefits you get when you bring
01:32people buyers and sellers together but
01:35there haven't been there haven't been a
01:36hundred at scale marketplaces and Lee
01:39has been wondering why and so maybe
01:42let's start talking about why aren't
01:44there a hundred at scale eBay sized
01:46marketplaces for buying services yeah
01:49well first I want to back that up with
01:52one statistic so I was doing research
01:54about services and how they how
01:57important they are to the economy so
01:59services represents two-thirds of
02:02private sector spend by consumers in the
02:05economy but when I looked up the
02:08statistics online I think the Bureau of
02:11Labor Statistics has a statistic
02:14only 7% of the services that we purchase
02:18meaning that they have been mediated by
02:20the internet so despite the huge amount
02:22of sponge that we have on services very
02:26few of them are actually occurring
02:27online and I think there's a four
02:31there's four reasons why that is the
02:33case I think services are obviously
02:37really complex when we go to a service
02:40provider and we're looking for a haircut
02:42or a babysitter or childcare or a hotel
02:47room there's so many different
02:49attributes that we're looking for
02:50there's so much complexity in our value
02:52function of things that we value for
02:56instance for a babysitter
02:57are they good with children do they
02:59speak other languages do they know CPR
03:02how kind are they services are super
03:05heterogeneous and and difficult to
03:07standardize into skews the way that
03:10goods are and that complexity makes it
03:13really difficult to capture the relevant
03:14information in a marketplace hard to
03:17skew of fine exactly many different
03:19attributes yeah I think another reason
03:21why services have sort of been under
03:24penetrated is that there's no clear
03:27metric of success or what quality looks
03:30like so what a good service experience
03:34has been for me it might not be viewed
03:37the same way by another consumer so what
03:40one consumer thinks of as a five star
03:42experience could be three stars for
03:44someone else going back to my point
03:46about the very complex value functions
03:50that consumers have the beauty is in the
03:54exactly yeah we think another reason why
03:58services have been slow to come online
04:00is because a lot of services are
04:03performed by small business owners and
04:06individual owner operators and these
04:09businesses typically don't have the
04:12tools or the resources to really invest
04:15going online setting up a webpage making
04:18sure that their availability is somehow
04:21listed online and that their bucket ball
04:24so a lot of it still remains offline for
04:26that reason I think of my favorite near
04:29neighborhood barber where I get my
04:30haircut Belmont Hair Studio it's just
04:33delightful family they're immigrant
04:34family and they love cutting hair and
04:36they love talking to people but their
04:39sign of the prices and services
04:41available they haven't touched that for
04:44yeah and often when you call them it's
04:45actually it's been updated and the
04:47prices have gone up versus what's online
04:49because that webpage is probably 20
04:51years old yeah it's the absolute last
04:53thing they think about yeah exactly and
04:55then I think another key aspect that
04:59distinguishes goods from services is
05:01that goods the stages of the value chain
05:06from manufacturing to production to
05:09shipping it to a warehouse to
05:12consumption and purchase by a consumer
05:14all of those stages and the value chain
05:16can sort of be split apart and broken up
05:20and serve and each tackled by a
05:24different company in a different
05:25marketplace so for instance Amazon is
05:28responsible for keeping things in stock
05:30and making sure that your orders get
05:32fulfilled services are a bit more
05:35complex because they involve this very
05:37synchronous creation and receiving a
05:41value by the consumer oftentimes
05:44services happen at the same place at the
05:46same location in an in-person meeting
05:49between the service provider and the
05:51consumer so there's difficulty in
05:55extracting pieces of the value chain
05:56that can actually be brought online like
06:00if you're going to go in for a haircut
06:01and meet with your barber and actually
06:05set up a time to do that then it's
06:08probably likely that it might be easier
06:12today to just schedule that in person or
06:15to do the payment in person while you're
06:18getting your haircut rather than
06:19breaking up those stages of the value
06:21chain into a marketplace mm-hmm so all
06:24of those are reasons that we don't have
06:26a hundred at scale marketplaces right
06:28there's lots of complexity in skew if I
06:32write products and attributes there's
06:34lots of complexities in delivering but a
