Alex Rampell

Supported by Greenhaven Road Capital, finding value off the beaten path.

Alex Rampell started TrialPay, a company that advertised to consumers at the point of purchase. Say you visited Fandango for information (movie showtimes), not consumption (ticket purchase). But you see an ad where Netflix offers a free ticket if you sign up. Well I was gonna sign up anyway the thinking goes. Click, sign-up, claim ticket. In the middle of Netflix, Fandango, and you was Rampell and TrialPay. The company was acquired by Visa in 2015.

Rampell is a general partner at a16z and we’ll look at some notes and quotes from his talks on YouTube.

Listen to this post as a podcast: iTunes, Overcast, or Soundcloud.

Homes. “In the future, you will buy your house from, or sell your house to, a company.”

Real estate is a weird business. There are two million real estate agents in the United States. The median number of homes sold is two, the modal number of homes sold is zero, the top 10% of agents sell 7 homes a year, and the top 1% sell 22 homes a year. Total commissions to all agents were around 100B dollars.

Sellers and buyers are connected by agents, except for for-sale-by-owner, about 7% of sales. Spencer Rascoff, sequentially Zillow’s CMO, CFO, and CEO said Zillow sold ads, not houses. That allowed them to have the P/L of an internet company and 1B in revenue and Rascoff joked that Zillow was the park bench for the internet.


New Zillow CEO Rich Barton wants to buy and sell like Rampell recommends. Why now? Expectations and capital. “Consumers are accustomed to everything on-demand…Consumers want to push a button on their phone and take an action,” said Rampell. And this new Zillow couldn’t have been built in 2004/2005. Zillow had to be a website, then an app, and then full-stack.

The company has adapted well. Rascoff said after seeing an app store demo in 2011, “We dropped dot-com from our name and did all the little things. It was all about mobile.”

As data became more important they recruited more people. Rascoff recalled, “We called the Expedia data science leader and he said, ‘I don’t know anything about real estate,’ we said this isn’t about real estate, this is a math problem.”

Websites, apps, and data were all incremental improvements on the same model. But what if people are ready for a new model?  When asked Why do we always do it this way? There’s got to be a better answer than We’ve always done it that way.

Tricia Wang saw how phone usage changed. Daryl Morey saw how basketball changed. Barton and Rampell think real estate will change.

Consider the job to be done. Sellers higher agents for speed and convenience, not price. And they deliver. Agents also have expertise, experience, and a mental database of when staging a home delivers value, when to list a property, and how to compute comparables.

Only for some of those things computers do the work much much better. This is a point hammered by machine learning experts, let machines do what they do well and let humans do what humans do well.

Rampell said, “Individual agents can look at the comps that have sold, but they don’t have a data science team. When should I list my house? November? January? March? Computers do a fundamentally better job answering these questions than people.”

Rascoff’s book, Zillow Talk, is stuff with this information. Location isn’t everything. Spring listings aren’t the same nationwide. Gentrification is not random. And my favorite chapter title, “It’s the worst house for a reason.”

The jobs people hire for don’t change much because they’re still people. We’ve been hiring for food, safety, companionship, entertainment, fulfillment, and signaling forever.  So startups should look wherever people are already doing whatever. “Rich parents have always helped their kids buy their homes,” said Rampell. It’s stuff that “happens on and off but there’s never been a marketplace for this.”

Rampell wants to fund companies that structure currently unstructured behavior.

His colleague, Chris Dixon blogged this in 2013. Ride-sharing began with Moes safely getting Hoes, where they need to goes. Jan Chipchase advises to, “find people who are already in extreme situations – people whose situations or context pushes them to make the most of what is currently available regardless of existing social or legal norms. Call it innovation by necessity.”

FinTech “If we don’t need branches or in-person communication (thanks to the internet) what type of businesses can we build because of that?”

As Tom Goodwin noted, we only call it FinTech because it’s new. Eventually, we just call it the thing. Electric light becomes light. Online dating becomes dating. Rampell said, “In twenty years you would think about it (FinTech) as your bank, your insurance, your retirement account, you wouldn’t necessarily think of it as your FinTech company.”

Yet we have banks on main streets with clean windows and nice lobbies. Why? Stability was the signal for safety. But, “The physical condition (is) a little bit more of an anachronism than a building block for banking success.”

Banking isn’t about money. Banking is about trust.

Our on-demand, tap-tap-do-it-app, adaptations are filtering to banking too. When I worked in minor league baseball there was a two-week stretch where because of the home games, weekends, and team commitments I couldn’t cash my paychecks during regular banking hours. So I used the night deposit via the ATM. Fast forward almost twenty years and I last used a bank for their notary services and while buying my most recent house didn’t enter a bank at all.

