> After paying for broker commissions, fronting costs, reinsurance, customer service, claims processing, there’s often around 50% of the original premium dollar left to pay claims – which is the primary purpose of an insurance company.
What about shareholders? One of the biggest problem I have with insurance companies as for-profit enterprises is the inherent conflict of interest that comes from trying to service claims and customers as best as possible and turning a profit for shareholders.
I've always felt that insurance companies should be run as not-for-profits, or at the very least co-ops..
Don't get me wrong here, still pay the employees and the executives competitively (you want things to run efficiently and by talented teams so you need to attract top talent), but otherwise the whole enterprise should be working hard to make sure every other dollar goes to helping the customers who pay the premiums, and that's it.
Disclosure: I've worked in both P&C and Reinsurance for several years. Also on Wall Street for a couple years. Also in Healthcare for a couple years.
Overall I think this is a nice briefing on the state of the insurance market in the modern economic landscape. It is extensively regulated with rates set and various nuances. All of this, of course, comprises part of a grand "data set" that looks quite appealing to modernization.
Unfortunately, I think there should be a strong expectation that the market (industry) will both be openly hostile to "disruption" oriented attitudes a la Uber, but laugh at any ability to raise capital to compete at any meangingful level.
I applaud your interest in perhaps improving a legally sanctioned form of graft (I prefer Mutual Organizations myself). Conversely, my experience leads me to laugh a little because I've seen the numbers and the complexity behind the scenes. I've got no interest in the industry beyond the paycheck it provided, but it is quite fascinating in numerous respects. Just the naming conventions alone once you get to Bermuda is a trip. Good luck.
I agree with you that there are going to be challenges in building these companies, but not all of existing players aren't as openly hostile as I think you expect. I've spoken with a lot of P&C folks at different companies, and they know things will have to change, they're just trying to figure out how.
Some of the more forward thinking companies are actively investing in new models and companies. I think that's going to accelerate.
Even without that, over time, new models and companies will succeed. Some of these are going to look like stock insurance, some will look like Mutuals and reciprocals, and some will solve insurance like problems in new ways.
We certainly agree on the bulk of your points (see my top-level post elsewhere in this thread). There is opportunity, though in my opinion, a really good way to get there is to work with an incumbent on a "modern" strategy. That's really hand-wavy, I know, but you gotta find the guy who's willing to piss off its brokers and start there.
Since you should know, is TFA's description: "Reinsurer – There are companies that purchase insurance risk from carriers." a conventional way to put this? ISTM a reinsurer is paid to take risk, which is the opposite of how one thinks of purchasing.
The biggest problem I have with insurance is the inherent disproportionate power relationship between the company and their customer.
Essentially, the moment a customer becomes more trouble than they are worth, they are dropped. This is true with other types of industries, but if your health insurance drops you when you get cancer, you can't get more health insurance, and you die.
Same with house insurance, car insurance, ect...
And it causes death or financial disaster all too often.
Some would argue that "this doesn't happen" or "it's illegal".
1) It happens ALL THE TIME.
2) It's illegal, but if you don't have the means or education to fight it. You are pretty much done.
It's the fundamental nature of insurance companies to milk healthy customers while dropping unhealthy customers. It's just too tempting and they are too protected by our legal system for them to not do it.
I know this is pessimistic, but as long as you realize this fundamental imbalance in the relationship with you and your insurance companies, you can mitigate it to a certain extent.
But, really, the only way to completely mitigate it is to be so rich that you don't even need insurance.
Or to be socially-connected to organizations that the health insurance company doesn't want to piss off. My family wasn't rich at all growing up, but we never had a problem with health insurance, because my mother worked for the government. The one organization that insurers will never piss off is the federal or state government, because then they will lose the right to do business in that state.
I do think that this increasing preference for the rule of power over the rule of law is pretty disturbing, and makes a mockery out of the claim that we live in a democracy. Sure we do...if you have leverage over the companies that would violate your rights. But that's the very definition of corruption, when you need to rely on inside information, relationships, or other proprietary tools to get people to do what you want, rather than your rights as a citizen.
I think you're confusing "health insurance" which has a strong social redistribution component, with pure insurance, which is a business transaction to pay off on the chance something happens.
Pure insurance is when a consumer buys a Playstation for $300, and then pays $5 for an extended warranty, or when a company insures their office building against a fire. I don't think your arguments apply to that. In such a case, is there unacceptable social harm if a insurance company determines I'm really bad at taking care of my Playstation, and refuses to insure me? Is a company that badly off if they need to write down the cost of a building if it gets burned down?
Is all insurance a social right? I feel that health insurance has the strongest claim to this, because it's not a pure business transaction and it's about who we're letting die in society.
> the moment a customer becomes more trouble than they are worth, they are dropped.
Insurance is a business. I don't think any business would want customers that are more trouble than they are worth.
> they are too protected by our legal system for them to not do it.
Solving the risk issue with patients who will never be able to afford their healthcare is not an insurance problem. These people are uninsurable. There's a known cost to pharmaceuticals associated with being a hemophiliac. There's a known cost to the outpatient care associated with kidney disease. There's no "risk" associated with those costs.
We need to re-frame the discussion around socialized healthcare, not health insurance. Health insurance, apart from high-deductible, low-cost plans, doesn't exist. The US already has socialized healthcare, the problem is that it is a split private/public model.
Re: health – that's because what you really want is a single-payer health care system funded through taxes, not commercial health insurance, but your country's ideology does not allow for that.
Given that insurers so far have been unsuccessful in legislating requirement for a customer to hand over any DNA test they have performed, it seems quite sensible solution that customer gets dropped, as otherwise they could exercise information asymmetry to their benefit, and great peril of insurance provider.
Of course this is sarcastic, but argument about information asymmetry, created by DNA testing, that supposedly threatens the existence of insurance business is quite real.
> "Essentially, the moment a customer becomes more trouble than they are worth, they are dropped [by the insurance company]."
In healthcare there are many different models, but one popular one is the concept of an ASO. This is the case where a healthcare insurance company (think Anthem) will provide administrative services only (ASO) and not (ultimately) be financially responsible for paying claims. The group responsible, in this ASO relationship, is generally an employer.
This is actually fairly common and in it the insurance company has no incentive to drop a "customer" (member), even if/when they legally could.
