Opportunity Amidst Innovation: Trends in Insurance as AI Reshapes Distribution | with TX-Zhuo

In this Episode 9 of Insurtech Unscripted, Ali Safavi talks about what if the future of insurtech is not about eliminating offline brokers, but empowering them? Prepare to have your perceptions about insurance distribution challenged as we welcome TX-Zhuo, the general partner at Fika Ventures, on our podcast. In a landscape often dominated by digital disruption, he advocates for a unique approach - leveraging artificial intelligence to empower traditional brokers and enhance their customer service.

As we navigate the evolving world of insurtech, we tackle the contentious debate between the embedded versus brokerage models, and the potential they hold. Drawing from TX-Zhuo's deep industry knowledge, we confront the prospect of bundling insurance with other financial products and how this could revolutionize the customer experience. We also delve into the complexities of the insurance market, discussing the potential for horizontal and vertical integration and the importance of focusing on customer demographics from the start.

But we don't just stop at strategies and models, we also discuss the realities and challenges of selling multiple insurance policies. From personal relationships in cross-selling to instant policy creation through data integration, our conversation delves into the nuts and bolts of the insurance industry. We also ponder over emerging trends such as net revenue multiples and AI-driven services and their potential impact on the future of insurance. Join us as we explore the future of insurance distribution and insurtech through the lens of a veteran industry insider.
-----------------------------------------------------------------------------------------------------------
COVU is founded and advised by a team of insurance, finance, and tech industry veterans with a passion for transparent and unbiased advice. COVU's mission is to help both everyday people and professional advisors manage risk and insurance smarter. We help our customers find and fill their coverage gaps and achieve true peace of mind.

Sign up: https://covu.co

COVU is founded and advised by a team of insurance, finance, and tech industry veterans with a passion for transparent and unbiased advice. COVU's mission is to help both everyday people and professional advisors manage risk and insurance smarter. We help our customers find and fill their coverage gaps and achieve true peace of mind.

Sign up: https://covu.com

Let’s stay connected!

⚡ Instagram: / covu_inc

👍 Facebook:
partner at FICO Ventures, todaywith me.
He's someone who's beeninvolved with insurance for a
while, who is a McKinseyinvolved with their insurance
practice, and he's beeninvesting in insurance for a
while.
I mean he invested in companieslike Policy Genius who, as you

(00:40):
know, are one of some of the OGsin the insurance space.
So he brings a lot of uniqueperspective into the podcast
with a lot of learnings thatobviously happened over the past
few years.
So let me start off with TX.
Welcome to the podcast.
Maybe you could tell us moreabout your background and where
your interests are coming from.

Speaker 2 (01:00):
Thanks for having me, ali.
It's such a pleasure to be heretoday.
So just very quickly, on mybackground, I was an
entrepreneur till in BC sostarted a very small bootstrap
company.
We're selling college textbooksonline, of all things.
So this was right out ofcollege.
We were very fortunate, endedup selling the company in 2006.
And at that point I realizedmaybe I was a good entrepreneur

(01:21):
but didn't have the rightdistance experience to scale a
company.
So spent a couple of years atMcKinsey and was very fortunate
to be part of their insurancepractice.
So worked with a lot of clientslike City Prudential, aig and
through that experience kind ofinformed how I wanted to invest
my time when I became a VC lateron.
And insurance tech hasdefinitely been a constant focus

(01:42):
within our portfolio since dayone.
So we've been investing ininsurance tech company since
2013.
It's about 10 years now.

Speaker 1 (01:49):
Very interesting.
So let me start off withsomeone who's been in insure
tech since 2013.
How do you see the evolution ofinsure tech?
Because I got involved ininsure tech beginning of 2016.
And I have a lot of thoughtsabout how insurance is going to
make for 2016 to 2020, like 19to 2022.

(02:09):
And even now, like if it's thesecond wave or the third wave or
whatever.
But you've been in it muchlonger, right?
So how do you see the phases ininsure tech for insurance
evolution?

