A new player has entered the Massachusetts homeowner’s marketplace. Not quite an insurtech, not quite a traditional homeowner’s insurance carrier, Swyfft’s goal is to transform the homeowner’s marketplace by working with traditional independent agents rather than by cutting them out. Using Big Data, AI and a “one-click” style interface, Swyfft aims to make agents able to quote homeowner’s more quickly and accurately than ever before.
With little fanfare, the company has already established a presence in seven states including Massachusetts, which it officially entered approximately six weeks ago. Agency Checklists had the opportunity to speak with co-Founder and CEO Sean Maher to learn more about the Swyfft philosophy and what it looks for when partnering with an independent agent.
Sean, could you give us a little background about yourself
I have only been in the insurance industry about seven years, my background is more in analytics and technology. Richard Trezza, the other co-founder, has been running MGA’s of one sort or another for about 25 years. He took over his mother’s insurance brokerage when he was 22 and turned it into an MGA. Has mostly been doing commercial property MGA’s for the past 20 years. We joined forces, about seven years ago, to really bring some more interesting analytics and underwriting to that business.
Could you tell us the name of the commercial MGA? Is it something that agents might already know?
Sure, it is called Core Programs. http://www.core-programs.com/#home
So how did Swyfft start from there?
We started Swyfft about two and a half years ago, but about four years prior to that, we actually started a commercial property MGA. That commercial property MGA is a traditional MGA in every respect except one, and that is we founded that company predicated on some analytics that we had developed, that we thought gave us a better bead on expected losses due to fire and to hurricanes. It was really that, those underwriting analytics that underlie that commercial company.
About three years ago, we realized that everything we were doing on the commercial side from an underwriting analytics perspective would probably work as well if not even better in homeowners.
We had this “aha moment” when we realized that not only did we think our analytics would work as well, if not better for homeowners, but there were data sources that had become available that [when] coupled with our analytics, could really change the entire user experience for buying insurance. By that, I mean something pretty simple and pretty specific, which is that we only need an address, no questions, and we can give you a bindable quote in 4-5 seconds.
“Old school” ways of buying insurance, whether you’re sitting in an agent’s office or doing it online at GEICO or Esurance, it’s answer 40, 50, 60 questions about yourself, your home, etc. before you even get a quote. That was fine. There’s nothing terribly wrong with that, but in this day and age, that’s not as acceptable a user experience, if you will. More and more products and services are going towards that Amazon one-click model, where people just are expecting things to be much quicker.
Primarily, the distribution model is 99% independent agent…We don’t market Swyfft directly to consumers at all.
A long story short—we thought we could actually do that—we thought that our analytics coupled with huge data sources where you can get information about homes would allow us to basically be able to [radically] change the user experience. Meaning, just enter your address and we’ll give you a bindable quote in three or four seconds, while still maintaining underwriting integrity.
So would you call Swyfft an insurtech?
We see ourselves as sort of this hybrid Insurtech. We are a traditional insurance company in the sense that we really come at it from an insurance company’s perspective: Meaning we know we have to underwrite properly, and write a profitable book of business at the end of the day.
The Insurtech bit is that we want to do that in a way that is much more modern, if you will.
We got funded about two and a half years ago, and launched, about a year ago, in our first state which was Alabama. Since then, we have been expanding into other states: New Jersey, California, Nevada, Illinois—Massachusetts being the most recent one. We are going to add Texas and New York to the mix pretty soon as well.
The other thing, which is worth noting and which I think really separates us from the other Insurtechs, like Lemonade, and Hippo, is that instead of creating this better user experience to go direct to consumer, we think it is just as interesting to offer this great experience and this great product to agents.
Who is funding Swyfft? Can you give specifics on Swyfft’s backers or investors?
I am going to answer that vaguely since our angels prefer to be the wizards behind the curtain, so-to-speak. It is not technically a secret, but all I can say is that they are not traditional tech VC’s, but rather very well respected insurance industry investors.
And Swyfft, is it a the company itself, or is it acting as a sort of a managing underwriter?
Technically, again, we are an MGA. We use a front company and then we have a large reinsurers that takes 100% of the underwriting risk.
So, who would the paper come out on, if we may ask?
It’s Clear Blue Insurance Company. [Agency Checklists: Best rating A-, Size VII] This is the admitted paper that we use in Massachusetts and elsewhere.
What insurance group Clear Blue is part of or is it a separate fronting company?
It is a separate fronting company. Jerome Breslin is the CEO. He used to worked for Bank of America and they essentially saw the success of the State-National model and said, “Hey, there’s room for another fronting company out there.” So they started that up about three or four years ago.
