00:00welcome to the a 16z podcast I'm Michael
00:03Copeland we are sitting here in our
00:06somewhat cramped podcast room with Frank
00:08Chen who runs our deal team and Mike
00:10Paulus who is on the deal team welcome
00:12guys thanks so much for having us great
00:15last year around this time you wrote a
00:17piece for our 16 things package and it
00:21was about insurance of all things and
00:24you know it was about this trend for the
00:27future and it turns out actually of all
00:29the things and/or I won't say of all the
00:31things but of some of those things it
00:33got this incredible response and Aulus
00:35and there were waves and waves of
00:37startups in insurance I have to say I
00:40was the most surprised to see the
00:42reaction to this person in fact Mike
00:44joined us after having read the post and
00:47saying hey look like it's an adventure
00:48for am interested insurance so insurance
00:51look I personally pay for insurance I
00:54don't get excited about insurance but
00:56what is going on why are there
00:58entrepreneurs and people like yourselves
01:00focused on what seems to be a fairly
01:03high bound and you know explainable
01:06obvious kind of thing well let me go
01:09back to the post and the original
01:10motivation which is you're right if you
01:12think about you're in your relationship
01:14with an insurance company I remember
01:16buying life insurance I literally filled
01:18out forms in triplicate after they wrote
01:21the policy they sent me letters that I
01:23swear were generated by their mainframe
01:25application because it comes in one font
01:27and it's all caps and that was the only
01:29app that the font available when they
01:31had the mainframe application and I
01:33really don't have any other interaction
01:35with them once a year I get this
01:36mainframe generated app that said a
01:38letter that says send me your premium
01:40and so if you think about the
01:42opportunity to completely reinvent all
01:44of that we can remesh with data we can
01:50reinvent how they find me as a customer
01:52we can invent the types of products that
01:55they sell me we can rejigger the prices
01:58that I'm paying for those policies we
02:01can reinvent how I change claims so if
02:03you just think about it as a start-up
02:05entrepreneur might think about it which
02:06is let's change every single aspect of
02:08the relationship an insurance company
02:10has with their insured let's just do it
02:12and I think that's a lot of it
02:14we're used to at the front end sing that
02:16you know the the Geico gecko and so
02:19there's a marketing aspect to it but I
02:21would I want to get deeper into what
02:23else can be changed and what the
02:24components of insurance are that in our
02:27software eats the world way can be
02:29attacked by software I think the really
02:32interesting thing about insurance is
02:33it's the first big data problem really
02:35what an insurance carrier is trying to
02:37do is evaluate you as a risk and
02:40traditionally it's been done based upon
02:42a lot of factors they don't evaluate you
02:44but someone kind of like you so what is
02:46your age what is your zip code these
02:48sorts of things but with the explosion
02:50of data and with the explosion of this
02:52this could this mobile phone that we
02:54always have with us we can really say
02:56who are you what's your activity what is
02:57your risk level and allows you to price
03:00someone in a way that's a lot more fair
03:01to them you can help them actually
03:03prevent claims and that's one really
03:06interesting thing about insurance
03:07companies is if they can use all this
03:08data to help my house not getting broken
03:11into or or help me live longer we're
03:14perfectly aligned there so it's a it's a
03:15really interesting opportunity both to
03:18better understand who you're insuring
03:20and better pricing and pricing in a more
03:22fair way and then also helping make
03:24people's lives better why is it that the
03:27only interaction you have with your life
03:28insurance company is basically their
03:29demand for your annual premium right
03:31because if I'm in their risk pool we are
03:34perfectly aligned I don't want to die
03:35they don't want me to die but the
03:41insurance company of all the customer
03:42company relations is I have they really
03:44don't want me to die right because then
03:46they'd have to pay out of policy and so
03:47why aren't they at the forefront of
03:49encouraging me to live a healthy
03:50lifestyle why aren't they sending me my
03:52jawbone of course they should be right
03:55but it's just it needs a change of
03:57thinking and I think it's gonna need
03:58startups to show the way of what
04:00insurance companies ought to do in their
04:02ongoing customer journey as opposed to
04:04the existing companies doing it so for
04:06many of the startups that come through
04:07here and that we talk about and think
04:09about data is at the core of things it