06:37couple companies have managed to emerge
06:39and I think what you did in the blog
06:40post was she started looking at the
06:42history of the marketplaces that have
06:44been successful so when you dug into the
06:46history what did you find yeah so we
06:49really assumed out and thought about the
06:51last say 30 years of marketplaces and
06:55what's been built so far and when we
06:57took that survey of the history of
07:00service marketplaces what we found was
07:02that there's there's pretty clearly been
07:06various eras of marketplaces specific
07:09specifically there's been four major
07:13paradigms or models for marketplaces
07:15that have been trying to tackle services
07:17so in the beginning in the early days of
07:20the internet in the 90s the first
07:23attempt of bringing services online were
07:26what we call the listings era basically
07:30webpages that were digitized versions of
07:35the yellow pages so in the yellow pages
07:38you just had you know an entire book
07:40full of service providers that you could
07:42go down and actually call marketplaces
07:45from this era were pretty similar to
07:48that so examples of this would be
07:50Craigslist or Yelp which basically gave
07:53consumers a huge list of all of the
07:56available service providers nearby but
08:00it really put the onus on the consumer
08:02to have to do the legwork and clicking
08:06on all those links finding their contact
08:09reaching out coordinating with the
08:12provider setting up a time to meet
08:14offline and actually transacting offline
08:17yeah and so there was definitely a set
08:19of people who loved browsing through
08:20Yellow Pages ads right the ones at the
08:22back or they loved looking through
08:24classifieds ads in newspapers and sort
08:26of highlighting them and clipping them
08:28out but like for most of the people
08:29that's a miserable experience like
08:31that's the last thing that you want to
08:32find a great hairdresser or to find a
08:34great daycare you don't want to go
08:36through lists and lists and lists and
08:37you kind of are suspicious of the big
08:39ads because you're like oh maybe
08:41like you know they're spending all their
08:42money on advertising and not on natural
08:44childcare right and so for most that was
08:47a pretty bad experience yeah so browsing
08:49listings is a pretty miserable
08:50experience yeah and I think listings
08:53it's not only tedious and a lot of work
08:56for consumers there were also problems
09:00so for services that required some
09:03component of trust where you were
09:05potentially letting someone into your
09:07home or letting someone interact with
09:09your child it was really difficult to go
09:12onto these listings platforms and have
09:14high confidence that the service
09:15provider that you were speaking to would
09:18actually be someone that you could trust
09:20yeah one of the fun facts from this era
09:22is if you can find yellow pages go look
09:26you might have to go to a museum to try
09:28to find one of these things but if you
09:29look at locksmiths you will see a lot of
09:32locksmiths named a and triple-a
09:34locksmiths and and the reason for that
09:36is they would actually show up first in
09:39the listings right and so if there was
09:41something like that you could do to gain
09:42traffic to your business it's gonna
09:46create trust problems right which is
09:47what makes triple-a locksmith any better
09:49than a zze locksmith but for for some
09:53reason we do have a lot of triple-a
09:54locksmiths and very few zze so there was
09:56definitely some gaming going on
09:58I think the modern-day counterpart to
10:00the a locksmith example is providers who
10:05are just listing their service over and
10:07over again on the Craigslist you'll see
10:09those duplicate listings if you go back
10:11a few pages because it's chronologically
10:14ordered and so the people who rise to
10:16the top are the ones that just post a
10:18ton of times yeah and they'll post small
10:20variants right I'm a home theater
10:22installer San Jose home theater
10:24installer Cupertino right so yeah you're
10:27like I know that's that's you same you
10:30so that was the listings errors so
10:32pretty miserable from an end-user
10:34experience point of view right like I'm
10:36browsing I don't know who's there I
10:38don't trust look what happened next like
10:41how did how did we get past the well
10:43collects list still exists but it's not
10:44the marketplace for everything yeah so
10:47what we started seeing next is companies
10:50started addressing those issues of trust
10:53of difficulty of finding providers and
10:57tediousness of actually reaching out to
10:59these folks by picking off various
11:02pieces of Craigslist and and creating
11:05what we call unbundled Craigslist
11:07marketplaces and so there's that famous