Speaker 2 (02:19):
Yeah, I think the first wave back in 2013 was
similar to other financialproduct categories.
I think we're just gettingpeople comfortable with
transacting and buying insuranceonline.
I think insurance traditionallyas you know, the average broker
today is still 60 years oldkind of meets a due in person.
So it's very uncommon and maybenot intuitive for people to be

(02:39):
able to self-select theinsurance policy.
So that education process anddiscovery was invicting in kind
of the 1.0 of insure tech.
So that was probably 2013, allthe way to 2017.
And I think after 2017,consumers got more sophisticated
, right.
So the second wave it goesbeyond discovery to price
comparison and finding what'sthe best policy for them.

(03:01):
So I think that's the secondwave between 2017 and 2021.
And now I think, with thismarket correction as a whole
within tech, I love the newfocus on profitability and
strong unit economics.
So I think we can dive intothis a bit later, ali.
But a lot of people arerealizing that single line
insurance products are verydifficult to scale, that the

(03:23):
unit economics might be prettyhard.
That's why a lot of thetraditional carriers MGA always
have multi-line products thatthey sell.
So a lot of the innovation, atleast in the last couple of
years is around kind of productfumbling, thinking about better
targeting across differentfinancial products.
So how can you bundle maybe, amortgage product together with

(03:43):
insurance, together with, maybe,trust and will?
So I think a lot of theinnovation now happens in being
able to use data and provideeach individual the exact suite
of financial products that theywant, and insurance obviously
serves as a very big categorywithin the suite of products.

Speaker 1 (04:01):
Very interesting.
So maybe we'll just jump intomy favorite question always,
which is what's one or two ofyour most controversial opinions
or ideas about insurance,something that you feel like
it's different than how themajority see insurance?

Speaker 2 (04:18):
Yeah, I actually think we need to go back and
empower these offline brokers.
I think, instead of competingwith them, I think we should
empower them.
I think what we see now is alot of these relationships with
offline brokers are still verypersistent, especially the older
demographic, and insurance is alot about trust.

(04:38):
It's one of these core productsthat a lot of people want
dedicated advice.
So, instead of competing withthem, we've actually invested in
certain platforms, like acompany called Fair Street that
helps Medicare agents kind ofimprove how they operate, and so
we're not competing with them.
We're not creating a new avenuefor carers to sell products.
We're actually empowering orsupercharging the existing

(05:01):
channel.
So that to me is like a veryinteresting category and we want
to do more investments insimilar spaces.

Speaker 1 (05:07):
So I wonder, because I think about this all the time
when I was thinking about Kovu,as you spoke about this briefly
but I'm very passionate aboutthe distribution space because I
believe transformation ofinsurance starts with
transforming distribution andhow do we advise people, guide
them, educate them to make thecorrect decisions?
That's, at least to me, likethe right way to start with

(05:30):
distribution, transformation ofinsurance.
But when I think aboutdistribution channels, there's a
huge controversy aroundembedded versus brokerage model,
or you know, between, I Wouldsay, the hottest topics.
Now I know you were talking alittle bit more about how
insurance gets involved and ideaof bundling with like financial

(05:52):
products and all that whichsounds more like embedded and to
me that's a little bitdifferent than like empowering
workers.
So how do you see, how do youlike navigate this?
Like the difference betweenlike, is insurance going to be
more focused on bundling or isit going to be more focused on
like, empowering the traditionalagency channel, on evolving
them?

Speaker 2 (06:12):
Yeah, I think it's a bit of everything.
Maybe maybe I'll break it downto three different parts.
I think embedded it's veryinteresting and maybe a
controversial view I have isthat everyone wants to be an
embedded insurance company, likewet-cell insurance, but I don't
think a lot of companies havethe right data right.
So, ultimately, if you want tobecome an embedded insurance
company where you can actuallycapture value, if you're willing

(06:33):
to underwrite some of the riskyourself, I think if you're just
going to distribute thesepolicies, you're no better than
an affiliate model.
So that's something that hasbeen tried and tested in the
past.
I think if you look atcompanies, for example, maybe
I'll name a company called hintAI, so Tim dot AI works in
marketplaces like rental carmarketplaces, like two row, but

(06:55):
there's actually a lot of datathat they're collecting about
Individual drivers and theirdriving records and experience
and they I think that's a uniqueadvantage for beyond Kind of
what the insurance carrierspublicly see you have a lot of
granular data that will informyou of how to better price the
policy.
So in that case, I think it'svery interesting for an embedded
insurance play.