04:12would seem to me that one of the
04:14advantages that these incumbent
04:16insurance companies have is they have
04:18all the data right they had this
04:19historic you know if you're some large
04:21insurance company you've been doing it
04:22for decades you know if not centuries
04:24you know go back to insuring you know
04:28or whatever Lloyd's of London yes is
04:31that advantage no longer such an
04:34advantage or is our startups and the way
04:37that they collect data able to sort of
04:38quickly catch up and also in some sense
04:40surpass what are you saying I think the
04:43data that were collecting today is very
04:44different things like your activity data
04:47from a mobile phone we've we've never
04:48had before things like cyber risk that
04:51have a very different profile in terms
04:52of what makes someone a good cyber risk
04:54you you can't look at that in the same
04:56way that you're trying to model out a
04:58catastrophe like a hurricane
04:59so as we look forward that the types of
05:02data that you can get around individuals
05:04are very different I would say the way
05:07to think about it is the data that
05:09insurance companies had will become less
05:11and less important over time as these
05:12new data streams get unlocked right so
05:15the old insurance company asked me in on
05:18paper in triplicate whether I was a I
05:20had ever piloted a plane literally this
05:23is one of the questions that you ask
05:24when you get life insurance my phone
05:26knows exactly when I'm going to get on a
05:28plane right and so why isn't that data
05:30being reflected back into the policies
05:34in fact why shouldn't it like when I
05:35booked that private plane trip why
05:37shouldn't the offer for a hey do you
05:39want to buy one time life insurance for
05:41you piloting this plane are you getting
05:42on this private plane of course it
05:45should right but insurance companies
05:46don't just don't think that way because
05:48historically they haven't been able to
05:50get that data therefore you can't
05:51generate the right price for that policy
05:53and you can't get it in front of the
05:55customer and now all of that has changed
05:57like you have the opportunity at the
05:59point where you get into a taxicab or
06:01you start your lyft driving experience
06:03or you get on a plane all of those could
06:06be insurable events so Mike and Frank
06:08you just mentioned sort of a new kind of
06:10insurance and Mike you mentioned cyber
06:12risk what are the new categories of
06:14insurance that you are seeing that are
06:16emerging because of our behavior because
06:18of the services that we're all using and
06:20because of the way we're honestly just
06:23able to do more I think that the sharing
06:25economy has been has been massive so
06:28uber and lyft and and the whole
06:30on-demand world have have really
06:32transformed and and and blurred this
06:35line between the commercial and personal
06:36driving worlds and there's a huge
06:38opportunity there same with your house
06:40you know the typical home owners
06:42you never thought that you'd be on
06:44Airbnb and be be wrenching out your
06:46house and then also be a consumer being
06:48another city and using Airbnb and hoping
06:50that you have property protection and
06:52things like that so the sharing economy
06:54has has really been tremendous and then
06:56also things like drones war men bought
07:00drones then then then I watches for
07:02Christmas and its really unknown what
07:05does it look like when the drone crashes
07:07into into the neighbor right new things
07:09like Bitcoin you know it's fascinating
07:12it's something like one third of every
07:13Bitcoin ever mined is is now lost but
07:16with some very common-sense protective
07:18measures it can be a very secure asset
07:20so how does how does something like
07:21Bitcoin get insured but there's a lot of
07:24folks out there that have wallets that
07:25they want to make sure that they have
07:26good insurance are you seeing any
07:28patterns I mean Airbnb is an interesting
07:30example because they're the ones who
07:32decided to sort of ensure everyone or at
07:34least the the hosts right and the guests
07:36who I guess to a certain extent and so I
07:39as an Airbnb customer I don't have to
07:41worry but are you seeing the kind of
07:44responsibility shifting because of any
07:47of these new trends yeah so if somebody
07:51if you're an Airbnb host and your house
07:53gets trashed right then Airbnb is
07:55insuring that the losses associated with
07:58that right so that's not something that
08:00your traditional homeowners insurance
08:01did it's part of the price of doing a
08:03transaction with Airbnb all right so
08:05existing products like home insurance
08:07are getting sort of modified for these
08:09new sharing economies and then we have
08:11it entirely new insurance products right
08:13so cyber insurance is basically for
08:16retailers who are afraid of getting
