Not so fast: Insurance will be heavily disrupted, but not from you would think.

Disclaimer: I have been working on the insurance business for the last two years and have a lot of inside information to support my points, but it is propietary and I am tied by professional secrecy, so for the purposes of the post here I will just refer to public available information and/or my own corpus of knowledge. 

There seems to be a lot of turmoil as of late regarding how the Insurance industry is going to be ripped to shreds by new incumbents. There is the customary Techcrunch article on the matter (Later regurgitated here by Enrique Dans with interesting conclusions put on top. I will come to this later). Personally I find this one far more enlightning, with Warren Buffet, Google and World Domination thrown in for good measure. A comment of this post, launched by my good friend Gonzalo Martín in a community of thinkers I belong, started the whole ball rolling for me to write this piece. He goes as far as implying that the whole autonomous car thing is a huge Trojan Horse set by Google regarding taking over the Insurance market. 

But there are no lack of examples all around the Web as you can check here and here.

And I cannot agree more. It's going to be a massacre.

I have been on the belly of the beast and all the warnings are there. And worse than that, the industry is not maniacally trying to reinvent themselves, like for example, banking is trying. They are playing the sleeping giant part, which of course, will only make things even worse.

So, Am I joining the ranks of the heralds of doom? Sure.

My only point is that the killers won't be who everyone thinks will be. There are two main suspects on the police line-up: Google and IoT. and sure, they look the part and will have something to do on the overall picture. But I strongly advise to keep an eye on that meek, milquetoast Keyser S?ze.

I have found that people are real suckers for good narratives, even if what is being told is basically, a huge bunch of crap. Checking the facts and questioning a fluid narrative is cumbersome. Brain is a lazy bastard and we are genetically programmed to connect the dots ASAP in order to provide a sound, tranquilizing, easy answer to complex solutions as soon as possible. Otherwise, we would get blocked when we need quick responses in order to act. 

You only have to check the examples on how well crafted business narratives can fool entire swarms of very intelligent people: Enron and Gowex comes to mind. Reality doesn't matter until it explodes upon you. 

Going to try to debunk several common narrative places which are a huge pile of manure. I will quote verbatim several contents from some of the aforementioned articles and try to get some facts right. 

Now, imagine if Google got into the insurance
business; what advantages would they have? Just imagine. Insurance is
the business of risk management and making informed assessments based on
the management and analysis of a variety of (preferably accurate) data
samples, variables and information. Google has access to millions of
people around the world and tons of private data. Google probably knows
more about you than you do. (That's an exaggeration, but you know what I
mean, and this trend will likely continue.) [From the SeekingAlplha article]

Until the part "Google probably knows" part everything went fine. Then suddenly the whole sentence turned into a fireball that suddenly exploded and came tumbling down into a mass of scorched and charred remains.

Been writing already on the matter extensively, so will reduce it to bullet points:

• Contrary to popular belief, Google do not know jackshit on you but a collection of random tidbits that could or not bould be relevant to the matters at hand
• By its own sniffing and behavioural nature, Google cannot reach beyond 20% of your real relevant data
• The "trend" will be eventually stopped from legal action coming, more likely from Europe. And the Browser wars from Microsoft will be dwarfed by the sheer size of the lawsuit.

In fact, private health insurers already work on a similar basis, using test results, medical reports, and other medical information. But of course health insurers cannot legally use this data to offer better prices based on gender, age, or the costs incurred by the client over time. As more and more of us start wearing devices to measure key health indicators such as heart rate, exercise, or the amount of hours we sleep, some insurers, particularly those that have added “wellbeing” to health, are looking at the possibility of working with their customers using a data base of such information to offer new or more attractive services, but face the same problem: the customer must trust that this data will not be used as an excuse to increase premiums. This is a complex issue, given the asymmetric nature of the relationship between insured and insurer, but that will inevitably create winners and losers in the insurance sector, where customer trust is the true competitive advantage. [From the Enrique Dans article]

Will try to separate wheat from chaff on this classical hodgepodge of real data, perceptions, biases, narrative of varied quality and not having a deep knowledge of a concrete subject.

