A pre-IPO interview with Daniel Schreiber of Lemonade

Blog : 21 min read

Last week, as part of the launch of our Insurtech 100 our CEO Matt Connolly and Jonathan Swift, the Content Editor of the UK’s Insurance Post, caught up with Lemonade’s co-founder & CEO Daniel Schreiber.

Daniel talks of Lemonade’s AI, European expansion and most interestingly, only days before they announced their IPO, his views on why an early IPO can help build an enduring, valuable business that will last the test of time.

To watch the full video: https://bit.ly/3cRHa7D
Transcript below.

Matt Connolly, Founder & CEO of Sønr
Daniel Schreiber, Founder & CEO of Lemonade
Jonathan Swift, Content Director of Insurance Post

Jonathan Swift: Hello and welcome to the latest episode on Insurance COVID Cast brought to you by Insurance Post and Insurance Age. Today we have the first in a series of specials linked with the Insurance Post and Sønr and the Insurtech 100 ranking, starting at the very top with the number one Insurtech: Lemonade. Now, I would like to say I have a co-host today, Sønr CEO, Matt Connolly. And also joining us from Lemonade we have its CEO and co-founder, Daniel Schreiber. Welcome both of you.

Daniel Schreiber: Delighted.

Jonathan Swift: So, Daniel, if I could just start by just asking to give, if you could give, us some idea about the latest numbers in terms of quality sales in premium Lemonade. And I know 3 years ago when we spoke, you mentioned about 90% of your product sales were first time buyers. Is that still the case?

Daniel Schreiber: So, good to be with you. The company has ballpark 800,000 customers today, ballpark $150 million of gross written premium, so there’s rough and ready numbers but they give you a good sense.

And yes remarkably, as best as we can tell, still about 90% of our customers are coming to us for the first time so we have an onboarding question where we ask, my urban bot asks, if you’d like any help in moving from another policy. Are you switching or are you not?

And about 90% of all customers get asked that question and about 90% of those who get asked say “I’m good.” So, yeah it does seem we’re over-indexing pretty hard on first-time buyers of insurance.

Matt Connolly:  Very cool. And I’m assuming that’s going to be very much true with the German market and also the Dutch. And, a question I had was regarding the launch in the Netherlands – which was in April if I’m right in saying, the lockdown hitting immediately after? How did that affect you guys?

Daniel Schreiber: We launched entirely remotely, so we do have a team in Amsterdam. We have a small team in Holland. That’s kind of our European headquarters.

So we launched in Germany and, to date, we still have zero employees on the ground in Germany. In Holland we did have people on the ground but it made no difference because they were all quarantined at home, so they might as well have been on Mars for all the difference that made. And it’s been really interesting and surprising to us, how much business has been able to operate uninterrupted, including the launch in Holland which happened while all of our offices were under lockdown.

But, beyond that, it’s been a pleasant surprise to see that the business has remained really strong – customers – and customer demand has remained really strong and all of our teams have been able to operate remotely with no real business interruption at all, so I feel pretty lucky about the way that played out.

Matt Connolly: Good. Well done. So, rewinding a little – Germany 2019, Netherlands 2020, what was it about those particular market conditions set that made them attractive for you guys? And, what next?

Daniel Schreiber: Germany has the largest insurance market in Europe. It is also one that is still very much dominated by traditional broker-based businesses. And so being a little bit controlling, we like the idea of going into the markets that are still very traditional. We think that give us the most pronounced difference between what consumers are used to and what we have to offer. And that is, of course, not true across Europe.

In the UK, that’s probably the most advanced, most digital, market across Europe. So that has advantages because you’re familiar with that mode of interaction, but it has disadvantages because the competition is more pronounced in that sense.

So, that’s why we went to Germany fast. And Holland was second for a couple of reasons. One, we really liked the demographic there in the early adopters of technology. The downside of Germany is that it’s a fairly conservative market, very traditional market. And we liked the alternative dynamics with Holland where it is still dominated by traditional players but the readiness of consumers there to adopt technology, and to look favorably upon American brands is perhaps different to what it is in Germany. So, there’s one dimension. The other one is that we are regulated by the Dutch regulators, and we committed to them that while we wouldn’t launch in Holland first, we would launch in Holland soon thereafter as it was practical, so those two things coincided for us.

Matt Connolly: Interesting. And, the second half of my question is that are there other territories across Europe that might be of interest to you? You mentioned the UK.

