Follow JRNY on LinkedInFollow JRNY on MediumFollow JRNY on Twitter

In this third segment of our interview with Glenn Renwick we talk about breaking the status quo and using data principles and technology to test new products, pricing, and improve transparency. Watch the video below or scroll on down to read through the transcript and pull out some quotes for takeaways.

One of the things you're known for at Progressive is bringing engineering and data principles on board. Can you explain a little bit about what that involved?

Sure, I'll give you a few examples and that might help fill in some things that we've talked about previously. I said that we changed our business model from returning cash to a consumer in the event that their car was damaged, to returning their car.

I won't go into the details of that but realistically that became a queueing theory issue and in many respects (which is an engineering discipline) consumers would take their car to a body shop and, or a panel beater, sorry I gotta use New Zealand language here, and they wouldn't know whether or not there were three or four or five or some number of cars ahead of them and at least from the environment I'm familiar with very few businessmen sort of said “no I can't take it” and they would say “the vehicle will be ready Thursday or Friday”. 

That was very common and statistically we actually have support for that. Consumer would call Thursday, Friday and it wasn't gonna be ready until the following week. OK, things happen maybe. Surprisingly at large scale we could see that that was happening an extraordinary amount. It is not generally possible for insurance companies to sort of get a car repaired cheaper than a competitor. Yeah there may be some arrangements that are there but for the most part we want to make sure everything is done absolutely correctly and the cost is the right cost and you want your panel beaters to be very good and they should get paid well.

So where do we take the friction costs out of the transaction? Now, again I'm using US examples. Well that weekend that we now incurred a rental car coverage so we were adding $75 or there abouts to each physical damage claim. So it was really an issue of well why can't we get the car back on that Thursday and Friday eliminate that friction cost which is not a cost that anyone values, except the insurance company but they're not part of this food chain, consumer and panel beater all we care about so we had to really analyze why was this happening and it was oftentimes a poor distribution of damaged vehicles to potential suppliers that could fix them. So we had a queueing model that was sort of multi que multi server, we needed to re-rationalise that so that damaged vehicles were getting to qualified suppliers but when they had the opportunity to fix them.

So we know just-in-time inventory type of models from whatever the 70s, this leveraged a lot of that thinking and I think that's enough detail for now but it really was a queueing theory just-in-time inventory kind of model. Yet to the consumer it was “Hey, I'm getting my car fixed ,faster”. That was a good objective and an outcome from the consumer the how didn't necessarily matter and we decided that it would be for our suppliers, our body shops panel beaters, it was all about quality the first time. And the number of times a vehicle had to go back for rework that was a product defect and so a lot of the classic models of quality control were implemented into the claims process but on the surface of things it doesn't seem like that's where you would use those tools.

Something else that we took a different approach to was, instead of saying “we'll collect a lot of information then we'll tell you what your price is”; we turned that around and sort of backwardly engineered and optimized the whole set by asking how much can you afford a month? And obviously zero and stupid answers didn't count but if someone said “I can afford $200 a month” just to use a number, then we would backwardly engineer our product so that we could solve for a $200 a month. The ‘but’ is, it had to be a valid and good product that met State requirements and was good for the consumer. But it gave the consumer a little more power, and we talked about trust previously, that this is about me, versus it's about you. Cos up to that point it's “gee I gave you a lot of my information and now you're just gonna tell me a price”; this turned that power around but it required sort of reverse engineering and optimization techniques which are sort of common in systems engineering and so on and so forth.

And one perhaps last example - rating can be complex so I won't go into great detail but it's also least in the United States done where you file a rate revision and you stay with that price until you file another rate revision and change your price. Well ironically when you think in terms of quantum's like that the price is only right, ‘theoretically right’, on any one day of its lifetime, and that would be the midpoint.

So is there a better way of matching inflation or trend with the premium? Not for an individual because they will get the price they got on that day, but if someone came three days later or let's say 30 days later, should the price have gone up just a very small amount to match continuous trend? So it was a way of taking something that had all of its genesis in written rate manuals which had to be almost by definition fixed and transforming the model to a more variable pricing model so it was a continuous model and any significant adjustments to that were only where you had missed the trend by an appreciable amount. 