(07:15):
Where's in others cases wherethe data might be more sparse,
so you're like a middlemantransactor where you actually
don't capture a granular leveltransaction data, then that that
case, I think there's less of awedge and or there's a less of
a long term mode for you to beable to better price risk within
that.
So that's the embedded category.
I think that's prettyinteresting.

(07:37):
I think the the other one that'spretty interesting, as we
talked about, being Empoweringexisting Brokers, existing
channels.
I think there are twocategories.
I break them down.
One is like a pure play let'sprovide more SAS tooling to
these existing channels.
The other is like a lot moredistribution.
We recently invested in acompany called outmarket that

(07:59):
helps a lot of carriers identifygood distribution partners.
So they identify Brokers thatthe carrier currently doesn't
work with.
Our MGS.
There are wholesalers that thecarrier doesn't work with today
but could be very strategicgiven the carriers footprint or
the carriers intent to expandtheir coverage in certain
regions.
So I think this model of like aAI co-pilot is becoming very

(08:21):
interesting when you think aboutEmpowering people with an
ecosystem very interesting.

Speaker 1 (08:26):
So let's just flash forward to, like, let's say, 20
years from now.
I know it's a little bit of astretch, but I just wonder, like
, how do you see the future?
Do you just distribution wise?
Do you see it as a market in 20years that is more dominated by
, I don't know like AI enabledbrokers, or like a model where
insurance is taking a bat'sbackseat of all these big
commercial brands?
Then is more just bundled intodistribution.

(08:49):
I don't know you buy your carinsurance from Toyota versus you
know like your broker.

Speaker 2 (08:56):
I really hope that.
I think, with the wealth ofdata and the cost of creating
these AI systems going down alot, that I think the Analogy
you gave about hey, you buy yourToyota and then, yeah, I
presented if you're your carinsurance and then I get your
physical address because I'mtaking down your driver's
license and your details and Ican offer your policy because
now I capture data about Kind ofwhere you live and kind of what

(09:18):
you do for a living and kind ofyour lifestyle trend, so I
think all of that is going to bethe future.
I don't think that's somethingwe get to in the next three to
five years.
It's probably gonna be 15 to 20.
The other thing which I thinkit's going to be interesting is
that, given a lot of the, theinsurance education can be
automated, that you can embedinsurance within other financial

(09:40):
products.
So, instead of just a cardealership, you walk in the bank
, you open a bank account Maybeit could be online now an online
account and at the same time,get asked you a few questions
that can prompt you to buyinsurance as Well.
So I think we see the bundlingof products not just with non
financial players, but also withother financial products.

Speaker 1 (10:00):
So I'll tell you my thesis and I want you to
basically correct me and say,like Ali Sure, I've thought
about this quite a bit andthere's two versions of the
features that I could see.
One version is that you havesomeone who's your risk advisor
that understands the risk you'reresponsible for and tells you

(10:20):
how to predict your downside.
And the other one is that theone that you mentioned, which is
you buy a car and then the carmanufacturer says I don't know.
Let's say, if I mean if youhave a total of cars, then
you're not driving, so it's notyour liability or risk.
Is the cars liable to you andrisk?
In that case they shouldprovide insurance, not you.
But if it's your risk, then Iguess the risk advisor makes

(10:43):
more sense to me then.
So I mean, we always talk aboutthe idea of bundling, but I
wonder why shouldn't ideabundling come from this angle,
which is like a holistic riskadvisor, as opposed to bundling
life insurance with mortgage orbundling I don't like auto with
the car manufacturer?

Speaker 2 (11:02):
Yeah, I love it.
I like the idea of a holistickind of risk advisor to help you
with like your entire life andyou're like everything else.
I do think that the one thinghas to change and maybe this is
not just insurance but otherfinancial products that
everyone's incentive buys to getyou to buy more financial
products.
So how do you take that away?
I think if a financial advisorsells you more insurance, you

(11:26):
take a higher level of coverage.
They ultimately make morecommissions.
So as we think about the riskmodel, like I think we need to
change kind of the commissionstructures or how we think about
compensating people withdistribute insurance so that the
end consumer gets what's rightfor them, not what's most
expensive for them.