11:10graphic that you can probably see online
11:12which is like the Craigslist homepage
11:15and basically every single category now
11:18has its own specialized marketplace that
11:21tackles just that particular category so
11:24for instance Home Services which used to
11:26be which is still its own Craigslist
11:28section now we have Angie's List and
11:32thumbtack and toss rabbit that are just
11:35doing home services in the childcare
11:37section we have things like CARICOM and
11:41there's a bunch of other examples so
11:43basically the next generation of
11:45marketplaces unbundled Craigslist and
11:48created verticalized marketplaces around
11:51a specific service category so that they
11:53could offer richer features better
11:56filtering technology better ways for
12:00users to be able to search and uncover
12:02the right provider for them so by
12:05specializing you could build features
12:07tailored to the service providers that
12:10you are trying to connect with consumers
12:12so what's a good example of that yeah so
12:15an example that I like is on Angie's
12:19List Angie's List actually has like a
12:22certification process that contractors
12:25and home service providers can actually
12:28go through so that they get a little
12:29badge on their listing which is like
12:31Angie's List certified exactly recommend
12:35it and a lot of these vertical vertical
12:38eyes marketplaces have some sort of
12:41similar credentialing or verification or
12:45recommended process because what they're
12:49trying to do is they're trying to help
12:50consumers figure out what's good quality
12:53versus what's poor quality what who are
12:56the service providers that have a lot of
13:01good customer satisfaction versus the
13:04ones that you know are more
13:05Brij and so this is this is trying to
13:08tackle that problem of search and
13:12discovery yeah so the badge of
13:14awesomeness helps preserve as providers
13:16stand out from each other
13:18I remember this era well because I was
13:20actually working at a consumer a company
13:22called respond comm which was a
13:24participant in this we were trying to be
13:26a marketplace that would let you find
13:28local service providers and the way it
13:30worked is you describe what you want in
13:31your own words and you'd fill out a form
13:33and then we would take that and send it
13:36to a list of local service providers and
13:39they would bid on your business so kind
13:41of a reverse auction you would say I
13:42will change that lock for $50 or you've
13:44got another one saying now I'll change
13:46the holic for $75 or something like that
13:48and one of the things we ended up having
13:50to do is that for every service provider
13:53they wanted to know different things
13:55the locksmith wants to know do you need
13:57me now do you need to change locks what
13:59do you need me to do whereas the window
14:02covering person needed to know how many
14:04windows and do you want blinds or do you
14:06want a curtains and do you want short or
14:08fuzzy right right and so we ended up
14:10having to build these specialized forms
14:12for every service provider so that when
14:15they got the request they could respond
14:17to it intelligently so the
14:19specialization helped sort of streamline
14:23the communication yeah between somebody
14:25wants something and then somebody who is
14:27providing something right good so that
14:29was sort of era to which as we moved out
14:32of straight listings and we moved to an
14:33era where there were specialized feature
14:35sets for the marketplace and then and
14:38then what happened what was next
14:39yeah so then starting in the late 2000s
14:44early half of the 2010s
14:46we saw the rise of a ton of on-demand
14:51marketplaces or what we call in the blog
14:54uber 4x marketplaces so these I think
14:59though the Y now for those that
15:01generation of marketplaces was really
15:03the advent of mobile and the rapid
15:06adoption of smartphones by consumers and
15:08service providers all of a sudden people
15:12could know where various providers were
15:16relative to them so you had location
15:20that you could do matching on people
15:24could respond to requests immediately
15:26rather than waiting until they got home
15:27to their computer and could check their
15:29email and so this was really conducive
15:31to an on demand model where you could
15:33receive a service very quickly almost
15:36immediately in real time without having
15:38to wait and to have a back and forth
15:41messaging process over potentially
15:44multiple days and so the on-demand model
15:47of marketplaces was really conducive to
15:51simple services that were very atomic
15:56and simple transactions that could you
16:00could do matching for in a very quick
16:02algorithmic way without you know
16:04matching on various different complex
16:07dimensions you didn't have to ask 27