For a start, insurers DO use data to offer better prices based on gender, age and the costs of the customer over time. That's their business. Arbitrage risks against premiums.  It's classical the use of data regarding age on health insurance. Here you can fin an article on the matter based on spanish market. I won't get here into deep legal mumbo jumbo, but the BOE is available to everyone. The law does not forbid adequating the premium to the real amount of risk involved, based on the available data.

In fact, the underlying disruption regarding insurance and Personetics (And here is where IoT plays a thing) is the jump from proxy measures and statistical data to realtime, consensuated relevant data between insurer and customer. 

The whole idea behind getting better data is that good drivers and people with low risk are subsidizing people with high risk via a mutualization of risk. For example, that, basically, hurts good drivers vs. bad drivers. There is a classic Progressive campaign regarding this basic fact. 

Therefore the capture of more data will work for good drivers and low risks in order to lower their premiums, exactly the opposite point that the article suggests, that data is going to be used to raise premiums.  

Insurance companies just want good risks. They are not interested in squeezing dry their customers, just to decant the high risks and therefore lower the combined ratio in non-life insurance, for example. And being able to lower the price and be more competitive.

What do they need to do precisely that? The right data to properly evaluate risks. And that data won't come alone from sensors and IoT. Again 80% of that necessary data is out of reach to behaviour observing probes and mechanisms. Which lets Google out of the equation, for example.  

But the blurt ends with a high note, as the real key of the matter is on the last paragraph. Even if it's not intended, all the right notes are pressed:

• What it is ending, as with most business models, is the asymmetric nature of the information. But from both sides, not just from the insurer, but also from the insured. Fraud comes from both sides of the fence.
• There will be winners and losers in the market. i.e: The ones who grasp how the basic insurance premise has changed vs. the ones who don't get it
• Customer trust is the real competitive advantage. Spot on. And the anal probe attack which is characteristic of the likes of Google is not really something trust-inducing. 

Even more, IoT lays a foundation for more customized insurance plans. The old model of tiered insurance lumps you into a plan that considers the “average” person’s risk (or your home’s or car’s) — which means you essentially pay for those average mistakes. But you aren’t average; you are unique. [From the TechCrunch article]

That's simply, an absolute beauty of a sentence and can't be truer. Problem is, it's followed by...

The new model will adopt strategies that observe your behavior and the specific measures you take to mitigate risk, which will allow for complete personalization to match your individual demands. [From the TechCrunch article]

Here comes the fireball again. Sorry, there is no "complete personalization" without heavy use of personetics. Yo need to put on the table the whole enchilada: Personetics, Human Data Models, Positive data exchange... All the arsenal I've been talking for ages. You can put a trillion sensors on the table and you will end up on the same place: 20% of relevant data at max.

I could be debunking standard narratives on the matter for ages, but I think my point is made. So, once we have the corpse on the table, the questions arises:

Who will be the killer? Who will be the real disruptor of the whole, classical insurance scene?

My thesis is that, likely, the killer will be some small, unknown, customer centric, personetic based, with a non run-of-the-mill loyalty proposal, a superb customer experience and a clear win/win proposition company outside of the classical insurance ranks and the Anal probers. A whole differential proposition coming from an unknown country. Which will take seriously and in strategic fashion the whole personetics architecture and uses science and technology to support its clear and defined strategy, not the other way around.

Or I could be completely wrong and the cake will be eaten either by Google or some big classic Insurance company. They just have well, to be born again and lose its basical fondational cultural patterns in the process. Unlikely, but could happen.

Because, as Peter Drucker once said:

Culture eats strategy for breakfast.

But hey, chances are. It's just my internal risk assessment system telling me to bet against it. Heavily.