Speaker 1 (11:44):
So we build an AI risk advisor, because that was
my original thesis and the ideawas a holistic, unbiased and
transparent advisor.
The idea that I have had, atleast as it, was that it's
supposed to be algorithmic.
So if it's algorithmic and hedoesn't have a human
intervention and it's kind oflike open source, so you know,
like what is recommended to what, that should solve the issue.

(12:05):
Because the combination that Ilike the most was that the AI
generates advice and then thehuman does the hand holding for
explaining it, because theexplanation I don't think AI I
don't know AI should be thereeventually, but people still
want that trust and human touchand human hand holding when it
comes to like these kind ofcomplex products which I think
might be able to solve that, tokind of provide some of that

(12:26):
transparency and some of thatunbiased incentive into the mix.

Speaker 2 (12:31):
Yeah, I think that's super interesting.
Maybe one kind of earlyinnovation of the space is that
we found that some of theseearly shopping tools for
insurance have taken off, likeplatforms like SaveBot.
I think what's interesting isthat I feel the incentives there
are more aligned with theconsumer.
So I like it that you'rerunning an algorithmic model in

(12:52):
this case, where you take awaythe biases and you're
recommending what's best for thecustomer.
So I think maybe there's ahappy medium or a compromise
that we can strike right becausea pure shopping tool again, you
don't want to just get thecheapest insurance policy.
You want what's best for you,but you also don't want to like
over commit the policies thatyou don't need.
So I think something in betweenis kind of what you're building

(13:13):
is definitely going to be thefuture for how people should be
thinking about insurance.

Speaker 1 (13:18):
Yeah, and, of course, the biggest issue here is that
people's relationship with riskis very complicated.
I think people are wired to buycertain policies, which is auto
, home and life, maybe to someextent, and they're not wired to
buy other policies likedisability insurance or all this
stuff that they're not used tobuying.
And convincing them to buysomething that they're not
generally wired to buy it's avery weird dynamic because it's

(13:41):
not a logical discussion.
It's like when I tell someonethat you have a 25% chance of
having some sort of disabilityin your life, they're like okay,
so there's a 75% chance,nothing's going to happen to me,
and that's like how a lot ofpeople treat it versus yeah, but
that's like the whole point ofinsurance is like we're talking
about, like low frequency, highservers, the items, and that's
what insurance is for.

(14:02):
So I think that was the biggestchallenge we ran into and
that's why I thought like thehuman advisors still need it to
kind of massage everything ontop of like unbiased
recommendation from anintelligent system.

Speaker 2 (14:13):
Yeah, I think the other thing, too, that will be
interesting is that I thinkthere'll be more product
innovation on the insurance side.
I think.
One example, ali, I know youand I have had offline
conversations about this, but Ithink the whole topic of
insurance for AI, I think it'sgoing to be very interesting.
I think a lot of people aretalking about AI for insurance,

(14:35):
but I think the opposite is veryinteresting If you think about
certain categories.
If I'm a medical practitioner,now I'm relying on AI to meet
memo grams for me.
Or if I'm a lawyer and using AIto drop drop letters for me to
some of my clients.
What if there's a mistake,though?
Is this covered by my generalliability policy, or do I need

(14:57):
something additional in AI?
So I think this is a veryinteresting product innovation
that will need to exist, atleast in the coming years.

Speaker 1 (15:05):
So I know tech ENO has existed, which is basically
right, is like you develop atech and you're covering its
errors and emissions.
I think this is just a morecomplicated tech ENO and the
question is Because theinsurance companies who write
tech and you are writing thesetech and those and I don't think
they have the money models todo a better job with writing it.

(15:25):
So the biggest losers in thespace are like when insurance
companies write this and thenrealize shit like that's like a
lot of you know.
So I think the so from acustomer lens, I feel like the
products are out there as techyou know from an insurance
company lens, I feel like theyneed to just be adapting much
quicker now with all these newsystems.