08:18hacked because they're reading the
08:19stories of everybody else getting hacked
08:21and they're looking at the costs
08:22associated with cleaning up offering
08:24everybody identity insurance those are
08:27real costs to the business not to
08:29mention the lost sales from people who
08:31no longer trust your online site and so
08:33this is going to be one of the fastest
08:35growing categories of an entirely new
08:36type of insurance which is I will insure
08:38you against cyber losses right and I can
08:41imagine the pricing and the risk
08:43assessment just must be hard to fathom
08:46honestly absolutely and so we're seeing
08:49startups trying to fill that void which
08:51is let me bring tools and data to this
08:53process so that people can write these
08:55policies intelligently
08:56because you're right how much will the
08:59policy cover how much will the losses
09:01actually be I like to joke that you know
09:03I get notifications all the time from
09:05the places that I've shop saying oops we
09:06lost your data here's your free year of
09:08identity insurance and I think I'm gonna
09:10get an identity insurance for the rest
09:11of my life one year at a time one hack
09:13at a time right so there's costs
09:15associated with that there's costs
09:17associated with plugging the security
09:19holes there's costs associated with
09:21hiring a brand new insurance team
09:22there's cost associated with lost
09:24customers so how far do the policies go
09:26and then how much should they cost it's
09:28also interesting that we see really a
09:30convergence of things like cyber risk
09:33into Property and Casualty traditional
09:36things like a fire flood something like
09:38that so you know we had in Germany the
09:40first factory that was attacked in a
09:43cyber way and it caused physical damage
09:45but as we see an explosion of connected
09:47devices where all of the machines in
09:49your factory are connected to the
09:50all the sudden that fire that flood or
09:53that shutdown or that interruption of
09:55your factory is as likely be caused by a
09:57hacker as it is to be caused by weather
10:00or something like that so it's not
10:02necessarily kind of new frontiers but
10:04this convergence if my if my car drives
10:07itself then that cyber policy is going
10:10to be the most important form of auto
10:11insurance that I have do you think in
10:14the same way that let's say I have a
10:16house that's on a hillside that's prone
10:18to landslides or it wasn't built to code
10:20its uninsurable because it doesn't meet
10:23certain standards do you feel like
10:25there's going to be standards that
10:27emerge around cyber risk and then let's
10:29get to the question of driverless cars
10:31um if you don't have XYZ in place the
10:34pillars the sort of underpinnings of
10:37good cyber security you know behavior
10:40you won't get insurance or will we see
10:43these things emerge as standards do you
10:45think I think standards will emerge over
10:47time and this is sort of what a lot of
10:49startups are trying to work on which is
10:50let's say you were the insurance company
10:53and you wanted to insure your favorite
10:55retailer against cyber losses an
10:56intuitive thing to assess would be well
10:59how vulnerable are they right you'd use
11:01all the tools that hackers would use to
11:03try to figure out like how easy it is to
11:05break in right and once I've broken in
11:06like what information can I get right so
11:09I think there will be a
11:10emerging standards that let people
11:11assess what level of maturity these
11:14retailers have on their online presences
11:16I think it won't be government
11:18necessarily I think it will be
11:20commercial standards right that sort of
11:23puts you into a risk category right are
11:25you risk one or risk six right right and
11:28you can either pay X or out the nose
11:30yeah exactly and then how do you get
11:32from six to one like if you wanted to be
11:33in a less risky category what things can
11:35you do to protect yourselves and you'd
11:37expect insurance companies to take a
11:38much more active role right in the way
11:40that I was describing instead of just
11:42the one year cough up your premium
11:44here's that hey let's work together to
11:47get your security risk down and I think
11:49if you go back to the the really the the
11:51beginnings of commercial insurance
11:52they've always had that role because the
11:54insurance company is is really the risk
11:56expert and when done right they're not
11:58just someone you pay premium to do
12:00insure your risk but they're actually
12:01your active partner so you know Hartford
12:03boiler came out of the Industrial
12:06Revolution the start of steam boilers
12:08and there used to be explosions
12:09everywhere and there were a lot of
12:10fatalities and and they invented the the
12:13really kind of the first modern
12:14insurance policy and they said we're
12:15gonna inspect every boiler only if it's