Daniel Schreiber: Yes. So, we do think of our ambitions as expansive, both in terms of product offering and in terms of geography.  Europe differs from America in two ways, one good and one bad from our perspective. The good is that it’s a single license. People outside of America tend to think of America as somewhat uniform or monolithic but from our perspective, it’s 51 different jurisdictions and each one has to grant a license independently and approve the rates and forms and the penalty, it’s incredibly cumbersome.

And Europe, for all its differences, is a single regulatory environment and one license, the Dutch license basically brings 450 million people under license for us. On the other hand, there are language differences and there are local legal differences and there are competitive dynamics that differ from territory to territory.

So, it’s a kind of mirror image of what we experienced in America where the market there is relatively uniform but the regulation there is incredibly fragmented. Europe is the other way around. So, we do plan to go and expand in Europe quite significantly. We’re not pre- announcing countries right now but that’s where our ambition stands.

Jonathan Swift: I think on your website there’s a drop-down where you can, you know, be alerted when you enter and I think it had most European countries, bar a few. I mean Russia, Turkey and Ukraine were three I saw that weren’t on there but there was a great spread. Can I ask Daniel, I know that one of your headlines that kind of was attracted by your entry into Germany, was the dispute over a little colourful magenta with Deutsche Telekom. Is there .. can you give us any update on that?

Daniel Schreiber: Yeah with pleasure. The Russia-Turkey thing is that they aren’t under the same license. I’m not going to go against those countries, per se. Coming back to your question about magenta. The dispute lives strong. We are embroiled in legal battles. We decided once Deutsche Telekom decided that they can’t hack an insurance startup using the color pink. We posed too big a risk, an existential risk, to Deutsche Telekom. It’s really kind of a… it’s such a silly story.

It’s been amazing to see how a company as large and respected as Deutsche Telekom has taken the battle to such extremes.

You know, I saw that there was a Kickstarter campaign by some company wanting to create a smartwatch and use the color pink in their ads. They took legal action against these poor people. There was an outfit of some 10 people in Bristol that was, where they’re doing servicing of Mac computers for educational institutions. They served them with a cease and desist. And you know, the list goes on. Blogs in America have been served with a cease and desist. Stuff that’s unrelated. If somebody came into Europe with a big color orange and they were a mobile operator, you’d understand why Orange would have a problem with that. But a smartwatch maker who uses the color orange. Surely not. And the same has to be true with pink.  The idea that my printer here has three colors and one of them is magenta and that is owned by Deutsche Telekom, just, it’s too silly for words.

So, having secured a court injunction against us in Germany, we obviously abided by the court injunction and we took down the magenta, and our website in Germany is red. But having complied with the court’s directive, we decided that this is just bully tactics, it really. This is just much too far. They’re not in the insurance space, we’re not in the mobile operators’ space. How far, how wide a birth do you have to give these guys before somebody is allowed to use the color pink?

So, we are doing a couple of things. We challenged.. they have a Pan-European trademark and we are challenging that. We’ve gone to court to argue that they should not have that trademark, that it has gone too far and in fact the law, the precedent, has changed meaningfully in recent years. There was an important KitKat decision by the European court that said that the shape of the KitKat couldn’t be protected – the shape in that case – if it wasn’t familiar in every country, every jurisdiction within the EU. It wasn’t, and they lost that trademark.

Surely, the color pink isn’t just associated with Deutsche Telekom in every country in the EU. So, on that basis we’re challenging their Pan-European trademark. They tried a bunch of bureaucratic things to get that delayed. Somebody else had challenged them on this 10 years ago and they managed to get it embroiled in legal nonsense for 10 years, but it looks like we’ve been able to overcome those hurdles and will have our day in court soon and, in parallel, we’ve started issuing challenges to their trademarks on a country-by-country basis. So, we’re challenging in the German courts, we are challenging in the French courts, and we’ll see if we can have some fun along the way.

Jonathan Swift:  So, how close are you then to offering insurance types in the US, in every state now in the US?

Daniel Schreiber: We are pretty close. So, we have licenses now in states that comprise about 90 percent of the US population. And we’re not live in each of those states for various reasons. The launching lags a little bit. It’s a dual step process. In the US you need your license and then you need to get your rates and forms approved, and so there is a bit of a time lag but we are live in states that comprise three-quarters of the US population. And that’s growing all the time, and the only sizable state that we haven’t launched in is Florida and I think that’s pretty imminent, so we’re feeling pretty good about closing out the final gap.

Matt Connolly: Daniel, I’ve got a question. I remember, back in 2016, I think your chatbot, Jim, made a bit of record speedy claim handle.

