Decision Quality Blog

These are my notes on a conversation between Tom Brown and Chris Larsen of Prosper.com. It occurred at the BAI Mavericks in Banking Conference.

Tom

Talk about starting E-loan

Chris

It started with taking out my first mortgage. It pissed me off how I was treated personally. It was motivated by anger. Anger and being poor. It helps blind you to following all that good advice you get. Ignoring the advice of the people you respect. You get an idea, an epiphany, and you only get that once in a venture. Then you spend the next years hacking through it. Any advice you get, you have to say, “no.” Take the financing advice, but not on the core idea.

Tom

Talk about what you did right and wrong?

Chris

First, ignoring advice.

I always partner with a cofounder, so you can help each other ignore all the good advice you’re getting.

We recognized early that finance and tax are different than entertainment. You have to bring together three domains: technology, risk, and regulatory. Then marketing. The good thing is that the tech heads in the valley can’t compete just on technology like they can do in music. This is what makes finance work more slowly on the Internet. But it gives you an advantage if you get it right. The barriers are so high.

Get domain experts, not great “athletes.” You need people that know what they’re doing and get them to work together.

Get the money in as early as you can and as much as you can. So many companies die of starvation. There are all these booms and busts. We were in a boom and now we’ll go into a bust. Teams work best in a trough by the way.

Mistakes? Not firing people who were only four stars (vs. five stars).

You need to iterate very quickly. Stay in that brutal Bataan death march of 4 to 6 week release cycles so you can innovate against your competitors and wear them out. Stay in that for years, and years, and years.

Companies will have a good idea and then they’ll outsource some part of what they do and then they really pay the piper because they can’t keep innovating.

Tom

You came public at the peak of the Internet craziness. How hard was it to run your business when the valuation cratered?

Chris

After the crash, we were worth less than our cap price. It was frustrating and depressing. We lost all the mercenaries, the people who just chase the money. We tried to focus on positives. We were about core values. We were radically pro consumer. If the financial values weren’t there, we went with the missionary values.

You try to focus on the long term.

Tom

As the company makes progress, there were some wild swings. How did you decide to sell?

Chris

It was the right thing to do. Personally, my value occurs early on. I think an entrepreneur gets in the way after a while. We had a great team. You check on the core values. Will they take care of the employees? I was already on to my next thing, Prosper. The time was right. This was 2004. We couldn’t do Prosper before that. ACH was there. Credit scores were transparent. Social networking. Timing was good.

Tom

Tell us about Prosper.

Chris

My wife was telling me about a concept in China called Lun-hui. . . we liked the idea at e-loan about creating an eBay for money. They thought about it. It was all these things coming together. We couldn’t do this at e-loan. We needed a new company to do it. Make it or don’t.

We wound up with a version of Wall Street meets online social networks. We couldn’t do a straight peer to peer. Then we got some really crazy ideas about social networking that probably went too far. So we scaled them back. We’re probably just finding out how it’s different in finance than entertainment.

Tom

You found that large groups turned out to be a negative.

Chris
We tried to learn the lessons about Friendster and MySpace. One the technology couldn’t keep up with the use. They also tried to get too involved with the users. They’re just a utility.

What really works is hands off. Don’t mess with the community. Let any group form. Let the market find its way. If you try to get in and mess with it, you’ll screw it up.

We wanted to build something where anyone could tap their group wherever it was. They can create their own virtual credit union. Where we screwed up was incenting people to go recruit people. Up to about 150 people groups work okay. Once you let groups get big, there are some very opportunistic people on the internet and they spot these things and milk them for whatever they can get. So they would find people who weren’t in a group and weren’t funded . . . in fact they would be prime borrowers but something wasn’t right. They’d pull them into their group but it didn’t work and the default rates were high. Adverse selection.

So we pulled it back.

Tom

How much in and out of group?

Chris

It’s now about 30% in a group. It was 60%.

In finance, unlike other social networks where accumulating a lot of friends is the thing, on Prosper, you’re not a friend unless someone bids on you. If you get friends to put money in the pot, default rates go down about 35%. We’re just now learning how all this works.

There is really cool stuff going on with combining social capital, friends, and scoring data, both for driving financial performance and customer engagement. There is a huge opportunity here for financial markets and for consumers to monetize their own financial data.

Google is working on something like this to help people monetize all their data: data as property. The young generation is much more open about things . . . much less ashamed about sharing data about personal finances.

Tom

Why are the demographics different than I expected? Borrowers are older. Lenders are younger.

Chris

Lending is scarier than borrowing on line. Both borrowers and lenders are consumers. The lending part is the most radical.

We suspect this will work through.

Tom

How are you attracting the customer bases on both sides?

Chris

On the borrower side, you can do all the classic methods. On the lender side, it’s trickier. It’s a different asset class. It’s more viral. There is social value in introducing someone to doing this. We’re still trying to figure this out. We now have something on Facebook. See how smart I am. Show your friends. Things like that.