12:18a really good boiler are we going to
12:19insure it and then we're going to
12:20continue to see if it's well maintained
12:22so not only did you have the birth of a
12:24massive line of insurance but you saw
12:26safety standards rise universally and I
12:29think a lot of companies benefit from
12:30their insurance broker that comes in
12:32whether it's practices of employment or
12:34your building or your car they're often
12:36the drivers of safety and partner so I
12:38think it's it's natural that as we go
12:40into the cyber world the insurance
12:42companies are going to be the folks with
12:43the most knowledge across the broadest
12:45set to sort of and have those standards
12:48but also make their clients a lot safer
12:49and I think that the insurance companies
12:51that that win in the in in in the next
12:55century will be those insurer least we
12:57and that they pay the most claims or
12:59they pay the fewest claims rather and
13:01there are the most preventive right we
13:03you've mentioned cars and driverless
13:05cars a few times this is one of those
13:07things that's a head-scratcher because
13:09who's liable if I'm not driving the car
13:11the cars driving itself you know who's
13:15the professional driver here it's not me
13:16that's exactly right I think I was
13:17looking at this New Yorker style comic
13:20and I know if it was actually New Yorker
13:23but it's a police car with another car
13:25that's been pulled over and the caption
13:28reads do you did your car know why my
13:29car pulled it over right like so right
13:32who's liable or imagine this rights it's
13:35only gonna be a matter of time before
13:37somebody comes in hacks your garage
13:38right your bluetooth-enabled garage
13:40lifts your garage door open and then
13:42hacks your car and drives it away right
13:45because you have a self-driving car
13:47so who's responsible in that case did
13:49the homeowner have liability because
13:51their garage door was so easy to hack
13:53and the car inside it was so easy to
13:55hack right is it the car manufacturer
13:57who's liable because Wow like that was
13:59just a easy security hole that you
14:02didn't plug when he manufactured the car
14:03so all these questions will get you know
14:05sort of actively decided as people write
14:07these policies as people pay for these
14:09policies and as the regulatory framework
14:11and the regulatory bodies get involved
14:13so it's gonna be fascinating to watch
14:14and I would say the the early trend has
14:17has been towards a policy for the
14:20manufacturer of the car
14:21so really Google Volvo and Mercedes have
14:24all said that for their cars while
14:26they're in assault driving mode they
14:28have the liability and then services
14:30like lyft and uber which will likely be
14:32the first delivery mechanism on the for
14:34the mass market for self-driving cars
14:36they're also covering that liability
14:38while there while they're in use so I
14:39think that that's it's more likely going
14:41to be the service provider or though or
14:43the manufacturer which which begs the
14:45question then this is what's interesting
14:46about car ownership etc if you know it
14:49might become incredibly expensive to
14:51actually be your own driver if the rest
14:53of the world is able to be insured
14:55because they're not driving their own
14:56cars you know I want to drive a car it's
15:00probably gonna in the future at least
15:01maybe cost me a fair bit yeah and why
15:04shouldn't it if if you know we're humans
15:06and we make mistakes and we're you know
15:09reckless drivers maybe it's a social
15:11good I I should save money which is the
15:13other way of looking at it if I if I let
15:15a much safer computer drive me around so
15:18where are we going to see our insurance
15:21kind of experience change first what's
15:23going to what's going to grab people is
15:25it going to be in the marketing and sort
15:27of the front end that all of a sudden
15:29were going to get new options how do you
15:30see this playing out for your average
15:32consumer yeah so I think what you'll see
15:34is you'll see insurance
15:36companies that deliver dramatically
15:38better customer experiences so how you
15:40were able to compare prices how are you
15:41were to make sure it was apples to
15:43apples so that will change I think you
15:45will see dramatic improvements in the
15:47way insurance companies do customer
15:49service which is submitting claims with
15:52iPhones taking pictures right will get a
15:55much more mainstream and then consumers
15:57will see new types of insurance products
15:59right so they'll get their existing
16:01products reinvented hopefully improved
16:03and then they'll see the new products
16:05that they hadn't thought about myself
16:08driving car insurance policy my might
16:10protect me against hacking my garage
16:12door opening and hopefully there's a lot
16:15of things that consumers experience but