And as that certainly put Lemonade on the map, as far as I was concerned. You know we’d been looking at the landscape, following the business but that was the real standout moment. And from then on, you know every client interaction we had, every presentation that we saw, it was all referencing you guys. It’s classically sort of famous in the Insurtech journey idea.

Are there any other bits of innovation that you guys have pushed out that haven’t been picked up or that haven’t been covered, you’re really proud of and that you would like to have shared?

Daniel Schreiber: There are a bunch that we’re proud of and happy to share them. I think they don’t have the same dramatic moment of somebody’s Parker coat being paid for in 3 seconds. That was a seminal moment for us, and I guess for the Insurtech, more broadly. And that’s become pedestrian in the best sense of the word, you know. It was exotic back then, and now we do it many, many times, a day or an hour. It’s happening the whole time, so that’s fabulous.

There are two or three things that I think you might find interesting even if they’re not as photogenic, so to speak. One is the power of all of the integrated technologies that we’ve built to lower our loss ratio. So, back in 2017, early 2017, we had a horrendously high loss ratio – it was 300% or something. In the last quarter it was down to 72%. A lot, for the last year published, you look at the top 20 insurance companies and homeowners returns in America, the average was 82%.

So, within the course of the last couple of years, we’ve just seen this precipitous decline in loss ratio consistently without a single quarter reverse in that trend, to the point that we’re now at par, if not advanced, relative to incumbents. And, we’ve been in the market 3 and a half years. And we’ve done that, you’ve seen that kind of shaving almost 300 points, 296 points off our loss ratio while we were growing at 450 percent compounded annual growth rate. If you just juxtapose these two trends, you would realize that something is happening that flies in the face of conventional thinking; this defies insurance orthodox. Because the orthodox position is, if it grows like the weed, it probably is.

And, large insurance companies talk about lowering their loss ratio. They talk about shrinking to excellence. And there is this conventional received wisdom that growing fast comes at a cost of quality. And I think that makes sense if you’re built on a human substrate. Certainly being overwhelmed means that people cut corners, they do a less good job, they don’t underwrite quite as well and quality would pay the price. I think the deep change, and this is a much deeper change in terms of Insurtech than painted claims in three seconds, the deep change is that we’re showing what can happen when you’re built on a digital substrate. And you actually define the gravitational pulls that govern insurance up until this point.

So, suddenly you see that we build a system using 21st century methodologies, then masses of data and rapid growth aren’t an impediment to increasing quality and good results. They are the enabler of it; it’s flipped that conventional thinking on its head.

And I think that draws, kind of, draws out what I do think as a real fission or crisis moment, although it will be stretching over so many of the people will feel it as a crisis moment, which is to say that insurance companies have lost dominion over data. Way back to, you know, from the coffee shop of Lloyd’s in London in the 1600s, insurance has co-evolved together with statistics and data. And you know, during that same time that Lloyds was creating his coffee shop, Bernoulli was coming up with the law of large numbers and pricing and building lots of probability theory came into place which co-evolved, as I stated, over the last 300 years. And if at any point during that time I had asked you who are the bastions of the world’s data and who is home to its final statisticians, you’d have said an insurance company.

And I think even as we see, as the 90s if I’d ask you that question, you’d have said an insurance company. You know, if I ask you that question today, it would not occur to you to say insurance company. You know there’s Google and Twitter and Facebook, and all of Silicon Valley.

So, insurance has lost dominion over the core building blocks of insurance and they’ve lost that to tech.

And what I think, the example that I just gave, is juxtaposed graphs suggest is that there is a way to claw a bit back.  If you rebuild  insurance from the ground up, on a digital substrate, free of all the legacy that encumbers the incumbents, you can get to an entirely new mode of underwriting policy rich – risk – which is the, obviously the very core of insurance.

Two or three years ago when we published that blog post about AI Jim, if you asked insurance executives about Lemonade, I think they would pass the backhanded compliment of saying “they’ve got a cool app, and they’re good at PR.” And the implication would be “but they don’t get the core of insurance, they don’t really get into insurance.

And I think that what we’re seeing is that technology can outperform expertise. I think this is true broadly, places where technology and Big Data is outperforming expertise – wine tasting. It’s amazing how many different fields where we thought were dependent on human experts. We’re seeing that Big Data machine only can outperform.