16:10questions it's kind of push button and
16:12something useful happens right car
16:14arrives a somebody who will pick up your
16:17stuff that Costco arrives or whatever
16:19right exactly and I think another thing
16:22that we saw happen with this generation
16:25of marketplaces is that because the
16:29matching happened in an algorithmic
16:32centralized way what happened was that
16:34the supply side sort of became more
16:38obscured and sort of abstracted away for
16:41consumers so as a consumer you would say
16:45something like my uber driver picked me
16:47up or I'm getting food delivered through
16:49Jordache you wouldn't think of the
16:51counterparty of the transaction as being
16:53that actual end provider the way that
16:57you did in previous generations of
16:59marketplaces yeah they didn't know their
17:00name you just wanted to know that like I
17:02know they had four stars or more or
17:03something like that right that was good
17:05enough yeah that's otherwise they were
17:07fungible right they're sort of
17:09and that's because these on-demand
17:11marketplaces really nailed in on a
17:13singular use case and took hold of that
17:16entire end-to-end transaction from
17:19discovery to matching to payment and
17:23actually a transaction and so because
17:25that entire service delivery process was
17:29Marketplace's people didn't really think
17:33of the the end provider was on top of
17:36mine for the consumer mm-hm they were
17:38sort of more or less replaceable
17:39thanks as long as you're reputable and
17:41strong enough to carry my boxes I don't
17:43really care who carat who shows up to
17:45carry my boxes yeah and we also saw I
17:48think a large reckoning with this
17:51generation of marketplaces as well where
17:54a lot of the startups that were trying
17:56to become over 4x didn't ultimately end
18:00up lasting and surviving of stand-alone
18:03companies and we think that that's
18:05because the model got applied probably a
18:08bit too liberally to serve as categories
18:11that weren't as conducive to to the on
18:15demand model so there's a lot of
18:17services out there that actually don't
18:19need to be delivered immediately that
18:21people are willing to wait for there's
18:24other service categories that are more
18:25complex where consumers actually want to
18:27know who they're interacting with at the
18:29end of the day rather than just the
18:31platform itself mm-hmm and that gave way
18:34to the next era of marketplaces exactly
18:38so where the person actually mattered
18:40they weren't fungible and so what do you
18:42guys call this era yeah so the most
18:45recent era of marketplaces is something
18:48we call the managed marketplace era so
18:51managed marketplaces went a step beyond
18:54all of the previous eras of service
18:56marketplaces and really providing a ton
18:59of value on to the consumer as well as
19:01to the service provider in tackling even
19:04more complex services and doing a
19:07heavier operational lift and so these
19:11managed marketplaces take on some
19:13additional value add or functionality
19:16that consumers and suppliers really
19:18value that they previously had probably
19:21been doing themselves so examples of
19:24this would be honor which is one of our
19:27portfolio companies that provides
19:30eldercare honor actually interviews that
19:34recruits caretakers who they employ as
19:38w-2 employees so that they meet some
19:42minimum standard bar and they can be
19:44trained and they can provide really high
19:47quality service and previously in you
19:51know in the CARICOM era of care
19:53marketplaces the end user the end
19:56consumer would have had to reach out to
19:58a bunch of different professionals
20:00interview all of them and do all of this
20:03work now a platform like honor is taking
20:05on that work and the operational heavier
20:07left on our behalf yeah Seth and crew
20:10are doing such a good job and he was
20:12explaining to me the last time I was
20:14chatting with him about how complicated
20:15it is to find you the perfect person to
20:18do the perfect task and interact with
20:21just the right person right there's so
20:22many variables like what language do
20:24they speak and can they lift heavy
20:26things and are they punctual or not and
20:29all of these things are very complicated
20:31and lead you to either have an okay or a
20:34great experience right so they actually
20:36have to use machine learning algorithms
20:38to match service providers to maximize
20:41the chances that you'll actually have a
20:42great experience because it is
20:44complicated yeah exactly I think of the
20:46managed marketplace era and this managed
20:49model as being a really good model for
20:53services that are complex that require
20:56higher trust that maybe have higher
20:59stakes involved or potentially like are
