From Sequoia to Democratizing Consumer Finance For the Underserved

Interview with Omni Prime CEO Dan Hu

Dan Hu, CEO of Omni Prime

Dan Hu has seen it all. From leading multiple successful investments for Sequoia China during the 2010–2014 tech boom to helping found ZhenFund (China’s top seed fund), Dan has always had an eye for finding the next big thing. That penchant for uncovering opportunities has led him to found Omni Prime, a mobile consumer finance service company that provides payday loans to low income entrants with a complex, app based risk evaluation system.

In China, hundreds of millions of consumers still do not have formal credit ratings. In particular, students, farmers, and blue collar workers lack credit scores, not because they are subprime debtors, but rather because they haven’t previously borrowed money or needed to make any repayments. That makes it impossible for traditional financial institutions to make assessments, and unlocks an interesting opportunity for internet companies that are more risk taking, that have lower operating costs and an ability to iterate more quickly on risk management.

On this Q&A, we explore this under-served segment to better understand the market opportunity, competitive dynamics, and commercial applications. Why are blue collars under-served by credit cards? How can 3rd party offerings compete against the consumer finance solutions provided by eCommerce giants (Alibaba and JD)? How do consumer finance players find the right entry point in a fragmented market where 80% of commerce still happens offline. What are the primary consumer scores and how do they differentiate (e.g. Omni Prime’s Octopus Credit vs. Alibaba’s Sesame Credit)? How will this market evolve in the near and medium term?

Listen to the podcast on SoundCloud (here)

Audio transcribed by Carina Zhang


Sequoia to ZhenFund

Adam: Dan, you’ve had a perfect career path to date. I mean, top MBA, top consultancy, one of the best VCs in the world, and now you’ve founded your own successful company, where do we even begin? And, I know many of us want to hear about Omni Prime and consumer financing in China. But I also noticed that you helped start ZhenFund, which is a very successful seed stage VC in China. So why don’t we start there, tell me more about your role in starting that venture.

Dan: Sure. So prior to Omni Prime, I worked at Sequoia Capital in China for three years as a venture capitalist. My focus was early stage TMT investments, especially on the internet side. So I was quite lucky because I worked from 2011–2014, which was a golden age of the Chinese internet industry. So a couple of my investments, including PPDai in the finance are, Jumei.com in the e-commerce area, and also a couple social media/social network companies went on to become quite successful.

When I was at Sequoia, the first priority for me was to spot good deals, very early stage deals, to make friends with angel investors, and try to source deals from angel investors’ existing portfolios. So I got familiar with Bob Xu, who is a very prominent angel investor in China and also made an early investment in Jumei.com. At that time, 6 years ago, Bob was doing angel investments on his own, but I pitched him this idea of starting a joint venture between himself and also Sequoia to make angel investing institutionalized. He was very happy about this idea.

So ZhenFund was started in 2011 as a half-and-half joint venture, 50% Sequoia and 50% Bob Xu. The goal and slogan for ZhenFund was to be the greatest small fund in the world. That means they wanted to invest in great companies in their very early stages, to be the first investment shareholders of those companies. We were able to attract the current CEO of ZhenFund, Anna Fang, to manage the fund, and the return of ZhenFund has been phenomenal to date.

That was my experience as an investor. My 3 years as an investor were very fruitful, not only in terms of the financial return that we generated for both ZhenFund and Sequoia, but also the experience to work with top founders and to help build (at least some financing and idea generation) for those great internet companies in China.


China’s Consumer Financing Market

Adam: I think we could probably spend a couple of episodes just talking about your experience as a venture capitalist in China, but why don’t we make a pivot to consumer finance and Omni Prime. We understand that the consumer financing market here is very hot, and in some ways quite different from the U.S. and other markets. So why don’t we start with your views on the market itself, and how you determined this opportunity.

Dan: So the consumer finance market in China is still very young. If we look at the consumer financing penetration in terms of a percentage of the total social consumption, it is a little bit above 10%. The number for the U.S. is 16–17%. So we have a long way to go. A lot of Chinese consumers are underleveraged, which is an interesting opportunity for us to explore. In China the credit card penetration is also low, and credit cards only serve white collars workers.