 But that’s, I think, what you’re beginning to see in some of Lemonade’s results, which is that three years later, zoom forward from that blog post about Jim, and what you’re seeing is that we’re actually outperforming 90 percent of the iceberg that is submerged, forget the 10 percent that’s above water. It’s a fundamental system or machinery of insurance that is being transformed in powerful ways. And that ultimately is what gets Insurtech its heft and it’s that 90 percent of the iceberg that ultimately sinks the titanic.

Matt Connolly: But if I’m right in saying, also the 10 percent, which is your Maya stat – that’s your customer satisfaction front end ticket service tech. That’s outperforming, in a relative sense the customer satisfaction scores from your human teams. So, is that some chatbot is good for humans, some is bad? It can’t be as simple as that?

Daniel Schreiber: No, it isn’t as simple as that. But I do think that building a company on a digital substrate allows you to do this, kind of, double whammy of lowering costs or increasing customer satisfaction.

This isn’t a low cost airline where you have to pay for your peanuts.  This is something where.. this is Uber, where you end up with a better experience, real-time information, click a button and lower costs.

And you see this, it is only technology that can deliver that. Dramatically, not at the expense of customer satisfaction but to the delight of customers. One of the things that we have tried to do, and Shai my co-founder is really a thorough leader on a global basis around, is how to meld bots and humans, because so many of our interactions do that so poorly.

You call up to check your flight, I fly United Airlines and every time I call, I have to feign an American accent for them to possibly understand what I’m saying, and you’re pulling out your hair like just get a human on the phone. And it was nice, a couple years ago somebody delivering a review of Lemonade and they said that Maya had much more personality than the person she spoke to at Liberty Mutual.

Matt Connolly: How’d she handle that?!

Daniel Schreiber: She was just fine, but I think what we try to do is only use technology in places where it can outperform humans, and not use it indiscriminately, so yes AI Jim will play a third of our claims instantaneously but two-thirds of our claims will have human intervention.

Maya will answer one-third of all customer enquiries instantaneously without any human intervention. And when she knows what she’s doing, that’s amazing because you typed noting, I’d like to add my spouse to my policy and she said sure what’s her name and did it for you, you get this instantaneous response, you’re happy. But if she doesn’t, if she hasn’t got an incredibly  high level of confidence that she will understand the question and knows how to answer it, it gets past  right to a human and you don’t know the difference, you just asked a question and you got an answer.

So, taking the time to think deeply about what are the things that technology outperforms humans in, and what are the things that create a seamless integration that weaves human and Bot. It’s something that’s Shai was quite insistent on from day one and I think it shows.

Jonathan Swift: So Daniel, you know, so you’ve talked about you know the handling of the part that claims in three seconds, I know where in diversification, that diversification hasn’t been talked about recently with Lemonade is petting insurance, and of course I think people love their pets more than they love their park because of how expensive they are. But why enter petting; what other sectors are you possibly think about launching into?

Daniel Schreiber: Yeah, they do love their pets a lot more than their partners. And you, over the last few decades I guess, that it’s become accelerated if anything. There’s been a deep difference or change in how people view pets, or how they once viewed. People don’t view pets as property, they think of themselves as pet parents. They walk after their dog in the street with a bag and pick up their poo. You don’t do that for your property.

And, pet insurance began live as horse insurance actually, in the 19th century, and really horse insurance was an early version of car insurance. So, it’s born of that kind of property and casualty perspective and a lot of policies and the way people insure pets has been, therefore, increasing the odds of how they think about their pets.

And this manifests very powerfully in the numbers in the United States. In the UK, pet insurance is pretty well-penetrated and common. I think something like 40 to 50 percent of people own a pet insurance. In the US, it’s less than 2 percent, so this is a woefully underserved sector. While we know that about two-thirds of our customers have pets, so we thought this is a great product, good fit to our existing customer base and the way we think about new products and pets is the first formula beyond homeowners but it’s not our last. So, a bit like the European countries, it’s not a question of “why this?” It’s a question of “why this first?”

And we just felt like this was an area, again being contrary, being underserved, it was not being given the kind of love and attention that the potential represents. And that is a great fit for our customers, and we do try to think in a customer-centric way, so we sell condo insurance, in our Parliament insurance.

But you could say that “well, why only do just commercial real estate insurance? Because you insure the apartments and not the building at large?” And that makes sense from one perspective but not from a customer-centric perspective because the customer who is insured with you in that apartment block isn’t thinking about who is insuring the building at large.

That’s just a totally different customer set of needs. They are thinking about the poodle around their ankles, so once we’re thinking in a customer-centric way, and we think about, well with, catering to these needs of our customers, which needs are under-catered for? Where are we forcing customers to go uninsured or to go insure with other insurance companies? And that’s really how we prioritize.