21:02involving higher value transactions so
21:07for instance open door is another
21:10example of a managed marketplace for
21:13buying and selling real estate this is
21:14another one of our recent investments
21:17but rather than following the existing
21:22model of real estate in which most real
21:25estate agents actually are contractor or
21:28independent contractors and they work on
21:30Commission open door actually I think
21:33employs people to go show these homes to
21:37help homeowners sell their homes and
21:39they actually they're super managed
21:42because they actually purchased the home
21:44from you if it doesn't solve within a
21:46certain time frame yeah one of the Alex
21:49who is the general partner who led the
21:51investments theses is that 10 or 20
21:54years from now we'll be buying homes
21:56corporations rather than people because
21:58corporations have processes technology
22:01they can bring to bear to sort of make
22:03sure every experience is awesome right
22:05they're pricing it right they're staging
22:07your home right they're showing you the
22:10right houses where as individual
22:12operators individual agents may or may
22:14not have the experience to do that
22:15there's a total sense that like we would
22:18professionalize things like that
22:20inferred the most important things in
22:23our lives right what house do we live in
22:25who takes care of our parents when they
22:26can't take care of themselves or who
22:29watches my kids while I'm at work right
22:31all of these things really need this era
22:34of management right you need to trust
22:36the service provider you need to trust
22:38the company bringing you just the right
22:40set of people yeah exactly I think of
22:43managed marketplaces as a model that
22:46works really well if the existing
22:49customer experience is really suboptimal
22:52in like a less managed version managed
22:55marketplaces can often create an
22:57extremely Wow customer experience that
23:00you would have never been able to
23:01accomplish before had the customer still
23:05been looking for a service provider on
23:06for instance the listings platform right
23:09and it's a lot easier for them right you
23:11don't get a list of six people that you
23:12then have to call an interview and
23:14schedule and figure out how to pay and
23:16all that yeah exactly you know awesome
23:19so that was sort of the managed
23:20marketplace era well that was a fun romp
23:23through history we like to do these
23:25historical walks through all of the
23:27types of investments we mean so we
23:29talked about listings from the 90s we
23:31talked about the unknown Craigslist era
23:33we talked about the uber Forex and then
23:35we sort of had just been talking about
23:37managed marketplaces where you need
23:38higher trust for the service provider on
23:40the other end what are you excited about
23:43now yeah so we ask ourselves that every
23:47day and thinking about what investments
23:49we should make and so zooming out of
23:54this entire historical picture I think
23:57are one of our observations about the
24:01best market place investments and the
24:02best market place companies are that
24:05they tend to be market places
24:08have more constrained supply than demand
24:11so the the biggest marketplace outcomes
24:14that we've seen or the biggest
24:16marketplace companies that we've seen
24:17are ones in which they've always
24:20struggled to keep up with the demand on
24:22the platform so really good examples of
24:25that would be Airbnb and uber so during
24:28the entire lifetimes of both of those
24:30companies there's always been an excess
24:32of demand and supply has always
24:34struggled to keep up there's never been
24:37enough home owners who are willing to
24:38open up their homes because there's just
24:42so many consumers who want that kind of
24:44authentic travel experience similar with
24:46uber ever since that business model was
24:48pioneered they've always had more
24:51trouble relatively more trouble finding
24:53drivers than finding consumers who
24:56actually want to ride and so we think
24:58that this is actually a really a key
25:00insight because we think that the best
25:03market place companies the ones that can
25:05grow to be the biggest and really scale
25:06are going to be the ones that tap into
25:09some sort of latent demand that already
25:11exists and consumers that consumers
25:13already really want but are unable to
25:15access today and so in other words we're
25:20looking for categories where there's a
25:23lot of demand that is being unfulfilled
25:25today and so that leads us to what we
25:28think is next which is regulated
25:29services and marketplaces that address
25:33regulated services so we the reason why
25:38we zeroed in on regulated services is
25:40because by definition there is an
25:44artificial constraint on supply in those