But it meant much less disruption to the system and much less what we called ‘rate revisions’ which were expensive from a reprogramming point of view. So we bought that dynamic of variability into our design and it actually was a cheaper way for us to do it but it also matched the market trending and there are times that market trending went negative which we could also include.

Those some sound like quite big innovations. I can imagine there must have been some tough parts of implementation. What would you say was the most difficult thing about innovations like that and how did you overcome them?

Well the most difficult is, is the obvious - it’s that status quo is very powerful and I would say probably for me I underestimated the power of status quo. Agents were happy with what existed consumers were happy with what existed - I mean they may complain and want something else but moving to a different model - well that's scary too, employees were comfortable that's their that's how they learnt their contribution to the company and skills and they were comfortable with that. So changing is really hard.

I could comment I don't know why United States is not a metric system but you could imagine the level of change that that represents. Whereas I think I can I can still remember the 10th of July 1967 someone could check that that was DC day we changed a New Zealand and you know a little bit of a issue at the time but eventually it becomes the norm, but that change point can be very stressful for a lot of people. 

So status quo and therefore how do you break status quo? Its leadership, and its leadership by going with us to regulators certainly to employees and being able to find from a marketing perspective a very rational way of relating to the consumer so that they go “oh that makes sense”. 

If it's something you have to explain in a very detailed way you're probably, my opinion, you're probably not gonna get past the first sentence. So for example now (this is a little bit like aged technology by today's standards) where we said you name your price we actually had a what we called a ‘price gun’ which was the old technology we put the price sticker on products and you go through and if you opened up a box in the supermarket you can put the price on top before barcodes and so on and so forth and a lot of consumers could still relate to that at the time we used it so they made it simple it's kind of like well hold it I get to set the dial and do the pricing. Fine, that's good enough for right now even though it was covering a great deal of technology.

So break the status quo, you need leadership, what's in it for me, and it's got to be simple. Einstein said “make things as simple as possible but no simpler”. A lot of power in that - so you can't just simplify things but you can make them as simple as possible, especially for the consumer and employee marketplace.

Was there anything that you wish you had done sooner or better?

Oh as you get old yeah the answer to that's everything, but I think we got the timing about right on a lot of things but that's looking now that's looking in the rear vision mirror. At the time you always feel like you wanted to go faster but I recognized the market sometimes takes some time to gravitate to that. 

I guess I wish we had known the full power of the direct model and how we could incorporate agents into that and why it would ultimately be powerful for them.

But it's awfully hard to go and sell a story that we're going to create a really powerful brand when we didn't know if our brand would be powerful. 

No I think in general the one thing that I might have been a little slow on, I think that's terrible I have I know slow on everything, but as a company we tried our own development of homeowners. So we were primarily, as I said earlier an auto insurer. One of the things that consumers liked to do was bundle their auto and their home together and my only real rational reason for wanting to be in the home insurance marketplace in the United States because I don't think it's a great insurance product, it’s not statistically is predictable as auto insurance - was to have a bigger addressable market place. If about half of the marketplace wanted to bundle we needed to start finding ways to address that marketplace.

A natural inclination was to develop things ourselves. And that was hugely diversionary on a lot of our technology staff and so on and so forth. I wish I had done what we have eventually done and that is go out and find the closest match to Progressive’s culture in a smallish homeowners company. 

And we actually found that and someone that had developed literally their company with, this is not an exaggeration, their culture they had used Progressive as the model including some very specific references in all of their early materials and they had developed a very nice homeowner's company - a company that we respected in every way from their early development they were much much smaller but it just felt like you were combining with Progressive’s. So the leadership of that company and I basically just did that, and we did it somewhat slowly with a little acquisition at a little you know part of the equity at first and then moved into it seeing how that opened up the marketplace for us.

Arguably I wish I'd done that a little faster, earlier. 

To use your words, looking in the rear vision mirror, are you able to single out the most impressive innovation that you saw at your time and also one that was perhaps a dud or it just didn't work?