So if we look at the consumer finance market as a whole, we can divide the customers into two major segments: the scored segment and the unscored segment. People in the scored segment usually have credit reports from the central bureau of China, the PBOC, to tell underwriters, banks and financial agencies their personal credit history. That way, financial institutions can make an underwriting decision and determine the right price of the private product that they want to provide to their consumers. But if we look at the scored segment, which is still not that big in China, they are all white collars, including some small/medium size business owners.

If we look at the blue collars — the students and the farmers — they all fall into this what we call the unscored segment. Basically, that means if you go check their credit scores from the central bureau, you can find nothing. So that’s the opportunity for us start-ups. Usually people think that these are sub-prime customers, but rather they just don’t have a score because they haven’t borrowed money from any institutions, and so they don’t have repayment behavior. That’s impossible for a traditional financial institution to judge. As a result, this becomes an interesting opportunity for internet companies, like ourselves, because we have low operating cost, top-notch technology, and can iterate more quickly in terms of risk management. So that’s why I decided to quit Sequoia and start Omni Prime to target the young blue collars in China.

Adam: Alright, Dan, the overall picture makes a lot of sense. But let’s drill a bit further into your user base and use cases. Let’s just look at blue collar workers for now. Could you talk a bit more about the value proposition for them, what kind of products the users buy, and what is the lifetime value of these users?

Dan: Right. If we look at a typical young blue collar, his leverage is quite low. Yet only 20% of the blue collars are served by either a credit or a personal loan by a bank. That means 80% of that population are unserved with low leverage, so its relatively safe for us to add some leverage to them. Second of all, there’re big consumers: they like to consume, they like to buy iPhones and in some cases undergo cosmetic surgeries. If we can find a shopping scenario that we can start with, then it’s relatively easy for us to convert them into higher purchase solution users. So this is how we sponsor and convert the users.

In terms of their lifetime value… their ability to make money will also increase with their age, with multiple stages and life events. Today they want to buy a phone, tomorrow they might get married, and a year later they are doing home decoration. We can engage the customers in multiple scenarios.

Adam: The value proposition overall for blue collar workers makes sense. But in terms of the actual process, can you tell us a bit more about where your users access Omni Prime. Can you describe how the process works?


How Does Omni Prime Work?

Dan: First of all, we only work with offline retailers. The reason being that online retail is very much consolidated in China. Alibaba, JD, and a few other big online e-commerce companies add up together to 99% of the market share. Because the internet giants all have their consumer finance apps, it’s impossible for a third party consumer finance solution to exist in the online world. But 80% of commerce still happens offline. For example, of nearly 400 million mobile headsets sold in China, 300 million are sold offline. This online-offline ratio is 1–4, and this ratio has been stable since last year. So the majority of people still like to buy stuff offline, especially for our consumers, blue collars.

If you think about the price of a mobile phone, it’s almost a month of their salary. If it’s an iPhone, that’s two months salary. So if they want to buy a cheap car, they want to see the product and have an offline spot for maintenance. That’s why we choose to work with offline retailers.

The great thing about offline retails is that the market is very much fragmented. The top four retailers in terms of mobile phone are GOME, Suning, Dixintong, and Leyu. They add up together to less than 5% of the market. So the market is very fragmented. That makes sense for us to exist, because we are dealing with small chain stores and single stores. It’s impossible for them to do their own consumer finance.

Adam: Got you, Dan. In that case, how do you go about partnering with all the retailers. A fragmented market signifies opportunity for you guys, but at the same time it makes your business hard to scale, so how do you address that?

Dan: We have a nation-wide BD team that cover 21s out of 31 provinces in China. And we cover 210 cities and 47,000 points of sales. So our BD team’s job is to go there, talk to the shop owners, pitch our products to make sure that they work with us. So we do a training session for the sales person in the shop, and we will also install a merchant app on the sales person’s mobile phone. So that when a customer walks into a store, the sales person of the store can open the app on his phone to generate a QR code to ask a potential customer to scan that QR code. By scanning that code, the customer will have our customer app on their mobile phone, and then the sales person will promote our high purchase plan to our customer by looking at our products, our installment solution, our interest rates, monthly payment plans, and these can be selected on the phone.

If the customer is interested, he will fill out his personal information and application in our app. Usually it takes 5 minutes. After the customer hits submit on his phone, then all the information will be uploaded to our cloud-based underwriting center. It takes us around 2.8 minutes on average to do the underwriting. If the answer is approval, then we will wire money to the merchant, and the customer can walk away with a new phone. As an incentive system, we will immediately send a Chinese red packet to the sales person who promoted this customer to us.