25:46service categories and there's
25:50regulation that exists to basically be
25:54able to say who is able and who is not
25:56able to provide the service to consumers
25:59and so we think those are really
26:01promising categories for people to build
26:03companies in what are some examples of
26:06services the service industries that are
26:08regulated yeah so you can go on to the
26:11Bureau of Labor Statistics and I'll
26:12actually give you like a huge long list
26:15of all of these professions but the
26:17major ones would be law accounting
26:21medicine and healthcare those are the
26:25major professions where a huge majority
26:28of those service workers are actually
26:32regulated and like her of the bar or
26:35certified by the state of California
26:37exactly or exactly has passed the
26:41Medical Board of California etc and then
26:44you get down to actually like
26:46professions where there's not as much
26:49risk to public health and and wellness
26:52if the provider is not certified so for
26:55instance in some markets and some
26:57geographies florists actually need to be
27:00certified in order to operate a flower
27:02shop and to practice as a florist I know
27:04that and people have been driven out of
27:06business because they can't afford to go
27:08through that licensing process another
27:11example of this is in many geographies
27:16interior designers actually need to be
27:19licensed in some way in order to
27:22decorate other people's homes so there's
27:25there's categories where we think
27:27licensing makes a lot of sense to
27:29protect the public health and then
27:32there's categories where we think that
27:34there could actually potentially be an
27:36unlicensed supply pool that meets a lot
27:39of the lists the latent demand that
27:40exists out there mm-hmm
27:42so a little like lyft when they thought
27:44about gee do you need a taxi medallion
27:46in order to drive your car around their
27:47argument was no you have a car yes a
27:50proven track record of safety because
27:52I'm not in any accidents so why don't we
27:54let you drive your car right exactly
27:56and without that kind of unlock without
27:59expanding it to the unlicensed pool of
28:02potential drivers there's no way that
28:05uber or lyft could be anywhere near the
28:08size they are today yeah so it sounds
28:11like the first thing that we're looking
28:12for is we don't necessarily have to have
28:15consumers learn a new habit right it's
28:17something that they do already get your
28:19haircut or whatever or go get a medical
28:23procedure right so it's something that
28:25you're doing already but we think that
28:27the crop of startups will make this
28:29experience better and why don't you
28:31provide some examples how will
28:34regulated service marketplaces make your
28:37life better first as a consumer of these
28:39services yeah I think we've all had
28:43experiences where we have tried to get
28:47access to a service provider in a
28:49regulated industry and we come upon
28:52challenges and accessing that service so
28:55for instance it might take months to get
28:58an appointment with a particular
29:01specialist in a medical field or perhaps
29:05it's cost prohibitive for you to receive
29:08interior design services because those
29:10people need to be licensed so this shows
29:12up in the consumers lives in the form of
29:15like decrease in access higher prices
29:18that you have to pay longer waiting
29:21lists or just generally having
29:23difficulty accessing that type of
29:24service yeah so that definitely sounds
29:26better right ready supply I can meet
29:29with somebody sooner I can have the same
29:32service at a lower price yep so what are
29:34the challenges startups have in making
29:36all that magic true right because they
29:38need to unlock the supply side right so
29:42there's a few strategies that we walk
29:46through in the blog post about how can
29:48marketplaces actually unlock the supply
29:51side in these regulated industries one
29:54of them is in some industries it's
29:57enough to just make discovery easier of
30:00these license providers if today there's
30:03no resource for people to turn to in
30:05finding licensed providers potentially
30:08just having having a place where you can
30:11go and be connected to them in fact two
30:13listings a good search right good search
30:17so an example of this is like dock dock
30:20which helps consumers find doctors
30:23previously where where consumers have to
30:26call and leave voicemails and try to
30:28make appointments they can now do that
30:30online so that's that's one strategy
30:32another strategy for unlocking supply is
30:37is hiring and managing the providers
30:40themselves as a platform go full stack
30:43exam your employees right so the managed
30:47and under the managed model the outcome
30:51of that is that companies can really