There is a lot of "bad" in the health insurance industry, but there is also a lot of "good" - people working from the inside who are trying to make things better for members/individuals. Things like improving quality of care, managing coordinated care models, identifying medical risk (eg. opioid abuse), etc.
I think first people need to better educate themselves on how the industry currently works, where the problems are (there are many), and where the more positive efforts are being made in the industry. Bottom-line - don't write it off, but get better informed and try to contribute.
Before you go thinking about how this applies to healthcare... health insurance in the U.S. is different from other types of insurance and personally I think we'd all be better off it were not even called "insurance."
Health insurance is a mix of pre-paying for predictable and certain expenses with tax-free dollars, a transfer/entitlement system to ensure that more people can afford insurance (by design, your premium does not match your expected risk--either you are pooled with others at your employer, or your exchange account is subject to rating band requirements which means, for example, that in many states old people can only be charged 3X more than young people even though old people are likely to be much more than 3x more expensive to insure), and actual insurance. I'm not sure what percentage of your premium reflects the cost of actually insuring you against uncertain future health events, but it's far from 100%.
This is an interesting article, and some of it applies to healthcare in the U.S., but much of it does not.
"health insurance in the U.S. is different from other types of insurance and personally I think we'd all be better off it were not even called "insurance.""
Agreed. If regular, predictable events are covered, it is not insurance. Regular, predictable events are not insurable. Be cause math.
Wellness checkups and scheduled preventative care and yearly mammograms/prostate are all fantastic things ... but they're not insurable. If someone is selling "insurance" for those things, you can be certain that you're paying 100% of the cost somewhere.
I've always wondered about US health insurance -- why can't the physician give me quotes about my personal obligation for various treatment options? It's frustrating that as soon as it's time to come up with a bill, poof there it is but prior to the bill being generated all I get is shrugs?
Is it because the insurance coverage algorithms are too complicated? Because the different entities involved in a single treatment plan is too complicated to navigate? Because physicians feel that cost is orthogonal to medicine and they prefer not to be involved/prefer to recommend the ideal treatment based on a predicted outcome? All of the above?
It feels like if there were a particular hospital group / physician group that had this feature, they would attract a lot of attention. Just imagine, "Your initial differential diagnosis will not exceed $150 and we'll discuss treatment options or more conclusive diagnostic tests afterwards."
All I've heard so far are physicians who don't accept insurance but instead have a straightforward "menu" for common items, which is interesting but not what I think most people want.
Nobody can tell you because nobody knows. And nobody can know.
Here is a simple example that my sister (a nurse) gave me yesterday. Suppose that you go in for an operation at the hospital, spend a week recovering, and develop diarrhea on day 2 while you are there. That diarrhea is a "hospital acquired infection" and insurance won't pay a dime for your operation. Therefore until you've been through the hospital, nobody knows whether you'll get paid.
Oh right, and the possibility of this happening is a reason for the hospital to kick you out of the hospital as quickly as possible. Average patient outcomes may be better if you stay a week, but their odds of getting paid are better if you're kicked out within 48 hours.
This is just the tip of the iceberg. She went on about how broken health care is for an hour...
The problem is there is a bunch of hidden complexity in medical coding and billing. There's 5 Evaluation & Maintenance (E&M) codes that may apply to an office visit, another 5 for ER visits, 4 for tele-medicine consults - which one is used depends on multiple factors that you can only know after the fact, and which one is applicable to the visit determines the expense. This doesn't factor in other procedures such as labs, and whether such charges are required to be bundled into the E&M fee by a specific payer or not - plus differing allowables from each payer. Add in the fact that nearly nobody bothers doing pre-authorizations except surgery centers and other outpatient services that payers require prior authorization for anyway and the whole situation turns into a giant clusterfuck.
Is it POSSIBLE to do all of this and give you a proper quote before a visit? Sure, but it requires some fairly complex software to do so and manual input of tons of different data specific to your insurance contract that almost nobody wants to do it (ironically, the billing company I work for DOES this - but since we aren't involved in patient care it's only utilized to ensure we get paid properly by insurance companies).
I think what's happening is that your doctor will attempt to get as much money from the insurance company as he/she can.
My wife has experience of the doctor's office asking insurance company for $X, and the insurance company comes back and say "no, max $Y". So then the final "cost" all of a sudden becomes $Y. Pay attention to the claims that your doctors send to insurance company and you might be able to see that.
So instead of telling you, yes, whatever procedure is definitely just going to cost $Y. They can't tell you how much things are. It depends on maneuvers with other players in the industry.
This is very different than say in Canada. If I want to get a teeth filling in Canada, my dentist straight up tells me how much before the procedure. If I want to price shop that, I can. In the US, nobody is willing to say how much, because "it depends".
It's not because intrinsically there can't be price transparency. It's because of all the messed up incentives that the industry has that causes US health care to be as such.
I agree, to me the second biggest industry acceptance that is counter intuitive is that insurers need to make money off of the float. Why can't an insurer come along who charges a fee for the service of ACTUALLY BEING ACCESSIBLE TO THE CUSTOMER. Intead of fax me this paper and wait two months. I am paying the insurer on the basis that they want to draw out any claim I have.
"All I've heard so far are physicians who don't accept insurance but instead have a straightforward "menu" for common items, which is interesting but not what I think most people want." - I think people want this but they are scared of going off of insurance in the event they need to see someone who doesn't offer this (chance occurance, expensive disease).
For routine things, your responsibility is a flat amount, usually printed on your insurance card. So you know, for example, if you go to your general physician, you pay this much. Go to a specialist, pay this much. Go to emergency room, pay this much.
For more complex things, it's usually:
1. Not possible to know in advance everything that will need to be done. Many medical procedures are not things that just go identically every single time, and complications can occur during the procedure. Having to call it off, re-quote, re-schedule, etc. is not optimal.
2. The doctor likely doesn't actually know how the procedure will be billed. Medical billing is done using standardized codes to describe procedures, and the doctor will have someone who knows how to do that, but that person may not even work in the same building as the doctor. And the sets of allowed codes and how to use them can change quarterly, and that's without getting into the arms race of doctors trying to "up-code" (rather than the most obvious code for a procedure, find a way to bill it as multiple procedures or as a plausible but higher-paying code, since doctors and insurance companies are locked in an eternal battle of doctors trying to make as much money as they'd like and insurance companies trying to pay as little money as they'd like).