Evaluating Credit Risk

Adam: Could you tell us more about how you’re able to evaluate a customer’s credit risk, especially when there may be lack of credit history. Are you also pulling in their social information on Wechat or Weibo, etc?

Dan: Because we are doing underwriting on site, we face challenges. We can’t expect the customer walking into the shop with his proof of job or proof of income or social security card. It is impossible, because our customers are random customers, like potential buyers of phones rather than loan applicants. So we only ask for 20 questions (20 variables). We ask the customers to take pictures of his ID card and bank card and take a selfie. A typical blue collar usually carries with them these pieces of information. We also ask customers to select one to three emergency contact person from his address book on his phone. We also ask him about his job information and home address.

So it comes down to 20 primary variables. We cannot do all the calculation with 20 variables, but we have the technology that helps us expand these 20 variables into 10,000 variables. We have online crawlers nd third-party data service providers, and we use social network analysis and user behavior tracking analysis. For example, by registering with our app, our customers provide us with their mobile phone numbers. Even if he only gives us his mobile phone number, we can do a lot of things with his number. For example, we can check if this number has already been used in our system, if this number is on the blacklist, if this number appears in another user’s address book who’s on our blacklist. By looking at this number, we can also tell how long he has been using this number by looking into third-party data. With this number, we can know his Wechat or Alipay account, and also his data on Wechat, including his city of origin he marked on Wechat. All the information is from the phone number only. We call this process variable expansion.

Adam: Got it. And can you tell us more about Octopus, your smart credit rating system?

Dan: Octopus as our smart credit rating system is comprised of several things. First is the data bounce, the variable expansion machine. Second is the spider that helps us scrape the relevant data from broader internet such as Baidu Forum and internet registering behavior. The third involves decision engines.

We put a lot of models and rule-based decision points into these engines so that the expanded variable can be processed through those decision points, and the decision points will come up with a decision — approval, reject, collection, etc. In my company, the decision is not made by humans because we have 10,000 applications per day, so we have to reply on the models and rules are designed by humans. We utilized both traditional logistic regression and machine learning. Our risk management team is more focused on designing the models and rules for the engine.

Adam: In terms of this credit rating system again, you guys use Octopus, and in the western world we also heard of a so-called sesame credit from Alibaba. So just more broadly speaking, when you take a step back and look at more credit rating systems that are available in China, how would you compare and contrast Octopus against the others. How do they relate to each other?

Dan: Sesame score is very powerful for Alibaba members because they have a lot of data — shopping behavior, delivery address, Alipay transaction information, etc. By looking at peoples’ transaction behavior, Alibaba can know a lot about a person and his or her user category. But actual credit scores are less relevant to the transaction itself and more relevant to repayment behavior.

With the sesame score, Alibaba can do a very good job in cross selling of products. However, for credit evaluation, especially for the blue collars, we don’t think it’s that useful. Many blue collar customers of ours do have sesame score, and we ran the sesame score in our system to compare the relevance to the probability of default of our customers, and the relevance is very low. Octopus is very different. The single purpose of Octopus is to find and quantify fault risk and credit risk in an offline shopping scenario for blue collars. That’s why our decision is much more accurate compared to (the ones based on sesame score) in a blue collar shopping scenario.


Market Growth Potential

Adam: Since you have a very clear value exchange between the users and the sellers involved, can you tell us more about where do you think this industry will go or perhaps where your business will be in a couple of years’ time. Currently, there are still a lot of blue collars from the countryside, but as China continues to develop and people become wealthier while shopping switches to online, how will Omni Prime be positioned in the long term?

Dan: First, we think online e-commerce is almost saturated, so the growth rate of the e-commerce, especially for big ticket item (for example, appliances, mobile phones), has stopped. There are three hundred million headsets sold offline every year. Many other products get sold offline, such including plastic surgery and e-motor cycles… close to 100% of them get sold offline. We tap into an endless flow of traffic of blue collar customers, so we don’t really worry about the drainage of offline traffic.

Now we have served 2 million customers, and by working with them and understanding their repayment behavior, we have gradually built a credit score system for them. For a user whose repayment behavior is very good and whose credit score is really high, we already provide these customer a line of credit. It is very much like the idea of a credit card. If the customer has a better behavior, we will give him or her a higher line-up credit, like 20,000RMB, so the customer can draw the money whenever he wants. Once he draws the money, he pays for the interest. It’s a virtual credit card for the blue collars.