(15:46):
To give you an example, like RAI, we have shown that it could
pass the licensing test, like 5%accuracy or whatever, and we
haven't even spent a lot of timeon it, like to improve it.
So for me, I asked the questionfrom a few commissioners and
like the regulatory space andinsurance is like look, this is
as good as like a licensed agent, like so what?
Like can we start using it likea licensed agent or how would

(16:09):
that work?
And obviously, I don't thinkthe world is equipped to be able
to answer these kind ofquestions and I feel like
there's going to be moreregulations coming into it,
there's going to be morelearnings, but I think the
foundation is for sure there,but it should be super exciting
what comes out.

Speaker 2 (16:25):
I think it's a lot of insurance companies, especially
the larger ones, think aboutwhere they want to spend money
on future development.
I think one huge benefit with AIis, I hope, that a lot of the
RPEX costs whether it's likecustomer service, whether it's
billing all of that should bestreamlined away and a lot of
the focus should be on futureproduct innovation or really

(16:46):
like understanding.
Like hey, how do we actuallytailor our policies to each
individual?
Maybe the other, the other pipedream I have is that even
though everyone tells you nowyou get a personalized quote for
insurance, you're still bucketthat into a few categories.
They're placing you, they'rechecking a few boxes, or you're
within this certain age group 40to 55, you live in these cities

(17:07):
or zip codes and will give youthe policy.
But it's not down to theindividual level and sometimes I
think insurance companies alsodon't have a holistic view of
like.
I think the whole idea of arisk advisor is pretty
interesting, like if you'recovered with maybe some of your
company policies, maybe yourpersonal insurance policy
doesn't need to be that high.
So I think this understandingor common database that's shared

(17:30):
across all insurance companiesbut at the same time remains
anonymous, might be superhelpful in helping the end
consumer get better pricepolicies.

Speaker 1 (17:39):
Makes sense.
So let me ask you this what doyou think investors got wrong
over the past few years?
And then I'm just saying, like,where I'm trying to go with
this conversation is after it'sgoing to be like what do you
think investors are gettingdrawn now?
And then I'm going to ask youabout what insurance companies
are getting wrong.
So let's start with whatinvestors and insurance

(18:00):
companies to some extent, havegot wrong before.

Speaker 2 (18:04):
Yeah, I can get overestimated the lifetime value
of each customer and the numberof additional products they
could upsell to each consumer.
I think there was a lot ofconfidence maybe some cases
blind confidence that here westart with one product, we can
upsell you three other productsand a lot of the VC dollars that
flowed into the space and againwe are guilty as charged to get

(18:24):
people we've done the samething that a lot of the
assumptions you're making on howprofitable each business could
be were based on theseassumptions that don't hold.
So I think the other assumptionwas that customer acquisition
costs for these people wouldhold constant at scale, which is
not true.
I think insurance is just likeevery other financial services
both the go, where at once youget to a certain amount of scale

(18:46):
, the cost of acquisitionactually increases a lot, so
that you need that buffer fromday one in order for you to find
a viable product and scale.
So that's what I think bothcarriers as well as insure texts
as well as VCs have gottenwrong over the last couple of
years, and a lot of it, I think,wasn't uncovered till we had
the market correction, wherepeople started to pay attention

(19:07):
hey, there's no more free $50million lying around for your
next round.
Like we really need to focus onmargins instead of just
focusing on popline revenuegrowth.
So before that was just poplinerevenue growth at every expense
right, which is not the caseanymore.

Speaker 1 (19:22):
Let me ask you this do you feel like these models,
the market?
Because, like, I'm not going tomention any names, but a lot of
people said like these marketcorrections.
It was more about like the like, just the price multiples.
But the question is that do youfeel these companies can save
themselves or do you feel likethey're fundamentally flawed
businesses?

Speaker 2 (19:43):
I think a lot of them are fundamentally challenged
businesses as standaloneentities, so I think they would
need to merge with anotherentity to get to that
exponential scale, and scale notjust in terms of revenues but
back office footprint, forexample.
If you combine two of thesecompanies, we could combine
operations where you couldstreamline probably 50% of the

(20:05):
cost.
So I think in order for some ofthese players to succeed, we
would need to see drasticconsolidation in this market,
which hasn't happened yet, butI'm predicting that it will in
the next one or two years.