[W]e don’t have underwriting restrictions, for example, we will write on Nantucket and we will write anywhere on the Cape.
Could you tell us about the reinsurers backing Swyfft through Clear Blue?
Not specifically. I can tell you that is not public record. The fronting company is, but all I can tell you about the reinsurers are that they are all large, well rated reinsurers…all A.M. Best A rated and financial size XV—meaning well-established.
How about the policy forms Swyfft uses, are they ISO forms?
Yes, they are. We write a standard HO3. It is pretty vanilla, pretty standard. We have really one endorsement that is worth mentioning and that is a guaranteed replacement cost endorsement. We offer that on every policy, we give guaranteed replacement costs on coverage A. We feel like that is our risk to take, not the policyholder’s. We just feel like if we all agree to $400,000 replacement costs, and the property burns down and it costs $490K to rebuild it, then we’ll pay $490K.
When we started, we didn’t like those stories about people paying their premiums for ten years and then, their house burns down and it turns out they get a check for $200,000 less than what it costs to rebuild the house.
What is Swyfft’s distribution model? Direct or independent agent, or a combination of the two?
Primarily, the distribution model is 99% independent agent. We do allow homeowners to buy direct, but that functionality is mostly as a benefit to the agent so they can just send a link to the quote page to the insured and the insured can complete the transaction themselves. In that case (when the agent sends a link), the customer is assigned to that agent. In the rare (less than 1% of cases) where someone has bought direct, we currently service those policies (although we’re exploring assigning them to agents). We don’t market Swyfft directly to consumers at all.
Maybe it’s because of our history with running traditional MGA’s, but we actually think that tech- enabling independent agents is, again, a more interesting combination. Notwithstanding all the stuff being done at Lemonade, and everyone else who is working on making the user experience better, at the end of the day, insurance is a complex financial product for the average consumer.
The whole industry is going to change and become more tech enabled. We think it has to. Unlike a lot of people, however, we don’t think independent agents are going away any time soon. We actually think there will be an ongoing role for quite a long time for agents, and in particular, independent agents.
I think you are going to hear a few hallelujahs from independent agents here, who only hear about new Insurtech ventures that will displace and replace independent agents.
It’s interesting. It is one thing to have a great user experience, but it’s still a financial product, right? Before you go to Lemonade, like a consumer, and just hit buy, there is still a lot of hesitation, right?
It’s like, “Oh, my gosh. Is this the right product? Is this really as good as … I mean, this is great. It’s cheaper. It’s easy, but is it really as good as Allstate or State Farm or Travelers. Who knows? The independent agent can answer all those questions quickly.
This is the way we see it at Swyfft. We think that the better user experience just allows Independent agents to write more business, because they are not spending so much of their time doing administrative work. We are trying to take most of that off their plates, while still letting them play the role as the advisor. We add value, add service, and help agents get rid of a lot of the administrative stuff.
And how are you going about appointing agents? Is there a volume requirement?
We are appointing them as producers. They have the ability to go in and use our system. We give them a log-in, they can go in. They can bind business through our system, but the contract does not give them binding authority. It is just sort of a traditional producer agreement that says, technically the binding authority is with us, but we give you access to our online system and we don’t have any volume or production requirements for producers that we accept.
We would love to write, Boston and Worcester. The way our models work, they do not have any strong preference for urban versus rural versus whatever.
[Agency Checklists: A producer application and a one-pager on Swyfft for agents follows this interview].
Do you have any specific requirements or limits on the number of agents you are looking to appoint in Massachusetts?
In terms of agent appointments… we are a fairly open, meaning we are not super picky about the kinds of agents we are looking to appoint. A lot of that is because we really do rely on our system and our analytics to do the underwriting. We want good agents, don’t get me wrong. We don’t want, sort of, bad actors particularly if they are going to purposely try to give us substandard business. We do consider ourselves a standard/preferred market and we think that’s the kind of coverage that we provide.
How about the commission structure? Could you tell us a little about that?
Yes, no problem. We pay 15% on new and 10% on renewal.
Do you offer Direct Bill?
Yes, we do.
Any particular terms on the direct bill?
No, we offer escrow, credit card or e-check. We do that on an annual or monthly basis. Well, on credit card and e-check, you basically can do annual or monthly withdrawals automatic from your credit card or we’ll do it escrow and sort of set that up with your bank.
Okay, but there is no direct bill in the mail from Swyfft. That is currently not an option?
Correct. We do not do snail mail right now. You just choose what you want and we bill you electronically.