A lot of good answers here. A few straightforward ones:
- Because healthcare providers negotiate different rates with providers, so the "list price" differs by your provider and plan.
- Because your personal cost is unknown to the doctor, as it would depend on factors between you and your insurance company (like deductible met), coverage types, etc.
Now, these are both solvable problems. And I agree with the sentiment of other posters here that it's predatory that medicine is one of the few fields where you simply don't know how much something will cost until you get the invoice.
Because it's sort of like hotels. There's a "rack rate" which is a high price that nobody pays, and a price floor is whatever Medicare pays. (With Hotels, GSA is the "normal" price floor, and cheaper rates are usually wholesale) If you price cheaper than Medicare, you get sued for fraud unless you charge that low cost to Medicare.
Everyone else has a bewildering discount scheme. The doctor literally has no idea what you pay.
In other cases you have HMOs, where primary care doctors get a monthly nut to take care of you and don't get a fee for service in most cases.
Health insurance is a very different business than P&C and one that I've spent a lot less time researching. Each customer interaction at least 3 participating actors (patient, provider, payer), each of which have different incentives, rules, and understanding of those rules.
I think there are great companies being built in the space (take a look at what Clover Health is doing https://www.cloverhealth.com/en/), but it's not an area I'm focusing on.
It's funny that I can take my car to a mechanic and get a free and pretty accurate diagnosis. But you go to a hospital, and you need pay for the diagnosis and it may not be accurate.
Even for a checkup, after the procedures are done, the office can't tell me the bill.
...yet we we call mechanics "wrench monkey" in a demeaning manor.
BTW, this is probably why services like Minute Clinic are getting popular...go in, get something done, pay a flat/low fee.
Mostly due to all of the variables that go into pricing a claim. And that logic on lives in the insurer's claim processing system.
Pieces that can impact the price. Your insurer and what product you have. These will affect who is considered in-network and the fee schedule to use. Different insures will have different arrangements. Depending on the product if you have a narrow network product they may or may not be in-network. It could also depend on the location. A provider can be in-network in one location but not in another.
Also the procedure that is actually performed may be slightly different from what was planned due to unforeseen circumstances.
This is assuming the provider is aware of what the actual costs are. In many cases they don't even know the ballpark price since that is not the portion that they deal with.
You might be interested in something like the Surgery Center of Oklahoma which has an upfront pricing page for their surgery procedures: https://surgerycenterok.com/pricing/
Even though the ycombinator blog post is discussing innovation etc with regard to insurance, I like the idea of innovation on the side of service providers. And it is somewhat sad that a list of prices is innovative.
I have Kaiser insurance which is an HMO with straightforward pricing for most things. One fixed copay for doctor visit, one for specialist visit, one for outpatient care, etc. The problem is you are stuck using only Kaiser facilities and doctors.
Even ignoring insurance, it can be impossible to get an accurate cash price depending upon the physician. And if they can't figure out the cash price, there is no way in hell they will figure out the insured price.
We recently went to a physician who works in an area that regularly isn't covered by insurance. For all procedures, the cash pricing was upfront and understandable.
In comparison, we tried to deal with another physician for a different procedure we knew our insurance didn't cover. Literally days worth of time was spent on the phone to try to figure it out and the day of the procedure we were told the prices were wrong and didn't account for some stuff.
Insurance companies have been into information processing in a big way since paper and pencil days. The first company to buy a commercial computer was Metropolitan Life.
In commercial insurance, there's a question of how intrusive the insurance company should be. My favorite insurance company, The Hartford Steam Boiler Insurance Company, established in 1866, was finally bought out by Munich Re a few years ago. Hartford Steam Boiler insures boilers and equipment in industrial plants. Most of their employees are inspectors. If you want to buy a policy from them, they come out and inspect the equipment. They give you a list of what you have to fix. Then they come back to inspect after everything is fixed. Then they sell you a policy. They also inspect again, randomly and unannounced. Cut corners on maintenance, and HSB cancels your policy. The premiums are low, because boilers inspected by Hartford Steam Boiler don't blow up.
Most companies hate that, even though the premiums are lower.
We're working on the modern commercial insurance brokerage at Abe (https://www.hiabe.com/).
Aaron's right that too much of the industry runs on pen and paper. It's confusing for the buyer and a massive headache for brokers.
Most of what we're building is behind the scenes to make the brokerage way more efficient. If you're an insurance expert with ideas to leverage tech or engineer interested in man+machine symbiosis, I'd love to chat (gordon [at] hiabe.com)!
After having spent the last 3 years in insurance with Allay trying to take on employer health insurance costs by making it easier for smaller groups to become self insured, I have found that the regulations are cumbersome but not huge blockers. The regulations are there to protect people and for the most part do that job correctly.
The hard part about this industry is that there is no single incumbent to disrupt, but thousand of very small businesses who have personal relationships with their clients. Also whereever you jump into the process, you have to deal with companies who do not value technology as much as the HN crowd would. These companies still print out PDFs and have automated very little of their business. No matter how fast you make your software, you are the behest of the companies below and above you in the chain.
If anybody wants to nerd out on the insurance industry, my contact is in my profile.
Healthcare needs to be funded through a non-profit community trust whose first priority is to secure the highest overall health for the population. Insurance is motivated by profit which will always seek to squeeze more money from the system as its first priority.
Americans pay 2.5x more than the average first-world single-payer system.
2.5x...
This funds demonstrably and starkly worse outcomes on the metrics we should collectively care about: life expectancy. Incidence of chronic disease. Infant morality.
The reason we pay much more for much less is simple: for-profit health care optimizes for profits, not outcomes. That is exactly what it has done; exactly what it will do.
The public sector already does single-payer in this country with an order of magnitude less overhead cost than the private sector.
When I discuss these facts with conservatives and libertarians, I usually discover that most of this is decreed "fine," because we disagree about whether or not health care is a right.
For that reason, I've stopped suggesting that it is, and focused on another stark fact:
Failure to provide basic health care to all, through straightforward means, merely means that it is provided through partially hidden means (like ER rooms) at vastly greater cost, not least because emergency care cannot perform preventive and chronic care, which in many cases would provide better outcomes for two orders of magnitude less cost.