Speaker 1 (20:19):
When you say these players we're all talking about,
like you still MBA slashcarriers right, like that's the
right.
Yes, yes, that's right.
So is consolidation betweenthem and other new stock
carriers, or is it going to bemore between them?
And I don't know liketraditional carriers.

Speaker 2 (20:35):
Yeah, it could be traditional carriers too, I
think.
If there's a very complementaryfootprint say they have a
strong hole in certain productlines, then I could see kind of
that.
I see vertical integrationbeing very helpful, but I think
horizontal integration, which iswhat we were talking about
initially, I think that it'salso a very viable path in order
to create kind of a much biggercompany.

Speaker 1 (21:00):
Makes sense.
And what do you think peopleare still getting wrong now
versus before.

Speaker 2 (21:07):
I think where what people are still getting wrong
is that a lot of investors arestill funding single line
insurance products and I justdon't think that single line
insurance products scale verywell.
I think these are likely goingto be.
You're going to tap out at, Iwould say, like 50 million in

(21:28):
revenues and then finding thatkind of second chapter in your
playbook is going to beextremely difficult.
So I think I think, from dayone, a lot of people are still
not focused on the rightbusiness models, are not
thinking about hey, instead ofbuilding one single product, how
can I build a suite of productsthat would serve the end
customer?
So I think where they getthings wrong is they, instead of

(21:50):
being customer led in theirproduct discovery, they're just
being led by a single line ofproduct.
I think you really need tothink about the customer
demographic you're serving fromday one and understand like, hey
, what else can I sell thisspecific customer?
I think that's a moreinteresting question to answer,
but a lot of insured techcompanies in the seed stage are
still not focused on that.

Speaker 1 (22:09):
The challenge that I see with that is that I just
feel like the model of sellingmultiple products at the same
time in the online channeldoesn't work, and even from an
embedded model doesn't work.

Speaker 2 (22:26):
I mean look to chat more though what examples you've
seen that happened, Because Ithink we actually think the
opposite with some of theseexamples.
We think would actually getskilled.

Speaker 1 (22:39):
I mean I spend a lot of time with customers trying to
understand how they work.
I mean I have customers indifferent categories, but one
category is of people who buyonline.
Let's talk about threedifferent channels.
One is B2C.
Imagine someone like Branch whodoes home on auto together.
I don't know if I'm a big fan ofa direct B2C play to go to

(23:00):
someone and say, day one, bundleyour home on auto with me
directly.
I just feel like the personwho's shopping online typically
is looking for one policy andthen try to sell them to
multiple things at the same time.
It's that much harder.
It's from a high-intendcustomer that you'll find online
.
If I'm talking about a bankcustomer or, let's say, a

(23:24):
financial product like 8 pointsor something, I'm trying to sell
them home.
Again, these are hardconversions as they are for
single line and high-intention.
And convert them in multiplepolicies at the same time is
even harder.
And if you just sell them onone policy, trying to bundle a
few other things after is alsovery hard, because most of these
customers you have very lowengagement with and there's not

(23:44):
much.
That's one of the reasons I gotinto the agency space anyway,
because it felt like in theagency space.
The renewals are really wherewe want to go after people for
bundling, because you basicallybring them in on a single line
and then you have thatrelationship with them that is
engaged, and you have thatpersonal relationship with them
as opposed to like a Acornsrelationship or banking
relationship or something, andthat's what you use to cross

(24:06):
sell.
At least that's my thesis.
I don't know if I'm getting itright or wrong, but that's how.

