Steve Yi: We were fortunate to become profitable within the second month. For that first month, we thought we could never make that math work. We were losing a lot of money in that first month. It’s specific to this business. You just have to go through that period where you’re willing to lose money.
Sramana Mitra: You also have to experiment.
Steve Yi: Yes, exactly. Is the website converting? We have to put traffic to it to see if it converts. What keywords work well? Auto insurance was a particularly hard vertical to do this in because every click that you buy from Google was $20. It was very expensive to get users to our site. We were spending $20 and making $2 or $3 for a long time. Slowly but surely, you get a 5% improvement here, a 10% increase in what the advertisers are willing to pay you, and the math starts to work out.
Then you just keep at it and start to have some breakthroughs where you have a good portfolio of advertisers who are willing to pay you the right amount. You develop a history and relationship with them. You also have history with Google so you’ll be able to acquire their traffic for a fair price. The model starts to really build. Fortunately, that’s what happened. At a high level, we basically focused on what we knew how to do. We brought it into a new industry but one that we knew would work along the same lines.
We took a leap of faith that we can make it work there and execute on a business plan that we knew we can make money with to support ourselves and also start to fund something bigger as we got to know the industry better. It’s hard to know what industry needs. Every industry and every little pocket needs improvement. It’s hard to know exactly what the areas of innovation are. In this case, what the pain points of insurance advertisers were unless you’re in it on a day-to-day basis.
Sramana Mitra: Could you give us an example of one or two things that you did that really made a difference in making that unit economics work?
Steve Yi: That’s the whole thing. There aren’t one or two things. There are dozens of things. The easy thing to point to is, we had a very well-designed conversion funnel that would qualify users. It was probably over designed. Users converted very well. When users came there, it wasn’t like a typical lead generation site that they didn’t want to have anything to do with. It provided a very smooth end user experience.
We also gave advertisers, on the backend, the ability to bid very granularly for all the traffic that was coming to our site. In the past, users gave out information. Before we started our website, no advertiser really had the ability to bid different amounts based on that information. Maybe, pay more for a married couple who had three cars and who are homeowners versus a 25-year-old single male who was renting. Those two things was really giving advertisers control. I’d be lying it those two things were it.
It was the execution of a hundred things that, cumulatively, added up to our ability to be successful in our very first year. That came from having done this before. We had done this for years before at a larger company. We took what we knew, executed on it, and then tried to figure out what was next after that.
Sramana Mitra: How did this ramp? The first year, you did $17 million in revenue.
Steve Yi: Yes.
Sramana Mitra: To get to $17 million, how much advertising did you have to put in?
Steve Yi: We had to put in something shy of that because all the advertising that we do is customer acquisition advertising.
Sramana Mitra: That’s my point. It’s an arbitrage.
Steve Yi: The typical arbitrage margin is between 15% and 30%. We were in the middle of that range.
Sramana Mitra: So 15% to 30% is the margin range that you were working with?
Steve Yi: Exactly. If you’re an SEO and you’re building out sites to get organic traffic, it’s different.