Morality is a matter of instinct and choice. The costs of the existing broken system are however obvious, and render any defense IMO irrational.
It's an excellent article and spot on. Insurance has remained fundamentally unchanged for centuries. In order to really change it, a complete re-architecture is needed. Insurance should be fully digital, hassle free, transparent and fair.
It's hard to really change it when you're not a fully licensed and regulated carrier. An MGA can create a beautiful UI on top of an old insurance product, but eventually, consumers will meet face to face with this old insurance product (for example in claims), and it's going to be the same old experience again.
As many of you stated here, there's an inherent conflict between the insurance company and the insured. Until that changes, the experience will stay the same.
Surprised not to see mention of the fundamental modern tech problem with regard to insurance.
As soon as you have big enough data, and artificially intelligent enough algorithms, the insurance becomes too predictive.
The whole point of insurance is to pay people when rare, bad events happen to them. If an insurer can predict well enough who will be the victims, it can refuse them insurance, and hence remove the entire purpose and benefit of the insurance.
This is the flip side to moral hazard. What does it even mean to offer commercial insurance, if it is only offered to the people who least need it?
This is least bad, for example, with a car predicting you're a bad driver as you can improve behaviour. It is really problematic with data such as gene sequencing, which you can't do anything about.
Only way out I can see is compulsory insurance, levied as a tax. Or maybe non-profit or Government AIs, trained to find a sweet spot between moral hazard and its opposite?
I've done a lot of systems design and development in the insurance industry, and it's a vast space with opportunities not just for insurers.
The most recent thing I worked on was a pricing and activation engine that sat behind a web site that acted as a broker for a number of insurers. That isn't new, but it was new for this market - life insurance. As such my employer was a single provider on a panel of providers.
The web site that provided the panel brought a number of innovations - one of them being an underwriting SaaS. Panel participants are able to enter their underwriting crown jewels into a 3rd party web site, secure in the knowledge that their IP wasn't going to be leaked or shared with others on the panel.
There were many more efficiencies that the model enabled, which were never realised (well, not by the time I left) because the SaaS provider couldn't reach financial agreements with some of the providers.
Greed was (is?) something that risked torpedoing the most innovative thing I've seen in life insurance, ever.
All that to say I agree that this market is brimming with opportunity. The market is so incredibly broad, and deep, and so complex... Regulation is definitely a thing, but hardly an impediment. I have, for example, spent many years in this industry, but never worked for (or with) reinsurers. I have a long-standing suspicion though, that that market is so convoluted that the front-line insurer can conceivable be it's own reinsurer, after having passed through like, 15 other reinsurers...
While much of what Aaron wrote is spot on, one thing missing is the importance of distribution. This a challenge for any startup, but is more pronounced in a regulated industry like insurance with largely homogenized products and with serious restrictions on how you can legally distribute your product (see Zenefits).
If you consider how most people and small businesses buy insurance, they typically make purchasing decisions one a year at most. As such, you need to get in front of them at the exact moment they want to purchase. GEICO and Progressive have done this really well, but have effectively bid up the cost of online advertising to make it prohibitively expensive. This is also why agents are such a powerful force in the industry (and because they effectively provide carriers with an initial underwriting screen which they don't need to file publicly).
It's important to get the product right, and there are many flaws with most P&C insurance today (chief among them that the forms haven't really changed in the past few decades), but I'd encourage any entrepreneurs to make sure they have an answer on distribution before spending time on product.
Disclosure: I've spent a lot of time looking at this as founder of a P&C insurance startup a few years ago.
Nice writeup. Certainly scratches the surface a bit (and I think that was the intent). I've spent the bulk of my career in the industry, both in P&C and A&H, and there is a ton of opportunity. You correctly conclude that it's not because we're idiots, even though some days I'm kind of an idiot.
Again, tons of opportunity, with some barriers that aren't unique to the industry but maybe unique to what the tech community might encounter (and this is by no means an exhaustive list) -
1. Capital requirements: You need a dumptruck full of money to do much in this industry, at least if your intent is to write or reinsure business, but even to a lesser degree if you're working in ancillary services.
2. Regulatory environment: Assuming US operations, 50 states with 50 different sets of regs need to be okay with what you're doing. In addition, federal law comes into play in certain spaces (GLB, HIPAA, etc).
3. Distribution: The current model compensates every layer of the distribution model very well. It's not as easy as you might imagine to disrupt when entrenched interests are all making a ton of money AND are interdependent upon their neighbors in the value chain to continue to do that. You can't just hack a piece off because that bothers their neighbor, who in other circumstances might otherwise be a competitor, but has a shared interest. The relationships get complex.
None of this is fatal, but you must navigate it and play by many of the rules, particularly with #2. As Zenefits learned very publicly, the insurance industry and its regulators weren't going to let someone do what Uber did to the taxi industry - which was essentially to operate in the grey/black and just ignore the calls to stop. Insurance is a subset of financial services, and financial services is a powerful industry. For any startup, my advice is to build compliance and regulatory relations in from day 1. It's the least exciting thing you'll ever do, but is extremely important. Anyhow, I'm rambling. By no means do I want to discourage anyone from trying, I'm just trying to highlight some of the big things that are "different" in our space.
I'm particularly interested in the data piece because that's what I've done and built a career on in this industry, but it never struck me as sexy enough for YC. Appreciate the article.
> For most of the insurance world, the hardest and most important thing to find is effective distribution and customer acquisition.
This is absolutely true. If you look at how regional, family-owned P&C carriers got their start, you'll find they started as brokers. These brokers found a profitable niche (call it motorcycles in California, or non-standard auto in Texas) that they wanted to own, moved on to become MGAs, and then admitted carriers offering multiple lines of business.
If you can find a way to own the customer as a distributor, you own the life-blood of the entire business downstream. As a result, P&C insurance companies are huge spenders on marketing (GEICO, spends $1.7B/year). It would be game-changing if this cash was used to provide utility to their customers above and beyond the insurance transaction.
I'm of the opinion that insurance premiums could be the new ad dollars - used to create products that promote lock-in, and much more consumer-centric insurance companies.
If you're going to innovate on the software side in P&C you would need to figure out what they're not doing or could not do that would allow you to get ahead of them.
My issue with insurance companies is that they're conflicted by design.