Speaker 2 (24:11):
I do it.
I definitely can see your angle.
I think what I mean maybe topiggyback on the example, like
maybe auto and home, it's notmaybe naturally like a
complimentary product, unlessyou bring in maybe a third
product like an umbrella product, and you can say like if you
get these two from us, yourumbrella premium goes down.
So I think, instead of sellingthem three policies at once, I

(24:33):
think there's a short-termopportunity to resell policies,
so getting them to go up out ofthe current policy and pick
something else in the short term.
And the other thing that'schallenge about selling multiple
products is the underwritingtimelines are not consistent.
So maybe that beckons anotherchallenge we have in today's

(24:53):
insurance industry.
So I think, if you think aboutlife insurance, it's so archaic
that a lot of it still requiresus getting a blood test and
everything.
Well, if you think about itlike, okay, most of us have gone
for medical checkup within thelast one or two years, right.
So I think we can rely on thedata and I think a lot of us
subscribe to online platforms,like one medical, for an example
.
So I guess why isn't there moreof a data integration between

(25:15):
this?
Why can't get an instant policy?
So I think a lot of thechallenges is also the data and
switch right.
You come back with a very lowpremium number and then two
weeks later they come up withcertain exclusions, so like,
okay, I can do this.
I think being upfront at thestart and getting pricing right
and underwriting periods shorteris going to be like the winning
formula, but I don't reallythink we've cracked the code yet

(25:37):
.

Speaker 1 (25:37):
Yeah, and why do you think investors are still
finding single line insuranceproducts Like is there, like
they haven't learned the lessonor they're betting on a certain
thing.
That?
What are they betting on?

Speaker 2 (25:49):
Like insurance, is such a such a big category.
So a lot of people think thatthe TAM, even with single line
products, is big enough.
So I think a lot of them arefocused on the TAM.
And I think the second thing isthey're focused on people who
have that experience in thespace as corporate executives
and like, hey, we know the spaceinside out, we have all the

(26:09):
right relationships, you'regoing to get the right
distribution channels, butfundamentally, are you building
a business that works?
So I think, as they think aboutsupply, they love investing in
people who have all the rightrelationships to all the
carriers and MGS.
But I think that's still thatconsumer acquisition fly we also
talked about is something thatis still very hard to fix.

Speaker 1 (26:29):
And another thing that I wonder about is that
before, when you think aboutroot or lemonade, people were
basically giving SaaS multipleson premium.
They counted premium likerevenue and like a 90 SaaS
multiple on like the low margin,unprofitable premium.

(26:51):
I just saw the announcementfrom Kin insurance that they are
unicorn now and I wonder whatare the multiples now for like a
single line in J or carriers?
It goes before.

Speaker 2 (27:04):
Yeah, it's much, much lower.
So maybe a lot of thebackground is that these first
batch of companies like RuthEliminate, as you mentioned,
these were, like, I think, thefirst in 10, 20 years of a new
insurance company coming on thepublic market.
So a lot of, I think, lessinformed public market investors
this is like the retail crowdthey didn't really understand

(27:26):
the difference between premiumsand regular SaaS multiples.
So I think revenues is like asmall fraction of premiums that
everyone finally understood.
So I think, as you look at thenewer companies, a lot of these
new companies are going to bebay-valued off kind of net
revenue multiples, which is thengoing to be more consistent
with SaaS multiples.
But if you look at liketop-line kind of revenue

(27:47):
multiples, I can't see anyonetrading for more than like three
, four, x kind of premiums,which is which is sounds crazy,
which is still very high in myopinion.
But I think everything is goingto be focused back on like
bottom-line revenue numbers.

Speaker 1 (28:02):
Yeah, but even the bottom-line revenue number is
not a SaaS margin business.
It's still like you know.
You have a huge customerservice and all that kind of
stuff.
So I would assume they're stillrunning at 30, 40% margin and
if they carry any risk, theyhave like huge unprofitability
issues now with sorry.
So it's still interesting, butI mean, can being insured to
unicorn?
I mean props to them in thismarket in that category.

(28:25):
I think I was here that news,but let's see, I wonder what,
like what are the multiples onthem?
So you spoke a little bit aboutlike areas of insuring that
you're excited about.
One of them was the idea ofensuring AI, which is a very new
trend.
Anything else that you're superexcited about these days.