There actually are quite a few innovations but I'm gonna choose the first one because it seems so commonplace today but it wasn't commonplace then. We were motivated by a change in the laws in California actually a proposition called proposition 103 I won't go into the details, and it was very negative to insurance companies specifically auto insurance companies in California. Coming from an insurance executive perspective. Now a man by the name of Ralph Nader didn't sort of see it that way. Ralph and I are now very good friends and he is a huge advocate of what Progressive did to ultimately respond to this but it took that kind of a awkward catalyst like “you're trying to put us out of business”.

We reverse engineered everybody's rates and put them in the public domain as 'here’s the top ten competitive rates for auto insurance in California'. And ultimately we extended that nationwide.

And while people were very interested in the rates, they were more interested in transparency. It's like “why are you as an industry so secretive about how you get my premium?”. And we sort of flipped that around and said “you know, look there are issues” - in fact this whole proposition passed because the notion was insurance companies aren't competitive. That was from the insurance industry's perspective an incorrect thesis - there were 300 competitors - how could we not be competitive? It was one of the most competitive marketplaces. From the consumerist, Ralph Nader in this case, he said well the consumers don't have access to that. So even if you're right, it's not competitive. And he and Harvey Rosenfeld were you know very dogmatic about that and ultimately that proposition passed because they were able to build in well consumers will get money back. So people tend to vote for things when they'll get money back.

And we took away from that proposition “boy we've got to change because we're a part of that industry”. We need to position the industry over here and ourselves over here. So and I happen to be in California at the time so we we did what seems to come naturally to our thinking it's like

“well if that's the problem, we’ll solve the problem; you say it’s not competitive, we'll make sure it is ultra competitive - so we'll publish all the rates”. Some of the competitors in California were not very happy with me, I will point that out, but we started to transform Progressive much more to a consumer company. And the fact that our competitors were more than not happy with me (I won't go into details) I was like “oh we are separating from the industry”. 

So it actually created this separation and that transparency has gone through everything that we do including being very transparent with our financials once a month to Wall Street. And I care more about Shareholders than the analysts, but we always took the attitude of look if we can make it transparent, we'll do so and if we can't we'll say we can't.

And were there any innovations that you perhaps started and they were a no-go?

One that I still haven't given up on, but I'm no longer in a position to do anything about it - we sort of used Archimedes principle of buoyancy. Now specifically in some of our Gulf state coasts and you've seen it on television with a V Houston Louisiana or Florida we get really really big hurricanes and oftentimes that is an incredibly damaging situation. So I wanted this, effectively a big waterproof tarpaulin we'll call it for now, that you drive on to - and there are specific places for your wheels to go, and then it wraps around the car in your garage because a lot of people had two cars and if there was a forced evacuation they only took one. So you would wrap this wrapper up around the doors and then sort of just anchor the car on the door or some other rafter of the garage whatever. And as the water filled up the car would actually float - it wouldn't need to be totally watertight all the way up it would just actually float.

And I have plenty of pictures where we have demonstrated that that actually works just fine. But it became very hard to commercialize it for the consumer and get the product, new material which was possible. And the idea was that frankly if we were going to lose let's say five thousand cars to flood damage and we saved them then it certainly would be worthwhile to produce the materials necessary and not to mention the fact that the brand value of someone having their car when they got back was positive.

Let's just say that doesn't happen today and so put that down in the category of good idea just didn't quite get there.

And I've got another one that sort of would protect the cars in the event of hail which is really bad for sheet metal and so on and so forth. And I hope somewhere in Progressive, I just got a couple of notes from some folks today but not on this, I I hope somewhere that that idea hasn't been lost because loss prevention is so much better than paying higher premiums. 

And I don't think loss prevention was the primary motivation for a lot of insurance companies - we just increased the premium to match whatever our costs were - and at some point if you can actually reduce those costs in very real and meaningful ways, like taking out rental car coverage or something, by getting claims sold faster, then you have a huge competitive advantage.

Interesting! There's a few themes that I've noticed cropping up over a couple of segments and I think communicating ‘the why’ has popped up in the context of culture, change and in trying to enforce an innovation, so that's really interesting, thank you. 

Next up in the following fourth segment we'll be playing a game of true or false where I'll be pressing Glenn for his brutally honest opinions on some industry cliches

Sharing is caring
FacebookLinkedInTwitter

What's next?

Ready to
start the JRNY?