As typically publicly traded companies, they have a fiduciary responsibility to their shareholders to maximize profits. And the only way to maximize profits is to deny coverage.
The purpose and value of insurance is amortizing risk. There are very few things that I think should be controlled by the federal government, but a single payer health care system is one of the few that actually makes sense from an incentives perspective.
In the words of Charlie Munger who's spent more time thinking about insurance than almost anyone: "Show me the incentive and I will show you the outcome."
Why does the US pay the highest prices for mediocre health care? Perverse incentives.
Unless you start with covering something completely new, I think it makes sense to start as analytics provider for insurance. Becoming a carrier is expensive, and MGA can only sell existing policies already created by a carrier.
If you sell existing policy you are at severe data disadvantage comparing to existing insurance companies, even with superior tech.
Shameless plug: At https://tensorflight.com we are working on P&C insurance and we focus on analytics for commercial properties mentioned in the article. Please get in touch at ( kozikow [at] tensorflight.com ) if you are interested in the subject.
The article spends many paragraphs on the complexities of the B2B and B2C players in insurance, their distribution channels, and how they get paid. It spends one sentence on noting how the pricing of risk is based on actuarial models, and then moves on. The pricing of risk is absolutely key to the entire industry, and it is stone age! It's all done in spreadsheets.
> The article spends many paragraphs on the complexities of the B2B and B2C players in insurance, their distribution channels, and how they get paid.
The article is primarily focused on commercial lines - primarily P&C, but there are similarities in A&H. The distribution stack is complex and is a natural starting point to start looking at opportunities. The general tech community probably isn't going to look at risk pricing and the ancillary services associated with that - there's a whole discipline built around it and I don't know how many in YC's catchment would be FSA/ASAs or excited by that particular aspect. Plus, for traditional products in P&C and the Disability side of A&H, the industry is pretty good at pricing the risk. Nobody's squeezing growth from their underwriting ratio, they're doing it through products and distribution.
Anyhow, not trying to give you a hard time, just another angle.
[+] [-] sbarre|8 years ago|reply
What about shareholders? One of the biggest problem I have with insurance companies as for-profit enterprises is the inherent conflict of interest that comes from trying to service claims and customers as best as possible and turning a profit for shareholders.
I've always felt that insurance companies should be run as not-for-profits, or at the very least co-ops..
Don't get me wrong here, still pay the employees and the executives competitively (you want things to run efficiently and by talented teams so you need to attract top talent), but otherwise the whole enterprise should be working hard to make sure every other dollar goes to helping the customers who pay the premiums, and that's it.
[+] [-] 6stringmerc|8 years ago|reply
Overall I think this is a nice briefing on the state of the insurance market in the modern economic landscape. It is extensively regulated with rates set and various nuances. All of this, of course, comprises part of a grand "data set" that looks quite appealing to modernization.
Unfortunately, I think there should be a strong expectation that the market (industry) will both be openly hostile to "disruption" oriented attitudes a la Uber, but laugh at any ability to raise capital to compete at any meangingful level.
I applaud your interest in perhaps improving a legally sanctioned form of graft (I prefer Mutual Organizations myself). Conversely, my experience leads me to laugh a little because I've seen the numbers and the complexity behind the scenes. I've got no interest in the industry beyond the paycheck it provided, but it is quite fascinating in numerous respects. Just the naming conventions alone once you get to Bermuda is a trip. Good luck.
[+] [-] akharris|8 years ago|reply
Some of the more forward thinking companies are actively investing in new models and companies. I think that's going to accelerate.
Even without that, over time, new models and companies will succeed. Some of these are going to look like stock insurance, some will look like Mutuals and reciprocals, and some will solve insurance like problems in new ways.
[+] [-] SmellTheGlove|8 years ago|reply
[+] [-] jessaustin|8 years ago|reply
[+] [-] socrates1998|8 years ago|reply
Essentially, the moment a customer becomes more trouble than they are worth, they are dropped. This is true with other types of industries, but if your health insurance drops you when you get cancer, you can't get more health insurance, and you die.
Same with house insurance, car insurance, ect...
And it causes death or financial disaster all too often.
Some would argue that "this doesn't happen" or "it's illegal".
1) It happens ALL THE TIME.
2) It's illegal, but if you don't have the means or education to fight it. You are pretty much done.
It's the fundamental nature of insurance companies to milk healthy customers while dropping unhealthy customers. It's just too tempting and they are too protected by our legal system for them to not do it.
I know this is pessimistic, but as long as you realize this fundamental imbalance in the relationship with you and your insurance companies, you can mitigate it to a certain extent.
But, really, the only way to completely mitigate it is to be so rich that you don't even need insurance.
[+] [-] nostrademons|8 years ago|reply
I do think that this increasing preference for the rule of power over the rule of law is pretty disturbing, and makes a mockery out of the claim that we live in a democracy. Sure we do...if you have leverage over the companies that would violate your rights. But that's the very definition of corruption, when you need to rely on inside information, relationships, or other proprietary tools to get people to do what you want, rather than your rights as a citizen.
[+] [-] fanzhang|8 years ago|reply
Pure insurance is when a consumer buys a Playstation for $300, and then pays $5 for an extended warranty, or when a company insures their office building against a fire. I don't think your arguments apply to that. In such a case, is there unacceptable social harm if a insurance company determines I'm really bad at taking care of my Playstation, and refuses to insure me? Is a company that badly off if they need to write down the cost of a building if it gets burned down?
Is all insurance a social right? I feel that health insurance has the strongest claim to this, because it's not a pure business transaction and it's about who we're letting die in society.
[+] [-] debacle|8 years ago|reply
Insurance is a business. I don't think any business would want customers that are more trouble than they are worth.
> they are too protected by our legal system for them to not do it.
Solving the risk issue with patients who will never be able to afford their healthcare is not an insurance problem. These people are uninsurable. There's a known cost to pharmaceuticals associated with being a hemophiliac. There's a known cost to the outpatient care associated with kidney disease. There's no "risk" associated with those costs.
We need to re-frame the discussion around socialized healthcare, not health insurance. Health insurance, apart from high-deductible, low-cost plans, doesn't exist. The US already has socialized healthcare, the problem is that it is a split private/public model.
