This post is an excerpt from Lean Into Lending, Vol. 2: Working with New Capital, the second volume of the ebook we wrote and published in 2022. Looking back, it seems we were thinking in the right direction—these ideas have only grown more relevant since it was published. We’ll post an excerpt from the third volume soon, but in the meantime you can read the excerpt from Vol. 1: Rise of the Borrower, or check out the whole ebook, free of charge.
Proof Convinces Investors
When investors put their money behind a company, they expect visibility into every aspect of their investment—from the technology powering the operation to the products offered and how they’re performing. But a long, hard look at a dumpster fire isn’t going to convince anyone to invest. Whether they’re a third party or an executive within your own company deciding how to best allocate resources, investors are drawn to programs that have both proven results and a potential for growth.
If something has a solid track record of growth and success, you can reasonably expect more success as you scale. If a product or program has no history, then you have no indication of how consumers will respond. Before charging ahead with a new product or business model, lenders (and any business, for that matter) should develop a proof of concept: a successful, small-scale version of the program that can then be expanded.
Why is a proof of concept important? Without a strong proof of concept, your idea may end up like CNN+ or Quibi. And if you’re unfamiliar with those streaming services, that’s a testament to how swiftly and magnificently they failed. Even an idea with record-level funding can crash and burn without proper planning and testing.
On the other hand, a well-designed proof of concept can pave the way for major success. It did for LoanPro. A little over a decade ago, when we were lenders ourselves and decided to develop the software we saw lacking in the marketplace, we didn’t start from day one trying to build the features and tools we offer today. We began with a much more narrow scope, focusing on servicing and collections tools for only a few types of credit product, and after we saw success with that initial product, we had a source of revenue that supported us as we built the modern LoanPro, a goal which would’ve been impossible without that successful proof of concept.
Fail Fast to Win Big
So how do you develop a proof of concept? Perhaps your company has been hearing a lot about new financing models, and you want to experiment. Rather than throwing all your resources into the first thing that comes to mind, research a few different options and develop small-scale programs for each. Issue a few accounts, keeping the total cost low enough that it won’t hurt if you lose it—the whole point here is to see how things work, so consider anything you lose an R&D cost. From there, you can see which models offer the highest and most consistent return on your investment.
And at the same time that you look for what works, you also need to keep an eye out for what fails.
Remember, failure isn’t a regrettable accident, but an integral part of the development process. When you’re running small-scale, controlled experiments, failures show you what to improve or what to scrap. If all you see are ‘successes’, that may mean you’re not experimenting with anything new—you’re playing it too safe and leaving exciting, innovative ideas untested. Trying those different models and methods will help you fail fast and small, paving the way for sound, confident investing once you’ve developed a sound prototype. As you issue these experimental accounts, see how consumers respond, and record your data. Do people apply for them? More importantly, do they pay you back? You can further experiment within the same type of accounts, trying different collections strategies to see what draws the greatest returns.
If you’re thinking that this sounds like an expensive fantasy, it might be time to upgrade your software. A lender could never have experimented like this just ten or twenty years ago. Testing different products might have required entirely different software suites, calculations, and even personnel to service them. But with modern, configuration-first software, you can create multiple products with different lifecycles and payment plans. Once the accounts are issued, your software should be able to track how they perform.
Good tech and software empower you to innovate. It puts the tools in your hands to refine what works and discard what doesn’t. Your company can become a sort of lending laboratory, constantly studying and experimenting with accounts as you continue to develop new products and ideas.
A Scientific Approach
We should stress that when we say ‘experiment’, we mean it. The biggest mistake you can make when developing a proof of concept is forgetting the scientific method. You could create a hundred or so accounts with your new model. When you turn a profit, you might be tempted to call it a success. But you haven’t yet proven that they’re any more successful than what you were already doing.
You should instead draw from a similar group of borrowers (perhaps current or previous customers who might be interested in another account) and randomly extend them an offer for either the credit model you already use (a control) or an experimental model (with different variables). Make sure that each variable is tracked independently. It won’t do you any favors to compare your current model to a new one that has a different interest rate, amount financed, payment schedule, and collection strategy; when differences in profit emerge, you won’t be able to trace which variable caused them. Instead, have one model that’s a control, and one model for each variable you want to test. Then, you can create additional models that pair different variables together, so you’ll be able to see their cumulative effects as well.
For any kind of meaningful analysis, you’ll need a strong reporting system as well, which can take information from accounts and either compare them directly or allow you to export that data into your own analysis tool. In addition to the variables you’re testing for, you can also look at how different types of borrowers react to different credit products. How do age, region, credit score, or previous experience with your company affect whether a potential borrower accepts a credit offer, and more importantly how well they repay you? You might learn that some aspects of your proof of concept work better with one group than others, directing you toward more targeted products and campaigns.
Going All In
Designing, developing, and testing a product are all critical steps, but they’re a means to an end. At a certain point, you begin asking yourself if the proof of concept is ready for a wide-scale launch. The right time will vary depending not just on your internal development, but also on external factors, like what type of credit products are growing in demand or what your competitors are doing.
Before you launch, you should make sure that you at least have a few boxes checked:
- It earns a significant profit on a small scale. If the test run is just barely profitable, you might see losses when you expand.
- It delivers a quality customer experience. New products often have some kinks to work out, so you should try to detect and correct those issues before scaling up.
- You have a plan to iterate. Even if the product works short-term, you need a plan to gather and analyze data, learning where you can improve and how you can adapt going forward.
If these three things are true, then you’re ready to take on an investment (whether from an outside entity or your own company) and build out your product. Perhaps even more important, though, is that you’re ready to win customers with an innovative product and delightful experience.
You can find more strategies for winning and making the most of investments in our free ebook, Lean into Lending, vol. 2: Working with New Capital.