And just tying this back to another question you answer that the 90 percent of customers who are first-time buyers of insurance, one of the fabulous things that we’re seeing is that we are catering, we are competing with non-consumption.

That’s what that implies. Our customers are not being taken from someone else. The entire insurance industry seems to be predicated on “I switched, and I saved” proposition and we’re not playing that game, we’re not playing that sport.

But what that also means is that we therefore have a bit of an opportunity and a responsibility to grow with our customers as their needs grow. So, an easy progression is they are renters and then they go and buy their first home,  but they also go through other life cycle, and so we want to be there to cater for their various insurance needs as they go through those joyous occasions.

Matt Connolly: So Daniel, now you can probably see my dog in the background here, proudly sitting throughout the interview, and if there’s one thing I can do with after lockdown is really some therapy for me and him.

I think he’s like the only one thing I’ve really struggled with, because of course we’re not going to the meeting, we’re just sitting here, so I’m with him 24/7 seemingly at the moment and our relationship’s getting a little frayed. Anyway, if that was part of the insurance package, I’m signing up. So, I had a question..

So, we’ve got insurers paying back a bunch of money for cars sitting on the drive doing nothing for months, and all the rest. We’re also all spending so much time at home. Have you seen, in particular around claims, and any sort of bizarre and unusual sort of maybe persons or unforeseen circumstances, you have experienced in the business?

Daniel Schreiber: Nothing too exotic. We do every now and again see dramatic claims but not ones that I can attribute directly to the COVID or to the lockdown. And I’m not sure entirely, you know what the dynamic is going to look like quite yet.

So, in broad strokes, insurance has seen a modest decline in claims, nothing for car insurance. But at the same time, a fair number of people have fled the city to wherever it is they flee, like a lot of apartments are empty at the moment and we will see what happens when they come back. So, we’re kind of waiting to see how this all comes out in the worst.

But so far, broad strokes, our business has been kind of business as usual. And obviously that’s a stark contrast to many areas of insurance, which either like car insurance has, you know, kind of swung wildly one way, or if you have the misfortune of insuring the Olympics, business continuity, as acts are just discovered in France or other places, that’s obviously swung way the other way.

We’re seeing our business has continued and most of our key metrics have continued to progress in a way that if we didn’t know that the pandemic and lockdown was going on, you wouldn’t know it by looking at our numbers.

Jonathon Swift: Daniel you’re obviously quite famous in your gift that you’ve switched back to Direct Relief. How is that going during the covid-19?

Daniel Schreiber: That’s great, we didn’t do a wholesale switch; we offered our customers if they would like to switch. This was really a customer-driven and choice.

Something like, if memory serve, 40 or 50 thousand customers made that switch. We will be publishing, our Give Back Day is July 1st, so we’ll be publishing our numbers when we get back quite shortly, and then you’ll get to see the numbers in a full impact. But it’s been a big deal. It’s been a big deal and we’re really pleased.

I think our community at large felt very good that we were able to leverage our business and to try and do our share, and that’s really all it is – you know, everyone does their little bit to try and move the process forward.

Jonathan Swift:  This you touched upon it when, with the kind of use, the pain and claim but looking back on the history of Lemonade, and why we’re short in space, if they evolved, how’d you expect, together so to speak?

Daniel Schreiber: I’m not sure they have. I’ve been a little bit surprised, with mixed feelings, not to see more, not to see more full-stack insurance companies, particularly in the US, and I understand that there are regulatory hurdles there and the Southwest hurdles, having you know developed all the scar tissue for them for all the things that it takes to get there, I have respect for the challenges.

But, had you asked me X years ago, you know, would we see much more fierce competition from other companies who understand the power of building this from scratch, the digital substrate etc., I would’ve been sure that the answer is yes.

And as I look around the landscape today, there are some great Insurtech companies out there, but I would have expected more, I would have expected more that to the full vertical integration rather than being MGAs and fronting father, you know, legacy insurance companies.

This is such a vast industry where there’s opportunities to do so many different things and add value in so many different ways, but that particular aspect of it was a.. has been a surprise to me. I still think there’s opportunities out there. I think that insurance remains the largest disreputable industry out there, and given that that is the case, I’m surprised that not more people have flocked trying and take the run at it.

Matt Connolly: Daniel, you created the digital substrate, have you looked at licensing that out to others for your business?