For some old customers, we already gave them a much higher line-up credit than before, and we have observed their borrowing and repayment behavior for over 20 months. Now we are confident to calculate the lifetime value of a customer. Surprisingly, a lifetime value of a blue collar customer to us is $500. It is as high as a white collar credit card holder who is served by China Merchants Bank, one of the best credit card providers in China. China Merchants bank has 35 million active cardholders, and we are still smaller than them. But we get 250 million blue collar customers, and our penetration is only 1% in the market. We are very confident that we will be able to acquire 35 million blue collar customers in a couple of years, and then our enterprise value will be as high as CMB credit card center. We will be a quite relevant company in the personal finance industry.

Adam: Those are incredible statistics for 1% market penetration. Could you tell us more about the retention of your customers and increasing that number, and secondly when it comes to the dynamics of this attractive market, who will be your primary competitors. Is it going to be the offline retailers, like Suning and GOME, or the smartphone sellers, like Xiaomi?

Dan: A couple of things. Among all the customers, we think 60% of them are very trustworthy, and we provide them with higher line-up credit. Among that 60%, over 40% of customers will use that line-up credit in six months. The overall conversion to the first loan order to the second order is ¼, which is already higher than a credit card company. Of all active credit cardholders, only 20% will become revolving credit users, which means they are profit-making customers for the bank. The people like me are loss-making customers, because I pay money back on time before the grave period when I borrow money from the credit center, so the bank has to subsidize my interest. The percentage of revolvers among our customers is already 25%, and we are quite confident that we can further increase that percentage to 30%, even 35%.

In terms of competition, we do have competitors on the cost items. There are four major cost items on the business. One is acquisition cost — we need to pay the shop owners and salespeople to promote our product. That’s a very transparent commission and fully competitive. This is a big chunk of cost. The second one is bad debt loss, which is related to the company’s risk management program. If the risk management gets better, you can control and further drive down the bad debt ratio. In our industry, 1% lower bad debt loss means 1% increase in ROA, which translates into 10% return on equity and that’s substantial. The third thing is the operating cost that involves underwriting, customer service, and collection. If you can use more machines and fewer men in underwriting and customer service, you can greatly drive down the operating cost. Currently, our entire operating cost is computer-based and automated. There’s very little human intervention in the underwriting. But we are still using human operators on customer service, and we are working on driving down this cost. The last one is cost of capital, which is a pretty fair competition for each player. Whoever can construct a better cost structure can win the game, because every penny we save on the cost structure, we can pay back to the customers, like lowering the interest rate, giving cash back, so that the customers can benefit and will choose whoever can bring them the best product.


Omni Prime’s Path Forward

Adam: It sounds like a sophisticated business to run with a relatively high barrier of entry. I am curious of your long-term plan, are you going to scale this business and eventually go public, or do you think about acquisition by a large strategic partner, like a large chain operator, that can directly benefit from your technology?

Dan: We are actually looking at both options. We see several fintech companies got listed in the U.S. market, but we are still understanding this market, because as a fintech company, there are always different ways of valuation. It can be based on PB or something else. To be acquired by a chain store is another option. One important thing is that although we are a tech-driven consumer-facing company, so a cheap source of capital is very important to us. In our future fundraising, it is my first priority to attract some institutional investors who have a finance background, such as banks, trust companies, or insurance companies.

Adam: Lastly, any views on P2P or any other aspects of consumer financing in China or globally?

Dan: P2P is very hot in China. The development of internet finance in China is 10x, if not 100x, compared to that in the western world. That’s because in the western world, traditional finance is already well-developed, and there is not that much room for business innovation. If we look at investment opportunities in the finance sector, it is more about technology innovation rather than business model innovation. But in China, the general finance, especially personal finance and consumer credit, is underdeveloped, so we see a lot of companies innovating here or there, in both technology and business model. That’s why there are already 1,000 P2P companies in China!

The government has realized that it is dangerous to leave these companies unregulated because these companies can take deposits for the public. If they don’t behave well, they could lose the money and cause social instability. The regulation is tightening up since last year, and the number of P2P players will decrease this year and next year. The top players will still exist and thrive, and the small players will be wiped out by the market.

Adam: Clearly there’s a lot of momentum in the Chinese market, which is as usual, very competitive, and often influenced by the government. Thanks for your time Dan!