With about 80% of Americans living paycheck to paycheck, a start-up e-commerce firm like Zebit makes perfect sense from a market-demand standpoint.

The company targets consumers with no access to credit who need short-term cash to make purchases larger than their paychecks can handle. It’s in that business alongside so-called “payday” lenders, but with some eye-opening differences.

Zebit charges no interest or late fees. It doesn’t check customers’ credit histories. It doesn’t take funds out of their bank accounts. And customers can pay off their purchases over six months.

Customers register for a credit line of up to $2,500 they can use to shop on Zebit.com. The site offers access to products and brands from a network of more than 80 popular vendors, including Apple, Black & Decker, Cuisinart, Nintendo, Samsung, and Whirlpool, and from such major wholesale suppliers as Best Buy, D&H, and Ingram Micro. Prominent product categories include electronics, appliances, furniture, beauty, kids/baby needs, sporting goods, tools, and jewelry.

Steve Lapin

The company, which started up operations in 2015, has more than doubled its revenue each year. It recently passed the 300,000-customer mark, expects to top the $100 million sales threshold this year (with a 27% gross profit margin), and, according to CFO Steve Lapin, projects 2020 sales of more than $250 million. To date, it has received about $39 million in venture funding.

But how does its business model — which includes extending free credit to people with poor or minimal credit records — make sense financially?

It is indeed a tricky dance, judging by Lapin’s comments. He recently discussed with CFO the model’s intricacies and how he runs the company’s finances. An edited version of the conversation follows.

Zebit’s website says it earns money the same way other retailers do: buying at wholesale prices and selling at retail prices. Is that the entire revenue model?

Yes, that’s the whole thing. We capture the full gross margin of each product. [The company also operates a drop ship vendor network, eliminating inventory costs.]

There’s a consumer review website called Trustpilot on which 86% of users rate Zebit’s service as “excellent.” But there are plenty of disgruntled folks as well, many of them complaining about having their registrations accepted but then suddenly canceled at the checkout point on the site.

We use big data analytics and predictive analytics to manage risk. But we don’t just use a consumer’s identity, income, and employment [to assess their creditworthiness]. We also look at the consumer’s behavior on the website. We’ve done hundreds of thousands of transactions, and some types of high-risk consumers exhibit similar behaviors.

We use that information to do predictive modeling that underwrites the consumer at the point of registration and also risk-scores that consumer at the point of sale. If customers come to the site with a temporary line of credit and then exhibit those behaviors, we decline them at checkout.

What are some examples of those behaviors?

I don’t want to give away too much, but maybe someone comes on the site and immediately maxes out their utilization. Maybe they use different addresses or different cards to check out. There’s a host of things that feed into the predictive analytics.

That’s not a foolproof process, right? Some people must be denied access who didn’t actually have any malicious intent.

That’s correct. There’s definitely a false positive rate. We’re working toward having some subjectivity that would allow some people we’ve canceled to come back.

Would you say predictive analytics is the main key to making the company work from a financial standpoint?

We have to be very good at underwriting risk. And we have to be very transparent and honest with our consumers about the value proposition we’re offering. We want those who are financially stressed to come back because we’re dangling a carrot vs. using a hammer-and-stick approach.

There are a number of comments on Trustpilot grumbling about Zebit’s high shipping-and-handling costs, and a few noting that products cost more than other retailers charge. Are those ways that you make up for not charging interest on the credit lines?

Shipping and handling is a net for us — we don’t make money on that. The reason we can’t do two-day free shipping like Amazon Prime is volume. As we scale, we’ll be able to lower shipping and handling costs.

Some of the drop shippers we work with use their own shipping accounts, and in those cases we can offer significantly lower-cost shipping. But that means even less volume going through our account.

But from a consumer’s perspective, they might be saying, “They’re not going to check my FICO, and they don’t charge interest, and I’ll pay for that with these higher costs.” Is that fair?

It is fair. However, you’ve also got to think about the alternatives for these consumers. They do shop at Amazon and Best Buy, but to get a power cord or headphones. They can’t make big purchases there, because they aren’t able to pay in full at checkout.

How do your credit-loss statistics compare with those of, say, credit card companies or Amazon?

We’re a deep subprime company, focused on consumers who don’t have access to credit cards. So from a credit-loss perspective, we really need to be compared more to high-APR lenders, the payday lenders who work with people with FICO scores in the 500s.

Our credit losses are significantly better than those of any payday lender out there. With no interest charges, late fees, or non-sufficient funds fees, our value proposition is significantly better.

What we’re really built for is to disrupt rental firms like Rent-A-Center and Aaron Rental Center.

With no late fees, why do customers bother to pay on time at all?

So they don’t lose access to their “ZebitLine.” As soon as a customer misses a payment, we freeze their ability to continue to purchase. Also, by making on-time payments they can increase their ZebitLine up to 5% of their gross income.

How does the business you’re in influence what you do as a CFO day to day?

Because we’re taking credit risk but not charging interest, we have a relatively thin spread on utilizing our cash flow.

We do 40% or more of our business during a six-week period in November and December. So managing our vendor relationships, managing our payables terms, and getting a working capital credit line in place so that we can have 30-day terms with all the vendors regardless, are super critical for making our equity capital work for us and getting the type of internal rate of return on our asset base that we’re looking for.

What do you mean by “making your equity capital work”?

It’s part of what makes this business unique. Say we buy a product from a vendor at the wholesale price of $750, and sell it to a consumer for $1,000. We give the consumer a $1,000 credit line to pay for it.

The consumer has to make a down payment of 20% to 30%; let’s say in this case it’s 25%. We receive $250 cash at the point of sale and ship the product immediately. The remaining $750 is amortized in six equal $125 installments over the next six months.

We owe the vendor in full at day 30, by which time we’ll have received one more, $125 payment from the consumer. Now we’ve received $375 in payments on our $750 purchase of the product, meaning our cost of goods sold is only $375 by the time we have to pay the vendor.

We have an asset-based credit facility that will lend us 90% of our total capital at risk, which will decrease over the six-month customer payment period. Our average capital at risk over that period is $225, and our credit facility covers $203.

That means we only have to put $22 of equity capital to work to make a $1,000 sale.

[Editor’s note: a reader who does an internet search for Zebit may find articles written several years ago, criticizing the business practices of another business by the same name. It was a U.K. business, owned by a company called Global Analytics, that existed for only a four-week test period.

Global Analytics later bought a U.S. company and spun out of it a new entity with a different business model that it also called Zebit. Neither Global Analytics nor any of its former management team is involved in the business.]