Daniel Schreiber: No, is the kind of mono-syllabic answer. We have built that, but we think that the true power is in the integrated system. The business model, the Give Back, the way we train our humans, the way they complement the bots, how each of those systems feed in and are fed by each of them.

So, every, my interaction & Jim’s interaction, every time AI handles a support inquiry, it is, every marketing campaign, all of those are generating tons of data. We’ve got over 2 billion data points that we’ve collected so far, so constantly being enriched but also constantly enriching each of those interactions and I think just taking out the piece claims in isolation from all that, is missing the true power.

It’s really a system, an organism with aligned interest. We take all of these components into play and hopefully increasing it being fine-tuned one or the other, so no.

Jonathan Swift: Have to ask though Daniel, since the last day there’s been a lot of speculation about an IPO and Lemonade. Could you shed some light on the truth here?

Daniel Schreiber: I’ve not ever commented on a future fundraising or IPO and I will not do that now either.

I will tell you that I’m predisposed, without telling anything about Lemonade, just to show you kind of my thinking about this.

I think younger IPOs are good. If you look at the class of 2019 and see how the dynamics have evolved in the technical industry, nothing to do in the Insurtech industry, and I’m not sure that the delay of IPOs has really been good to anybody – investors or the company themselves.

In the late 90s companies like Amazon, and shortly after that, Google and Microsoft, back in the day. When they went public, they were like four years old and did a hundred million dollars of business. And 10 years later, it was 14 years old was the average.

So, there’s been a sort of protracted adolescence, this delayed adolescence if you like. And just like delayed adolescence with humans isn’t necessarily a wonderful thing. I’m not sure it does all that much good for the long-term development or health of the companies, that’s been fueled by an abundance of capital.

You can raise huge amounts of money and better valuations with less scrutiny when it gets to the public market. It is very tempting to remain in a cocoon, but as I said, if you think about building something for the next 20 years, and you ask yourself not what will be nicer, better, cheaper for the next 12 months, but what is the best platform to build an enduring, valuable business that will last the test of time?

I do think that militates in favour of you taking your pain early and growing in that environment. And in general, some of these companies that did go public in 2019, they did all of their growth in the private market, they heaved themselves on to the public market at an already fairly mature state, and the investors threw up all over that, didn’t approve of that, and then we.. you lose the power that Amazon had, and others had, that they went early enough to reward investors with their growth, which then go back to the market capital to fund more growth. And that created a virtual cycle that has propelled those companies for decades. So, for all of those reasons, I am predisposed to look favourably upon younger rather than older, kind of an age over, IPO.

Matt Connolly: Interesting. Okay now last question from me. So first of all, huge congratulations on the number one insurer on Insurtech 100. Well done you guys. We’ve touched a little bit on the European market, but Lemonade in its entirety, what’s next in the business?

Daniel Schreiber: We hope to continue to progress along each of the trajectories that we think of. So, geographic expansion is one. So, there’s no question that more countries, more states, more geographies is somewhere we’d like to go. It is kind of striking, but I mentioned earlier that we launched in Germany, our first European foray, and we have zero employees on the ground there but even in the US, our two largest markets on Texas and California, we have zero employees in either of those states.

So, again coming back to this alternative strategy, an alternative way of doing things that militates in favor of being able to grow faster for geographic expanse, because the traditional mode of coming into town, building skyscrapers and putting ten thousand people in it, inhibits geographic expansion and we’re not inhibited in the same way. And if we launch in a country where it doesn’t work out, that’s okay, because it’s on a bet the company moves, so we can be much more experimentalist and move faster. Some of the best market research is just to do it and find out from the market, rather than analyze it. So that’s one.

The same philosophy is guiding us along product launches. So we mentioned patent would be public about the fact that we will launch perhaps soon, and we’re still being cagey about what comes after that but we are working on future products, so that’s another one.

And then, you have to continue to build down deep. So, some of the, that kind of, 90 percent of the iceberg that’s submerged when you’re more than doubling your business year-on-year. You’re hiring a lot of people,  you’re creating new tools and infrastructures and new products, you constantly got be able to have enough resources to invest in that infrastructure to make sure that you’re not cutting quality, that yes as you earlier pointed out Matt, as your business grows the lights grow faster, loss ratio declines faster. We never want to be in a situation where we’re selling dollars for 90 cents where our growth is being funded by bad business practices or at the expense of  customer satisfaction, and that requires a fair amount of  inwardly focused innovation as well.

Jonathan Swift:  Thanks so much for your time Daniel and also thank you very much Matt.