Speaker 2 (28:44):
I think we're very excited again.
We, I think for the last, Iwould say the last five years a
lot of people have focused onmore consumer innovation within
insurance.
I think we are now more excitedabout helping a lot of the
existing players largewholesalers and carriers fix
their back office operations,which is something very

(29:04):
interesting, I think.
One big innovation again not tokind of go back to AI again,
but I think in the past we wereonly able to leverage our PA
technology, so a lot of theinnovation was with regards to
like okay, we can ingest datafaster and process data faster.
So we can ingest a PDF form and, okay, spit out the data into a
database.
But now I think we call itsmart processing or smart

(29:27):
contextualization.
But not only can you capturethe data, the data can be used
to inform other decisions, andthis is only kind of made
possible more recently with AI.
So a lot of this informationthat we're pretty excited about,
like the example I told you, weinvested in this company called
Outmarket.
That's a, it's a co-pilot for alot of carriers and wholesalers

(29:48):
to improve their distributionreach by finding existing
brokers that could actually helpthem and could be additive to
their current portfolio.
So I think that's going to bepretty interesting.
So the areas of focus was a lot, I would say, like, back office
, a lot of distributionautomation, a lot of customer
service automation, a lot ofreceivables automation.

(30:09):
That's all pretty interestingto us today.

Speaker 1 (30:12):
And what are some of the trends that you feel like
now have a huge impact oninsurance just outside of the
insurance space that peopleshould be aware of?
Do you expect like a big impacton insurance or insurance?

Speaker 2 (30:26):
Yeah, I think.
I think a lot of the trends inthat are happening more on the
macro level are going to make animpact on insurance as well.
Right, I think if you thinkabout the cost of capital for
all businesses, it's gone up alot.
So I think, as you think aboutthe whole underwriting and the
MGA model, I think kind ofpeople who are sitting in the

(30:46):
wrong part of the stack aregoing to get compressed as well.
I think the other thing that'sgoing to affect the insurance
industry a lot too is, just nowthat there's more data about
understanding premiums, a lot ofthe carriers are quickly
finding unprofitable markets.
I mean, I'm sure you know this,but State Farm kind of moved out
of California as well for theirhome insurance.

(31:07):
They're realizing like, okay,all these policies that we
underwrote, now we have data onit, they're like gee, not that
profitable, especially with allthe fire risks that now we can
predict.
I think before that a lot ofthe wildfire risk was not really
factored into these models.
I think with the data a lot ofthe carriers are getting a lot
smarter.
So I think there's going to bekind of, I would say, a more

(31:28):
variation in pricing which isgoing to affect the insurance
industry.
We're not going to see ahomogeneous kind of like set up
for most of these product lines.
We're going to see very bigvariability and I think a lot of
underlying kind of risk takerswithin the space people
underwriting the risk are goingto ask more questions for them
to price the policies.
So I think that's going to bethe evolution of insurance.

Speaker 1 (31:51):
And who do you think are going to be the biggest
winners in this?
When do you think about thefuture of insurance and all
these trends and changes and soon?

Speaker 2 (32:00):
We're going to be the biggest winners, I think,
people who could get the scale,and I know it's a cop out, but I
really like the rollout modeland the insurance.
I think that's very, veryinteresting.
So I think that there's a groupcalled XPT Insurance which is
out in New York.
So they've been buying a lot ofspecialty wholesalers and this
is more traditional kind ofoffline, but they've done

(32:20):
extremely well.
I think what you realize isthat the back office operations
and the distributions are verysimilar across different types
of businesses but none of themhave ever got the scale because
insurance is still a veryoutside of the top kind of few
carriers, wholesalers and MTAsstill a very, very fragmented
market.
So I think this whole strategyof a rollout play in insurance

(32:41):
it's going to be a veryinteresting one.

Speaker 1 (32:44):
Very cool.
That's it for most of myquestions.
I'm going to ask about thetrends and so on.
I don't know if you haveanything else that you want to
share or add that we haven'tcovered.

Speaker 2 (32:57):
No, this is super fun , ali.
I had a lot of fun chattingabout insurance and I know a lot
of different trends havehappened over the last two years
, but still very excited aboutthe future of insurance.

Speaker 1 (33:07):
Perfect, no well.
So thank you so much forsharing all your insights and
thank you everyone for listeningTX.
It was great having you on thepodcast.

Speaker 2 (33:18):
Thank you again, Ali.

Speaker 1 (33:20):
Thanks, thank you.