Tom

You talked about Phase 2 and helping lenders improve their understanding of credit.

Chris

We switched over in October going to default data based on Prosper data. Before, we had to use Experian data. Finally we had $100mm lent, lots of units, 2 years of history, we could start popping up data on how loans like this have performed. That was a pretty big breakout.

We have to be careful about not guiding people too much. If we manage too much, we’re a security. There are all kinds of things we’d like to do on one side that are gated on another side. Again, the cool thing about finance because of the complexity is that all those things move like that.

Tom

Is there an optimum loan size?

Chris

We go from $1000 to $25,000. We go after unsecured business. I don’t see why we couldn’t go in other areas as the market grows. Our average is about $6,800. We’re square in the middle of replacing credit cards.

At some point we could get into a secondary market which would allow for longer terms. We think that the platform can accommodate all kinds of products.

Tom

Any meaningful changes in lenders?

Chris

Early on you had a wild west where so we had all early adopters. We were very hands off. So you saw some funky behavior. A lot of irrational exuberance. People bidding on the picture of the puppy and they hd horrible default rates.

Later, it changed. People pulled back. We can’t step in because we’re just a market. We saw a change in lender behavior. We had a reevaluation so people were looking at 75% financial return and 25% of engagement. That seems about right.

Now we’re seeing really cool stories emerging that we had nothing to do with.

Question

What are your defaults?

Chris

We run about 5 to 6%. We have more to come. We do collections for the first 120 days, and then we sell the bad debt off to bidders.

Last year we had about 30% subprime. Now it’s about 5%. A radical shift in what lenders will fund. We’re trying to get mainstream buyers, just another channel.

Question

We’re sensitized to selling non-insured products through our networks. Have you been sued yet?

Chris

We’ve been very lucky. We always want to let people know the risks are. People get that. You’re giving them the power to make the call. We looked at the insurance products, but we don’t see how we could do that in the US. They’re better off diversifying their portfolios of loans.

Question

What about other entrants into the space?

Chris

Quite the opposite. Because of the complexity and regulatory issues, we’re still the only open market place. We’re the only ones that allow you to make your own bid. We also think there’s a point, unlike e-loan, where we think this will be like eBay: one marketplace because of scale and network effects.

Question

Describe our organization vs. traditional lending institution.

Chris

Very heavy technology. Half our people. We believe in owning all our technology. We want to build everything ourselves. We build on Microsoft platform. We have a huge list of things we want to do and we want to stay on that brutal four to six week release cycle.

The other thing is that we don’t do underwriting. But we do a lot of fraud detection. We work closely with law enforcement and get people arrested. It sends a message in the blogging community.

Regulatory we look differently. We work closely with the regulators. We push the envelope. We work closely with consumer advocates. Regulators look differently at people pushing the edge for bad reasons vs. good reasons.

Question

Financial metrics?

Chris

Biggest is growth of the market and performance of the company. We have to look at both sides: borrowers and lenders. We have to be a referee . . . make sure it’s safe. Like eBay in that way. Keep on eye on things. Did we screw up here? Do we need more data? More transparency?

Question

How do you break down between business and personal?

Chris

All are personal in that they’re all on that person’s credit. For regulatory reasons, they have to be personal. About 30% are using it to fund a business. The rest is credit card replacement.

Question

Are people using Prosper to “live on?” Like an eBay power seller?

Chris

We have three roles, borrower, seller, and group leader. Like a mini-banker. You can make money as a group leader, but we don’t have the formula right yet. But we’re confident that we’ll figure it out.

Question

Are your members predominately US based? International?

Chris

We’re only US and we don’t allow lenders from overseas. We are forming a market in Japan with SoftBank. The intent would be to connect the markets. I don’t think we’ll go to Europe soon. We’ll go to Asia first. We love the idea of a global marketplace.

Tom

What are your thoughts about going public?

Chris

When you’re private, you can focus all your time on the product. I think this next wave of the internet is all about product. Google taught us all that. Heads down and don’t have all the bluster. There are so many ways to fund a company now. Now you need real numbers. It takes awhile to get there.

Question

Talk about your leadership styles.

Chris

The first go around we were a little bit more anger oriented. Very anti-bank. It was very effective. Complaining is different than doing something about it. It can work against you.

E-loan was a channel competing against the bank. Now, we focus on the products.

Early on you really have to baby the venture or it will fall away. You have to force feed it for awhile. Then you have to build the team. There comes a point wher no matter how much effort you put in, you can’t move the ball by yourself. You need to focus on team. Then the domain experts who are 5-star. Then ruthlessly get rid of the non-5 star players.

Tom

Is there a point of maximum discouragement?

Chris

You get tired after two years. Nobody knows who you are. You’re dealing with all the muck. That’s why you need a co-founder. A core team. Someone you can bounce the frustration off of. It needs to be equals.

Tags: ChrisLarsen, e-loan, Prosper, e-bay, Tom Brown, BankStocks, BAI, Mavericks in Banking