[+] [-] raspie|8 years ago|reply
[+] [-] DanBC|8 years ago|reply
And the principle of "utmost good faith", which insurance companies abuse to remove cover after a claim, unfairly disadvantages many people.
https://en.wikipedia.org/wiki/Insurance_law#Utmost_good_fait...
https://en.wikipedia.org/wiki/Uberrima_fides
[+] [-] wuch|8 years ago|reply
Of course this is sarcastic, but argument about information asymmetry, created by DNA testing, that supposedly threatens the existence of insurance business is quite real.
[+] [-] forgotmypwd|8 years ago|reply
In healthcare there are many different models, but one popular one is the concept of an ASO. This is the case where a healthcare insurance company (think Anthem) will provide administrative services only (ASO) and not (ultimately) be financially responsible for paying claims. The group responsible, in this ASO relationship, is generally an employer.
This is actually fairly common and in it the insurance company has no incentive to drop a "customer" (member), even if/when they legally could.
There is a lot of "bad" in the health insurance industry, but there is also a lot of "good" - people working from the inside who are trying to make things better for members/individuals. Things like improving quality of care, managing coordinated care models, identifying medical risk (eg. opioid abuse), etc.
I think first people need to better educate themselves on how the industry currently works, where the problems are (there are many), and where the more positive efforts are being made in the industry. Bottom-line - don't write it off, but get better informed and try to contribute.
[+] [-] asr|8 years ago|reply
Health insurance is a mix of pre-paying for predictable and certain expenses with tax-free dollars, a transfer/entitlement system to ensure that more people can afford insurance (by design, your premium does not match your expected risk--either you are pooled with others at your employer, or your exchange account is subject to rating band requirements which means, for example, that in many states old people can only be charged 3X more than young people even though old people are likely to be much more than 3x more expensive to insure), and actual insurance. I'm not sure what percentage of your premium reflects the cost of actually insuring you against uncertain future health events, but it's far from 100%.
This is an interesting article, and some of it applies to healthcare in the U.S., but much of it does not.
[+] [-] rsync|8 years ago|reply
Agreed. If regular, predictable events are covered, it is not insurance. Regular, predictable events are not insurable. Be cause math.
Wellness checkups and scheduled preventative care and yearly mammograms/prostate are all fantastic things ... but they're not insurable. If someone is selling "insurance" for those things, you can be certain that you're paying 100% of the cost somewhere.
[+] [-] joshfraser|8 years ago|reply
[+] [-] colanderman|8 years ago|reply
[1] https://en.wikipedia.org/wiki/Chargemaster
[+] [-] wyldfire|8 years ago|reply
Is it because the insurance coverage algorithms are too complicated? Because the different entities involved in a single treatment plan is too complicated to navigate? Because physicians feel that cost is orthogonal to medicine and they prefer not to be involved/prefer to recommend the ideal treatment based on a predicted outcome? All of the above?
It feels like if there were a particular hospital group / physician group that had this feature, they would attract a lot of attention. Just imagine, "Your initial differential diagnosis will not exceed $150 and we'll discuss treatment options or more conclusive diagnostic tests afterwards."
All I've heard so far are physicians who don't accept insurance but instead have a straightforward "menu" for common items, which is interesting but not what I think most people want.
[+] [-] btilly|8 years ago|reply
Here is a simple example that my sister (a nurse) gave me yesterday. Suppose that you go in for an operation at the hospital, spend a week recovering, and develop diarrhea on day 2 while you are there. That diarrhea is a "hospital acquired infection" and insurance won't pay a dime for your operation. Therefore until you've been through the hospital, nobody knows whether you'll get paid.
Oh right, and the possibility of this happening is a reason for the hospital to kick you out of the hospital as quickly as possible. Average patient outcomes may be better if you stay a week, but their odds of getting paid are better if you're kicked out within 48 hours.
This is just the tip of the iceberg. She went on about how broken health care is for an hour...
[+] [-] snuxoll|8 years ago|reply
Is it POSSIBLE to do all of this and give you a proper quote before a visit? Sure, but it requires some fairly complex software to do so and manual input of tons of different data specific to your insurance contract that almost nobody wants to do it (ironically, the billing company I work for DOES this - but since we aren't involved in patient care it's only utilized to ensure we get paid properly by insurance companies).
[+] [-] dannysu|8 years ago|reply
My wife has experience of the doctor's office asking insurance company for $X, and the insurance company comes back and say "no, max $Y". So then the final "cost" all of a sudden becomes $Y. Pay attention to the claims that your doctors send to insurance company and you might be able to see that.
So instead of telling you, yes, whatever procedure is definitely just going to cost $Y. They can't tell you how much things are. It depends on maneuvers with other players in the industry.
This is very different than say in Canada. If I want to get a teeth filling in Canada, my dentist straight up tells me how much before the procedure. If I want to price shop that, I can. In the US, nobody is willing to say how much, because "it depends".
It's not because intrinsically there can't be price transparency. It's because of all the messed up incentives that the industry has that causes US health care to be as such.
[+] [-] danvoell|8 years ago|reply
"All I've heard so far are physicians who don't accept insurance but instead have a straightforward "menu" for common items, which is interesting but not what I think most people want." - I think people want this but they are scared of going off of insurance in the event they need to see someone who doesn't offer this (chance occurance, expensive disease).
[+] [-] ubernostrum|8 years ago|reply
For more complex things, it's usually:
1. Not possible to know in advance everything that will need to be done. Many medical procedures are not things that just go identically every single time, and complications can occur during the procedure. Having to call it off, re-quote, re-schedule, etc. is not optimal.
2. The doctor likely doesn't actually know how the procedure will be billed. Medical billing is done using standardized codes to describe procedures, and the doctor will have someone who knows how to do that, but that person may not even work in the same building as the doctor. And the sets of allowed codes and how to use them can change quarterly, and that's without getting into the arms race of doctors trying to "up-code" (rather than the most obvious code for a procedure, find a way to bill it as multiple procedures or as a plausible but higher-paying code, since doctors and insurance companies are locked in an eternal battle of doctors trying to make as much money as they'd like and insurance companies trying to pay as little money as they'd like).
[+] [-] basseq|8 years ago|reply
- Because healthcare providers negotiate different rates with providers, so the "list price" differs by your provider and plan.