On the policy billing, if it is monthly, is there no charge or is there a fee?
That is a good question and it varies by state. But, I believe there is a nominal charge.
What about your pricing as compared to other carriers?
I think our pricing is competitive. We think we can become a real player, especially in terms of coastal insurance. I don’t know how this is going to play necessarily in Massachusetts, but I can tell you in Alabama and in New Jersey having an A minus seven admitted carrier that really doesn’t have any coastal restrictions has worked pretty well.
Again, I say that meaning we don’t have underwriting restrictions, for example, we will write on Nantucket and we will write anywhere on the Cape. The way our analytics work, we try to price risk accordingly.
What we are seeing early days in Massachusetts is that it seems like the agents that we have writing with us right now, seem to be sort of playing well on the coast, out on the Cape, a little bit on the islands and kind of up the coast towards Boston. It’s early days, though. We are writing some business inland as well, but it does seem like … and we have seen this other places where if you look in New Jersey, we’re writing business throughout the state, but definitely sort of skewed towards the coast. I kind of expect the same thing to happen in Massachusetts where I think we’ll, I suspect, that from what we’re seeing early on, that we’ll be a good market for agents to write coastal business.
Besides writing coastal risks, how about writing urban risks?
Again, we have no problem with them. We would love to write, Boston and Worcester The way our models work, they do not have any strong preference for urban versus rural versus whatever. They are idiosyncratic, if you will.
Can you tell us how many agents Swyfft already has in Massachusetts or overall?
I can tell you in broad strokes, across six states, we have about 1100 agents appointed right now. Maybe, so far, 700 to 800 of those are actually quoting writing business with us. Massachusetts has been our newest state, so I think we only probably have 40 or 50 agents at this point in Massachusetts, is my best guess. We have a couple hundred in New Jersey, about 300 in Alabama and maybe 150 so far in California. I can definitely tell you that with respect to Massachusetts, even though we have only been there six weeks or so, definitely we are seeing growth.
How is the company doing profit-wise?
Our new business sales, as of last month, are 10x what they were in January, just seven months ago. So, we are definitely playing well. Our bind to quote ratio is in the low 20’s which is about where we want it to be, meaning we are not just out there undercutting the market, winning everything.
With the bind to quote ratios in the low 20’s, that’s a good sign there’s a nice balance between production and underwriting.
If you get too much lower than that, you start to worry that your pricing is too high. If you get bind to quote ratios too much higher than that, then yeah you worry that you might have an underwriting problem. If you’re out there winning 60% or 70% of the quotes that you put out. Empirically, you kind of say, a lot of companies where that’s the case, they end up having underwriting problems at the end of the day.
We’re growing very nicely in all of our states, but all of the metrics that we look at tell us that it’s good growth. It’s good quality growth. And we put it all down to that user experience and the analytics that kind of drives the pricing a little bit different than other players in the market.
What would you tell an agent asking about the underwriting engine and what makes it so special that it produces a quote so fast and, as you believe, so accurately?
I will give you one example. Some of our models are proprietary, so I don’t want to say too much about them but, one of them is public and we have a patent. I think it will give you a feel for maybe how we’re different from traditional matrix underwriting.
It has to do with wind in coastal areas and in particular or mostly hurricanes, but any high winds. I’m going to tell the story, sort of, for hurricanes specifically. I’m going to use Florida as an example instead of Massachusetts.
If you hear a hurricane is hitting Miami and the weatherman says that it’s 120 mile an hour hurricane that’s going to hit Miami. What the weatherman’s actually quoting is called the open terrain wind speed, which means out over the water or over a wide open area, maybe the airport, something like that, about ten meters above the ground, the wind speed’s going to be 120 miles an hour. But if you went into Miami and actually measured the wind speed on the top of, say, five different houses that were all a mile from the beach, and all maybe within a mile of one another, you probably wouldn’t find 120 mile an hour wind actually hitting any of those homes. You’d probably find wind speeds as low as 100 and maybe as high as 125. What’s driving that variation in wind speed is the natural environment around that home. Very simply put, other buildings and trees can block the wind before it hits a particular building, or like I said, if you’ve got a home that’s right next to the airport, there might be nothing that’s blocking the wind before it actually hits your building.
The reason that that’s so important is because expected losses are very nonlinear with respect to wind speed, meaning if you’ve got a 25% increase in wind speed from 100 miles an hour to 125, you don’t see a 25% increase in expected losses, you see about a 400% increase in expected losses.