16:18don't necessarily see so for example if
16:20I can use big data to more easily
16:22identify who's trying to defraud me as
16:25an insurer it means for everyone else
16:26there's a much better claims process or
16:28when I'm buying insurance on my phone it
16:30doesn't ask me any questions that it
16:32should already know because it's smart
16:34so just those you know like hopefully
16:37those those experiences that you have
16:39that are technological enable that are
16:40delightful in their in their cleverness
16:43or their simplicity does the University
16:46of insurance if if let's think about the
16:48Internet of Things if everything is
16:50smart and connected and therefore data
16:52is sort of ubiquitous does the universe
16:55can I insure anything at that point you
16:57know or in some ways bet on anything
16:59like do you see that happening are we
17:02still for the near-term gonna be focused
17:04on kind of the the more fundamental
17:07pillars of insurance that we that we
17:09know so in theory if there were more
17:11data freely available you ought to have
17:14new insurance products right to
17:15capitalize on that and I think of the
17:17the whether a company climate core did a
17:19series of experiments where you could
17:20buy insurance on weather events right so
17:23if I'm a farmer in Iowa and I want to
17:25say look if it's a hundred degrees seven
17:27days in a row that's very bad for my
17:28corn I want to buy a policy on that
17:30particular event it turned out not to be
17:33a great business for the climate core so
17:35I think we need to run more of these
17:37experiments because in theory we ought
17:39to see a Cambrian explosion of insurance
17:41types because data is so easy to get
17:44right we haven't seen that yet so and I
17:47think that the other really interesting
17:49sector of folks that are cut off from
17:51the insurance markets today you know for
17:53example insurance is very expensive for
17:56teenage girls to get for example because
17:58their male counterparts can be very
17:59irresponsible I know I know I wasn't a
18:01great driver when I was six but through
18:03telematics I can show hey I'm not the
18:06irresponsible 16 year old you're saying
18:08that teenage girls were treated as if
18:09they were teenage boys exactly and they
18:11were there's a company in the UK called
18:13drive like a girl and it said look you
18:15know we'll put a telematic device in
18:17your car if you can show that you drive
18:18like a girl I you're safer less
18:20aggressive will will save you a lot on
18:23your insurance so all the sudden a lot
18:24of people were able to save a lot of
18:25money I think a more serious example is
18:28diabetics today it's it's a huge part of
18:30our population and it's very difficult
18:32to get decent health or life insurance
18:33but the vast majority of diabetics start
18:36managing there they're following their
18:37doctor's orders you're exercising and
18:39they're actually pretty good risks and
18:41we can use wearables and devices now
18:43have to say hey look I might be diabetic
18:44but I'm actually a really good risk cuz
18:46I'm doing the things that that I'm
18:48supposed to be doing and all of a sudden
18:49those sorts of folks have access to the
18:51two insurance policies and markets so
18:54this idea of rewarding virtuous behavior
18:57right less risky behavior is great and I
19:00one of the really interesting trends is
19:02this is going to extend to algorithms in
19:04other words there are going to be
19:05ethical Pergamon that we have to do for
19:08our cars right so the product a couple
19:09example is look you're building the
19:11self-driving car and you're in the
19:12situation where you can make one of
19:14decisions right kill five kids on the
19:16sidewalk or kill 20 adults if you turn
19:20left right and so eventually we're going
19:23to need to program our cars to make
19:25these decisions and so the this idea of
19:28sort of ethical decision making right
19:30doing something that's sort of better
19:31for a specific audience is actually
19:33going to come back and influence
19:35insurance rates to what kind of ethical
19:36programming does your car have and
19:38therefore what kind of insurance is
19:40appropriate for it you what you need in
19:41that situation is the speed racer' Mach
19:44five button words and jumps over five
19:47kids and xx Alton like it's the problem
19:50solution so give me some examples of
19:53some entrepreneurs and companies that
19:54are at this pushing the envelope on
19:56insurance well we got to start with
19:58zenefits I think ok we let's look
19:59our portfolio and nobody can talk about
20:01other companies did one of the
20:02fastest-growing companies in our
20:04portfolio who would have thought it's an
20:06insurance company right and they figured
20:08out a magical way to deliver value to
20:11customers in the form of HR software and
20:13collect insurance money basically as a
20:16way to monetize it and fantastic right