- Because your personal cost is unknown to the doctor, as it would depend on factors between you and your insurance company (like deductible met), coverage types, etc.
Now, these are both solvable problems. And I agree with the sentiment of other posters here that it's predatory that medicine is one of the few fields where you simply don't know how much something will cost until you get the invoice.
[+] [-] Spooky23|8 years ago|reply
Everyone else has a bewildering discount scheme. The doctor literally has no idea what you pay.
In other cases you have HMOs, where primary care doctors get a monthly nut to take care of you and don't get a fee for service in most cases.
[+] [-] akharris|8 years ago|reply
I think there are great companies being built in the space (take a look at what Clover Health is doing https://www.cloverhealth.com/en/), but it's not an area I'm focusing on.
[+] [-] fma|8 years ago|reply
Even for a checkup, after the procedures are done, the office can't tell me the bill.
...yet we we call mechanics "wrench monkey" in a demeaning manor.
BTW, this is probably why services like Minute Clinic are getting popular...go in, get something done, pay a flat/low fee.
[+] [-] Mz|8 years ago|reply
http://micheleincalifornia.blogspot.com/2017/01/direct-prima...
[+] [-] rgoddard|8 years ago|reply
Pieces that can impact the price. Your insurer and what product you have. These will affect who is considered in-network and the fee schedule to use. Different insures will have different arrangements. Depending on the product if you have a narrow network product they may or may not be in-network. It could also depend on the location. A provider can be in-network in one location but not in another.
Also the procedure that is actually performed may be slightly different from what was planned due to unforeseen circumstances.
This is assuming the provider is aware of what the actual costs are. In many cases they don't even know the ballpark price since that is not the portion that they deal with.
[+] [-] TommyBombadil|8 years ago|reply
Even though the ycombinator blog post is discussing innovation etc with regard to insurance, I like the idea of innovation on the side of service providers. And it is somewhat sad that a list of prices is innovative.
[+] [-] pkaye|8 years ago|reply
[+] [-] sbov|8 years ago|reply
We recently went to a physician who works in an area that regularly isn't covered by insurance. For all procedures, the cash pricing was upfront and understandable.
In comparison, we tried to deal with another physician for a different procedure we knew our insurance didn't cover. Literally days worth of time was spent on the phone to try to figure it out and the day of the procedure we were told the prices were wrong and didn't account for some stuff.
[+] [-] Animats|8 years ago|reply
In commercial insurance, there's a question of how intrusive the insurance company should be. My favorite insurance company, The Hartford Steam Boiler Insurance Company, established in 1866, was finally bought out by Munich Re a few years ago. Hartford Steam Boiler insures boilers and equipment in industrial plants. Most of their employees are inspectors. If you want to buy a policy from them, they come out and inspect the equipment. They give you a list of what you have to fix. Then they come back to inspect after everything is fixed. Then they sell you a policy. They also inspect again, randomly and unannounced. Cut corners on maintenance, and HSB cancels your policy. The premiums are low, because boilers inspected by Hartford Steam Boiler don't blow up.
Most companies hate that, even though the premiums are lower.
[+] [-] gwintrob|8 years ago|reply
Aaron's right that too much of the industry runs on pen and paper. It's confusing for the buyer and a massive headache for brokers.
Most of what we're building is behind the scenes to make the brokerage way more efficient. If you're an insurance expert with ideas to leverage tech or engineer interested in man+machine symbiosis, I'd love to chat (gordon [at] hiabe.com)!
[+] [-] togasystems|8 years ago|reply
The hard part about this industry is that there is no single incumbent to disrupt, but thousand of very small businesses who have personal relationships with their clients. Also whereever you jump into the process, you have to deal with companies who do not value technology as much as the HN crowd would. These companies still print out PDFs and have automated very little of their business. No matter how fast you make your software, you are the behest of the companies below and above you in the chain.
If anybody wants to nerd out on the insurance industry, my contact is in my profile.
[+] [-] toddwprice|8 years ago|reply
[+] [-] aaroninsf|8 years ago|reply
Americans pay 2.5x more than the average first-world single-payer system.
2.5x...
This funds demonstrably and starkly worse outcomes on the metrics we should collectively care about: life expectancy. Incidence of chronic disease. Infant morality.
The reason we pay much more for much less is simple: for-profit health care optimizes for profits, not outcomes. That is exactly what it has done; exactly what it will do.
The public sector already does single-payer in this country with an order of magnitude less overhead cost than the private sector.
When I discuss these facts with conservatives and libertarians, I usually discover that most of this is decreed "fine," because we disagree about whether or not health care is a right.
For that reason, I've stopped suggesting that it is, and focused on another stark fact:
Failure to provide basic health care to all, through straightforward means, merely means that it is provided through partially hidden means (like ER rooms) at vastly greater cost, not least because emergency care cannot perform preventive and chronic care, which in many cases would provide better outcomes for two orders of magnitude less cost.
Morality is a matter of instinct and choice. The costs of the existing broken system are however obvious, and render any defense IMO irrational.
[+] [-] refurb|8 years ago|reply
Over 50% of healthcare in the US is paid for by the gov't and those costs are still higher than other systems that are single payer and non-profit.
Doesn't quite align with your theory though.
[+] [-] gilsadis|8 years ago|reply
It's hard to really change it when you're not a fully licensed and regulated carrier. An MGA can create a beautiful UI on top of an old insurance product, but eventually, consumers will meet face to face with this old insurance product (for example in claims), and it's going to be the same old experience again.
As many of you stated here, there's an inherent conflict between the insurance company and the insured. Until that changes, the experience will stay the same.
In Lemonade, we are changing that. Will love to get your feedback. Here's how it works - https://www.youtube.com/watch?v=6U08uhV8c6Y&t=9s
Disclaimer: I'm head of product at Lemonade (lemonade.com)
[+] [-] frabcus|8 years ago|reply
As soon as you have big enough data, and artificially intelligent enough algorithms, the insurance becomes too predictive.
The whole point of insurance is to pay people when rare, bad events happen to them. If an insurer can predict well enough who will be the victims, it can refuse them insurance, and hence remove the entire purpose and benefit of the insurance.
This is the flip side to moral hazard. What does it even mean to offer commercial insurance, if it is only offered to the people who least need it?