If the branch doesn’t matter what else is changing? User Interface. Andy Rachleff said about Wealthfront “millennials pay us not to talk to them.” Rampell said, “My parents like going to the banking website, it’s a feature for them but it’s a bug for people under a certain age.”

What else? The data. Rampell talks about SoFi’s focus on HENRYs. These High Earning, Not Rich Yet consumers are filtered by degree – oooh mechanical engineering from Cal Tech – and offered attractive terms. Instead of bricks and tellers, SoFi uses reams and programmers.

SoFi wants only the best customers. Art majors default more than STEM majors. But credit card debt may not matter much. With more data, a company can pick off customers whose default risk is lower.

Then sell them other things. If you find a high-income doctor who pays back medical school loans on time it’s likely she’ll also pay back home loans on time and car loans on time and so on down the line. Not only that, SoFi has the chance to be the default choice thanks to a good UI (the new main street) and low rates (because only HENRYs are in the pool).

New data – like college major and university – isn’t limited to the United States. Some of the most interesting work is being done in Africa or the Caribbean and using mobile phones.

In one example, only 34% of adults had bank accounts. If banking is about trust how do you trust someone with “a thin file”? But 89% of adults had a mobile phone. That’s sort of a file. Right?

One study found that borrowers with smaller networks (infrequent and fewer calls to homogenous numbers) were more likely to default.  Consistent SMS use signals repayment. Another study looked at how frequently customers who’d been denied loans “topped off” their prepay mobile number. If someone topped off often they were more likely to repay than those who’d been approved using traditional credit checks.

It used to be that if you built a better mousetrap the world would beat a path to your door. Now businesses face the TiVo Problem.

Think of it as a race. Can innovators get to the market before incumbents get to the innovation? About FinTech he said, “In the US the big ones are Betterment and Wealthfront who were the innovators. The challenge it turns out that Vanguard and Fidelity and Schwab have now replicated what they’re doing and has a massive distribution.”

Rampell is the co-founder of Affirm, an organization that has avoided the TiVo problem. When asked about barriers to entry, CEO Max Levchin said about big banks, “Those guys cannot enter our business because they’re addicted to the (late fee, interest rate, and hidden fees) income.”

There’s an aphorism that something can be good, fast, or cheap but only two. Well, startups need to ignore that. Ideally, a startup will acquire good customers, in a fast manner, for a cheap cost.

Which brings us to the next part.

Channels & CAC

Customers hire businesses to do jobs. Businesses deliver solutions to customers.

How does a business get hired? Their customers get acquired. Hopefully cheaply.  Open Table, said fellow a16zer Jeff Jordan, had a kind of negative CAC. The patrons requested the restaurants to offer online booking.

Because of the TiVo problem startups must grow cheaply, “because it’s too complicated or too expensive to compete with well-capitalized companies.” Paid acquisition, said Rob Fitzpatrick, “is mind shattering expensive. Geico spends a billion dollars a year in advertising, how can anyone compete with that?

Startups can compete because strengths are weaknesses. Where does Geico advertise? Television. Who watched television? Everybody. 

What if instead of selling insurance to everyone a company sold insurance to only the best drivers? The healthiest people? What if a company only loaned money to people who were unlikely to default?

That’s what SoFi did. That’s who HENRYs are.

A business needs to have a plan as to how they can find their HENRYs. For SoFi, Rampell said, “Think of this from the psychology of the borrower, how come I’m paying the same rate of the person who’s going to default. Then, positive selection vs negative selection.” 

Without a good answer to how a startup is going to acquire customers, Rampell suggests buying Google or Facebook stock, where many of these companies advertise.

One company doing a good job is TransferWise. “At the end of a transaction, TransferWise will probably tell you how much every bank would have charged you for the transaction. Why do they do that? The transaction already happened. They give you that reinforcement and then they say, tell your friends.”

A delightful example was shared in Season 8, Episode 16 of the Under the Influence Podcast.

Jerry Murrell used the WOM test to see if his burger business would work. The Zestimate was born after Bill Gurley challenged the Zillow team to reduce their ad spend. Part of the reason Beanie Babies went viral was because of their size.  “(Founder Ty) Warner thought that the play value of a stuffed animal with mostly beans and a little stuffing would appeal to kids, and he also thought that a stuffed animal that could be slipped into a backpack would spread through word of mouth far more quickly than one that required at-home play dates.”


Thanks for reading.

4 thoughts on “Alex Rampell”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.