30:53raise the level of service and quality
30:55and accessibility and make these service
30:59providers a lot easier to access and
31:02have a consistent standards bar for them
31:05to really elevate the consumer
31:07experience got it so down this path
31:10you're likely raising more money because
31:12having them on your books is more
31:14expensive rather than having them as
31:16contractors but your argument is that in
31:19some of these marketplaces you have to
31:21have that extra increment of quality for
31:24the service to work right yeah
31:26potentially there is just huge variance
31:28among the licensed supply pool such that
31:31a managed marketplace can really come in
31:33and relieve consumers of that problem of
31:37finding the high-quality supply the next
31:41strategy that we identified in unlocking
31:43supply is companies that are augmenting
31:47or expanding the license supply pool so
31:51a perfect example of this is lyft or
31:53uber so expanding beyond the current
31:56pool of professional livery drivers who
32:00are other people in the population that
32:01could provide a similar service perhaps
32:04not the same one that consumers had
32:06experienced before but something that
32:08met the same the same consumer need and
32:11you can apply this model to other
32:13industries as well so a startup that we
32:16recently met with is building a
32:20marketplace for buying a dog as a pet so
32:24I don't have any dogs but what I've
32:28heard is that it's just really difficult
32:30for consumers to find a good responsible
32:33dog breeder and a healthy dog exactly
32:36and a lot of the dog breeders live in
32:38other geographies or other states and
32:40it's not easy to go and visit them and
32:42so what good dog is doing is it's
32:45actually finding these breeders in the
32:48middle of the country
32:50that otherwise you couldn't get
32:51connected to and creating its own set of
32:54standards as an alternative to
32:57alternative to the licensing that exists
33:00doesn't necessarily amount to what
33:02consumers actually value and expanding
33:05the supply pool there
33:07so rather than the traditional license
33:09you trust good dog to find a good
33:11breeder and you trust it because they're
33:13doing all the vetting and all of the
33:14standards and all the training that the
33:16breeder you get connected to will be
33:20and then the last strategy to unlock
33:25supply in these regulated industries is
33:28using a ein automation so probably your
33:32favorite category so today in order to
33:37be a truck driver in the US you have to
33:39be licensed and this is the most common
33:41profession in most states in America and
33:44yet we're still extremely supply
33:46constrained when it comes to the truck
33:48driving industry there just needs to be
33:50a lot more drivers on the road than
33:51there are today and so there's a number
33:54of companies working on self-driving
33:55trucks to relieve the industry of that
33:58supply challenge I've also seen startups
34:01that are applying AI to dermatology and
34:05skin care so there was a company I met
34:08with recently called MD acne in which
34:11you take a selfie of your face and it
34:14uses AI to detect pimples or
34:17discoloration and instead of sending
34:20that to a dermatologist to review they
34:22use AI to review it and to determine the
34:25correct product the correct ingredients
34:27that you should be using on your skin
34:28that's a clever way to get around the
34:32constraints apply a board-certified
34:34dermatologists exactly you can sort of
34:36replace the basics of what they do right
34:38it's not like an advanced thing it's
34:40sort of the very basics what creams and
34:42what what should you do with your skin
34:45precisely and for for companies that are
34:49you know not not issuing prescription
34:52ingredients and asking consumers to go
34:56get a prescription I think that kind of
34:58solution is perfect because otherwise
35:00there's a lot of people out there who
35:03would never have the time or the
35:05resources to go to an actual
35:06dermatologist for help got it so
35:10of all of the companies that you've sort
35:12of met in this regulated services arena
35:15what are some of your favorite
35:16categories like what are you personally
35:20so personally I am really excited about
35:24the mental health and wellness piece so
35:28therapy today is a product that is
35:31really what I think of as a luxury
35:34product so mental health is something
35:37that people can really only access if
35:40they are really wealthy and have extra
35:43time and money to pay for it
35:46most psychiatrists are out of network
35:48and so they can charge prices of you
35:51know even $600 an hour and so this is
35:55really a product that most people can't
35:57access and so I'm excited about the
36:01potential for a marketplace that perhaps