This is least bad, for example, with a car predicting you're a bad driver as you can improve behaviour. It is really problematic with data such as gene sequencing, which you can't do anything about.
Only way out I can see is compulsory insurance, levied as a tax. Or maybe non-profit or Government AIs, trained to find a sweet spot between moral hazard and its opposite?
[+] [-] Spearchucker|8 years ago|reply
The most recent thing I worked on was a pricing and activation engine that sat behind a web site that acted as a broker for a number of insurers. That isn't new, but it was new for this market - life insurance. As such my employer was a single provider on a panel of providers.
The web site that provided the panel brought a number of innovations - one of them being an underwriting SaaS. Panel participants are able to enter their underwriting crown jewels into a 3rd party web site, secure in the knowledge that their IP wasn't going to be leaked or shared with others on the panel.
There were many more efficiencies that the model enabled, which were never realised (well, not by the time I left) because the SaaS provider couldn't reach financial agreements with some of the providers.
Greed was (is?) something that risked torpedoing the most innovative thing I've seen in life insurance, ever.
All that to say I agree that this market is brimming with opportunity. The market is so incredibly broad, and deep, and so complex... Regulation is definitely a thing, but hardly an impediment. I have, for example, spent many years in this industry, but never worked for (or with) reinsurers. I have a long-standing suspicion though, that that market is so convoluted that the front-line insurer can conceivable be it's own reinsurer, after having passed through like, 15 other reinsurers...
[+] [-] zstiefler|8 years ago|reply
If you consider how most people and small businesses buy insurance, they typically make purchasing decisions one a year at most. As such, you need to get in front of them at the exact moment they want to purchase. GEICO and Progressive have done this really well, but have effectively bid up the cost of online advertising to make it prohibitively expensive. This is also why agents are such a powerful force in the industry (and because they effectively provide carriers with an initial underwriting screen which they don't need to file publicly).
It's important to get the product right, and there are many flaws with most P&C insurance today (chief among them that the forms haven't really changed in the past few decades), but I'd encourage any entrepreneurs to make sure they have an answer on distribution before spending time on product.
Disclosure: I've spent a lot of time looking at this as founder of a P&C insurance startup a few years ago.
[+] [-] SmellTheGlove|8 years ago|reply
Again, tons of opportunity, with some barriers that aren't unique to the industry but maybe unique to what the tech community might encounter (and this is by no means an exhaustive list) -
1. Capital requirements: You need a dumptruck full of money to do much in this industry, at least if your intent is to write or reinsure business, but even to a lesser degree if you're working in ancillary services.
2. Regulatory environment: Assuming US operations, 50 states with 50 different sets of regs need to be okay with what you're doing. In addition, federal law comes into play in certain spaces (GLB, HIPAA, etc).
3. Distribution: The current model compensates every layer of the distribution model very well. It's not as easy as you might imagine to disrupt when entrenched interests are all making a ton of money AND are interdependent upon their neighbors in the value chain to continue to do that. You can't just hack a piece off because that bothers their neighbor, who in other circumstances might otherwise be a competitor, but has a shared interest. The relationships get complex.
None of this is fatal, but you must navigate it and play by many of the rules, particularly with #2. As Zenefits learned very publicly, the insurance industry and its regulators weren't going to let someone do what Uber did to the taxi industry - which was essentially to operate in the grey/black and just ignore the calls to stop. Insurance is a subset of financial services, and financial services is a powerful industry. For any startup, my advice is to build compliance and regulatory relations in from day 1. It's the least exciting thing you'll ever do, but is extremely important. Anyhow, I'm rambling. By no means do I want to discourage anyone from trying, I'm just trying to highlight some of the big things that are "different" in our space.
I'm particularly interested in the data piece because that's what I've done and built a career on in this industry, but it never struck me as sexy enough for YC. Appreciate the article.
[+] [-] ksar|8 years ago|reply
This is absolutely true. If you look at how regional, family-owned P&C carriers got their start, you'll find they started as brokers. These brokers found a profitable niche (call it motorcycles in California, or non-standard auto in Texas) that they wanted to own, moved on to become MGAs, and then admitted carriers offering multiple lines of business.
If you can find a way to own the customer as a distributor, you own the life-blood of the entire business downstream. As a result, P&C insurance companies are huge spenders on marketing (GEICO, spends $1.7B/year). It would be game-changing if this cash was used to provide utility to their customers above and beyond the insurance transaction.
I'm of the opinion that insurance premiums could be the new ad dollars - used to create products that promote lock-in, and much more consumer-centric insurance companies.
[+] [-] sharemywin|8 years ago|reply
https://www.guidewire.com/
If you're going to innovate on the software side in P&C you would need to figure out what they're not doing or could not do that would allow you to get ahead of them.
[+] [-] joshfraser|8 years ago|reply
As typically publicly traded companies, they have a fiduciary responsibility to their shareholders to maximize profits. And the only way to maximize profits is to deny coverage.
The purpose and value of insurance is amortizing risk. There are very few things that I think should be controlled by the federal government, but a single payer health care system is one of the few that actually makes sense from an incentives perspective.
In the words of Charlie Munger who's spent more time thinking about insurance than almost anyone: "Show me the incentive and I will show you the outcome."
Why does the US pay the highest prices for mediocre health care? Perverse incentives.
[+] [-] kozikow|8 years ago|reply
Shameless plug: At https://tensorflight.com we are working on P&C insurance and we focus on analytics for commercial properties mentioned in the article. Please get in touch at ( kozikow [at] tensorflight.com ) if you are interested in the subject.
[+] [-] osullivj|8 years ago|reply
[+] [-] SmellTheGlove|8 years ago|reply
The article is primarily focused on commercial lines - primarily P&C, but there are similarities in A&H. The distribution stack is complex and is a natural starting point to start looking at opportunities. The general tech community probably isn't going to look at risk pricing and the ancillary services associated with that - there's a whole discipline built around it and I don't know how many in YC's catchment would be FSA/ASAs or excited by that particular aspect. Plus, for traditional products in P&C and the Disability side of A&H, the industry is pretty good at pricing the risk. Nobody's squeezing growth from their underwriting ratio, they're doing it through products and distribution.
Anyhow, not trying to give you a hard time, just another angle.
> It's all done in spreadsheets.
No it isn't!
[+] [-] x0x0|8 years ago|reply