That expected loss number goes right through to the adequacy of the premium. Depending on what the environment is around a home, you could have the very same home next to the airport in Miami where the adequate premium might be $4,000. You could take that very same home, same construction, same everything, move it half a mile away into an area where there’s other buildings and trees around it that are blocking the wind and the adequate premium could be as low as $1,000. That sort of non-linearity in wind speed is really important in figuring out what your expected loss is and what you really need to charge as an insurance company.
What I just said, that concept, is not lost on anybody that actually insures hurricane exposed areas. If you went and talked to any underwriter at Lloyd’s, or wherever, that’s been doing this for 20 years, he’d say, “Yes, totally agree with all of that.” The problem is nobody has really, until now found a way to put that into practice so that when you are actually quoting a specific home, you can take that into account.
We actually have a patent on using what is called LIDAR data that essentially allows you very quickly—only needing an address—to build a 3D model of the area around a home and know if this home next to the airport and therefore likely to experience higher wind speeds if there is a storm, or is it in a more mature neighborhood that has buildings and trees around that are going to block the wind? We actually use that to drive pricing.
On a place like Martha’s Vineyard or the Cape, it’s not as extreme as Miami, but the same thing, whether it’s a nor’easter or hurricane, kind of where different homes on the cape really have different wind exposure depending on what the environment around them is. We actually take that into account in our pricing and nobody else does.
At the end of the day, what that means, I think, is that we can go out on the Cape and sort of say, “Hey, you know what, we actually like this because we think it’s more protected.” But, only we know that, so maybe [the Fair Plan] or whoever else is charging $2,000 for a home, we can come in and say, “You know what, we’re only going to charge $1,700 for that.” We save the consumer money, give them just as good or better coverage and that works for the consumer.
On the other hand, we are going to have some places on the Cape where an insurer is charging $2,000, and our quote is going to be $2,300. People are going to look at that and go, “Nope, I think I’ll go with the other carrier, It’s good coverage and they’re cheaper. That’s the essence of it. We have very different ways of looking at risk. At the end of the day, that means sometimes we’re cheaper, sometimes we’re more expensive, but we think that coupled with the very quick user experience, agents will just kind of say “Hey, it’s so easy to find out if we’re better than who you got now. You might as well put in your address and find out and if we’re more expensive, well you’ve wasted a grand total of ten seconds of your time. If we are cheaper, then you might save a few hundred bucks or so.”
So, how big is your company now? How many people do you have working for you and where is Swyfft officially headquartered?
The company is headquartered in Morristown, New Jersey. That is also where our commercial MGA is headquartered. Our technology team is out in Seattle, mostly because there’s just a lot more technology folks out there. I think total right now, I think last head count was 27 people, 28 I think. We just made another hire. We’re definitely growing quickly. Like I said, our premium is growing and we’re definitely hiring
And with Swyfft, are you planning just to specialize in homeowners or do you have grand plans to go into other areas?
That is one of the $64,000 questions we debate internally, because we, like I said, we do have this other commercial MGA, and we definitely think that we could bring a very similar kind of experience to commercial packages, so we are considering doing that. The bigger question is … we have some very interesting auto analytics that I think would allow us to write auto profitably, even knowing that that is such a challenging business, especially in Massachusetts. We would love to go into it, but we are not sure if we are going to.
We suspect many reading this will say “Hey, it’s great to underwrite. It’s great to get these quotes so quick, but I got the whole account for this person and how is this company going to be on claims?” So how is Swyfft at handling claims?
I think that is a very fair question. Since we are small, we use a TPA. We manage it, but we do use a TPA, which is North America Risk Services. We have used them on the commercial side for a long time. We know them personally. We like them.
Are we the best at claims out there? I honestly don’t know. I don’t know if USAA’s better than us at handling claims, they might be, but I think we’re very good. I think, most importantly, we pay our claims and I think we pay them very fairly. I also think we pay them fairly quickly. Again, like I said, I don’t want to oversell it because I think it’s a very valid concern. I think, to be honest, agents should … Agents that are worried about that should honestly probably ask people other than me how we handle claims, because I think we do a good job.
I think the major point from the agents point of view would be if an insured comes in to me and there’s an issue with their claim, am I going to be able to get a hold of somebody at your company?
On that note, I think, we’re actually good because we’re small, meaning…things can get escalated svery quickly if there is a real problem with the claim. Relative to bigger companies where it takes you forever to get to the guy that actually can make a decision on something that’s a little complicated, that is not the case here. If need be, stuff can get escalated to myself or Rich within a day. That’s definitely something that I don’t think is a problem. I suspect we’re actually good on that because we’re small.