36:04leverages a automation perhaps make some
36:10of the service delivery asynchronous
36:13rather than synchronous for a
36:15marketplace to really come in and
36:17improve access to that kind of service
36:18and I've seen a few attempts at this so
36:22far that I think are pretty clever and
36:24interesting so one startup based in San
36:28Francisco is called basis and they're
36:30using unlicensed but trained people in
36:35the community to basically serve as
36:36therapists and so the concept is that
36:39you have an hour-long long conversation
36:42with someone who isn't actually a
36:45therapist who knows the techniques to
36:47help you identify problems or actually
36:51tackle issues in your life that can
36:54really improve your mental health and
36:55your happiness and so let me let me play
37:00a board-certified psychotherapist and
37:02give you the emoji which is oh my
37:06goodness we're really gonna let
37:07untrained on board certified
37:09psychotherapists out into the wild and
37:12like coach people like what could
37:13possibly go wrong so how do we answer
37:16that question yeah I think that's going
37:19to be something that they need to tackle
37:22really figure out how to address because
37:24I think there is definitely a limit of
37:27the types of cases and the types of
37:30patients that the unlicensed but trained
37:32model can be applied to I think for a
37:36large portion of the population that's
37:38just going through everyday stresses and
37:40anxiety having someone to talk to
37:44is just a great resource in and of
37:46itself so maybe we pick patients that
37:49aren't most at risk exactly yeah and
37:51perhaps for the patients that are most
37:54at risk they can be then routed to or
37:57sent to you know an actual licensed
38:02therapist so mental health is one
38:06category you're interested in what else
38:07yeah another vertical that I think this
38:12kind of regulated service marketplace
38:15could be really applicable to one that
38:17is very near and dear to my heart is
38:19something in the beauty space so a lot
38:23of makeup artists today have to be
38:25licensed as cosmetologist and they have
38:27to go and receive that licensing and pay
38:30the fees associated with that and so
38:34makeup artistry is actually also a
38:36luxury good today so we think of
38:38celebrities as people with you know
38:40makeup artists that they have following
38:41them around doing their makeup every day
38:43maybe for special events women will
38:46splurge for a makeup artist but on a day
38:48to day basis we have to do our own
38:51makeup and we have to figure it figure
38:53it out ourselves and so how can we give
38:56every woman that experience of having a
38:59personal makeup artist who helps them
39:03look their best and feel confident every
39:06day I think that would be a really
39:09interesting application of AI and a lot
39:13of the themes that we talked about in
39:15this conversation to really improve
39:17access to that kind of service yeah I
39:19love the idea that you're gonna
39:20democratize the service by bringing the
39:23prices down and I love the idea of sort
39:25of maybe makeup artists building their
39:27brand by YouTube and so like you're kind
39:29of following them anyway and then now
39:31they could actually do your makeup I
39:33think that's a great idea and then if
39:36for encouragement to entrepreneurs who
39:38are starting companies in this space
39:39what what would you say to them yeah I
39:42would just say that there's so many
39:44opportunities that are still left
39:46there's a table in our blog posts with
39:49the various regulated professions and
39:52examples of startups that are tackling
39:54them and you'll probably notice that
39:56many of them are actually blank and
39:58don't have any names listed next to them
40:00and that's because we couldn't find any
40:02startups that are doing it and so I
40:05would just say to entrepreneur is that
40:07there's so many service categories that
40:09are still left to build marketplaces in
40:13marketplaces for services are a huge
40:18opportunity because services represent
40:2180% of the US GDP and so there's a ton
40:26of opportunities left and if you're
40:28working on any of them we'd love to hear
40:31all right startup land if one of those
40:33empty spaces in the blog post has your
40:36name on it because you can't stop
40:38thinking about it you're at the
40:39breakfast table and you're like I can't
40:41stop thinking about this then think
40:43about starting a company and come Sealy
40:45well thanks Lee for joining us in this
40:48hallway conversation thank you we hope
40:50to do lots more of these and so if you
40:53liked this conversation leave us
40:55comments subscribe to the channel and if
40:58you have thoughts about what other
41:00hallway conversations you'd like to hear
41:01about let us know so see you next time