20:19so they figured out the magical business
20:21model and we'd love innovation like that
20:23doesn't have to be a breakthrough
20:24outgrow them all the time it was a
20:26breakthrough business model right you
20:27know we have a couple others in the
20:28portfolio that you might not initially
20:30think of as insurance companies but the
20:32applications have been numerous so so
20:34smart cars doing some really interesting
20:36things around an API to access your car
20:38and the insurance applications there are
20:41really interesting human api which
20:43allows you to have access to all of your
20:45health health data in a more electronic
20:47way has had some really interesting
20:49early insurance opportunities making it
20:52a lot easier if you can under itten for
20:53example or so these are both like in the
20:55in the first example it's how my driving
20:57habits in the second example it's my
20:59health habits exactly and your
21:02healthcare record right and then there's
21:03some really interesting things you know
21:05we've seen in in the u.s. at least I
21:07think clover and Oscar are two really
21:09interesting approaches to try to solve a
21:12healthcare experience that is expensive
21:14and probably suboptimal from a patient
21:17perspective and taking really
21:19interesting approaches I think another
21:20really interesting company that's this
21:23Florida under the radar a little bit as
21:25is Ong on and that's an insurance
21:26company backed by Alibaba and $0.10
21:29that's raised almost a billion dollars
21:31and an eight billion dollar valuation
21:33which would make it one of the most
21:35valuable private companies in the world
21:36really bringing digital insurance to the
21:39Chinese market using messaging apps and
21:42and a lot of social distribution
21:44fascinating and really disruptive
21:46company and they're insuring what I mean
21:48all kinds of things or yes so they
21:50started out insuring the shipping cost
21:52to return a digital good so you get
21:55something ship from $0.10 or or Alibaba
21:57and they didn't sure the cost of you
21:59shipping it back to them and now they're
22:00doing home auto all the different lines
22:04you might want and of that billion
22:05dollars raised as much as that just a
22:07sort of the pool of capital that they
22:08need to to follow through on there and
22:11it's promises I would say in this
22:14particular case there definitely is a
22:15regulatory component it's most of that's
22:18gonna go out and conquer the Chinese
22:20market right I think it's one of the
22:22best examples people are always
22:23wondering one the Amazon or the Google
22:25or the really big tech players and in
22:27the US are gonna make their their real
22:29splash and I think this is a really
22:31intriguing example in China where
22:32towards happening in the insurance game
22:34yeah because you know what you need is
22:36money and then you need to be able to
22:38market to people and then you know you
22:39need to be able to underwrite risk the
22:41right way but there's nothing I guess
22:43proprietary about those things exactly
22:46and so you know what it really sort of
22:48foreshadows for us here in the States is
22:51where do I buy my insurance product
22:53right and I think we're gonna see a lot
22:55of disruption today traditionally I buy
22:57it from a broker who's highly trained
22:59and sent it and in fact makes a ton of
23:01commissions when they sell me insurance
23:02products why aren't I buy more insurance
23:04products at Costco or Amazon or through
23:06facebook Messenger and so like that will
23:09come right do my back buy from Google or
23:11do I sort of rent it every time I take a
23:13ride or have an experience yeah that's
23:16exactly right so maybe we unbundle the
23:18insurance products right they become
23:19more event-driven rather than I cover
23:21your entire life or the other way to do
23:23it is basically the insurance product
23:25flexes as my life changes right which is
23:28the way life insurance is sold today is
23:31basically you figure out like how much
23:32money will you need like if you die your
23:34kids need to go to college and so on and
23:36they're like let's add up all those
23:37numbers and multiply by some safety
23:39factor and you buy that policy but the
23:41reality is your needs change over time
23:43as you make money and save it and your
23:45kids get older and maybe you get your
23:46kids through college and so how about a
23:48life insurance product that
23:49automatically adjusts based on your life
23:51situations because they're already
23:53talking to your bank and they know your
23:54life situation right that would be
23:56okay well this this is fascinating I I
24:00have to say I've never been so
24:02interested in insurance in my life
24:03mission accomplished Frank Mike thank
24:07you guys so much and as this continues
24:09and it will we will be talking about it
24:11more fantastic great thanks so much
