Investing in Ophelos

Posted: 5 Jul 2021

Building better debt collection

I really didn’t want to invest in this company.

Yes, you read that right. Probably not the opener you’d expect but (sadly) it’s true.

I was introduced to Ophelos by two separate people — both of whom I trust and both had strong endorsements of the founding team — and yet both times I turned down the initial invitation.

Ophelos were pitching themselves as a “challenger debt collection agency”. My lazy lizard brain told me I didn’t want to be in the debt collection business and as for building an ‘agency’, well, it’s almost a dirty word amongst technology startups. Two easy reasons to say no. And, as it turns out, two reasons I was wrong. I’ll explain why but first let me set the scene with the founding team and origin story.

Note to self: Keep an open mind. Powerful outcomes can come from overcoming my preconceptions.

(*BIG thanks to @alexflamant for the initial intro and @mattwichrowski for persisting with me. We’re delighted to have led Ophelos’s $2.3m pre-seed round alongside our friends at Fly Ventures.)

Origin story...

Amon (co-founder CEO) and Paul (co-founder COO) met whilst working at ASAPP, a New York-based venture-backed AI startup. Amon was a director of their EMEA business and Paul was Head of International. They were the first of the ASAPP team to land in London and quickly became close. ASAPP’s proposition was AI enablement for enterprise contact centres. Whilst selling into the financial services industry, Amon & Paul saw first-hand how large the internal debt collection teams were as part of these contact centres (e.g. in one case, there were 4,000 call centre agents devoted solely to debt collection). They also saw how analogue the phone-based practices were within these teams. Amon became obsessed with the idea of applying technology and machine learning to optimise and solve many of the problems associated with debt collection (both enterprise and end-customer problems). Paul had previously been Head of Watson Group at IBM in EMEA and didn’t need much persuading. As part of their learning, they realised there had only been a handful of technical research papers written on the debt collection subject in the last 50 years. Only one of these papers was from the last 5 years — enter Qingchen (Ophelos co-founder and Chief Scientist). Qingchen, who is now also a professor at the University of Hong Kong, had written a paper titled: Data-driven Consumer Debt Collection via Machine Learning and Approximate Dynamic Programming. Not the catchiest of titles but his unique knowledge and experience meant that the stage was finally set for Ophelos.

Why I was wrong...

Your own reaction when reading “debt collection” might have been similar to mine. It’s definitely not sexy. It conjures up all sorts of images we’d prefer not to think about — bailiffs, fear tactics, injustice and ultimately an uncomfortable amount of personal hardship. I’m not sure anyone grows up wanting to innovate in this market. And yet this is exactly the reason not to bury our collective heads in the sand.

There are some markets we’d prefer not to exist because of the societal consequences they create. Debt collection isn’t one of them. So long as we want to keep paying for products and services whilst we consume them, there will be debt. And for as long as there is debt (regardless of what you call it) there will be debt collection — because life isn’t predictable or within our individual control. Quite simply, our economy and society can’t function without debt or indeed the practice of enforcing unpaid debts.

Rather than pretending debt collection doesn’t exist, the most powerful thing we can do is focus on the quality of people, products, and services tackling this market. Or as the Ophelos team would say: “Reimagining debt collection — for good.”

How Ophelos fits our investment thesis…

Opinionated products

The debt collection industry is dominated by incumbent agencies that employ analogue and labour-intensive processes. The combination of letters, SMS, human call centre agents, and lots of persistence, has historically resulted in the best debt recovery rates. There is logic in this approach if you default to the lowest common denominator and believe that all people with overdue debt are in some way ‘bad’ (i.e. have done something wrong or are intentionally not paying their debts).

The Ophelos founders have a different point of view on the world. The beliefs that struck me when I first met them, and the things I believe will define the products they build can be summarised as follows:

  • People want to pay off their debt — trite but true. This simple belief has major implications for the solution one would design to tackle this market. It’s akin to the eBay founding principle — “people are basically good” — a belief that allowed eBay to bet on building online peer-to-peer trust. People in debt are not criminals. The vast majority of overdue debts are due to changes in circumstances that are outside of a person’s control. If you build the right tools to help them get out of debt, this will naturally become the path of least resistance.
  • People want to self serve — money is one of the most personal parts of our lives. And issues with money can frequently be shame-ridden. People often don’t want to talk to their closest friends and family about money issues, let alone a complete stranger on the phone. The Ophelos team believes that people want the chance to self-serve their way out of debt, AND they deserve the same quality of digital user journey that exists in nearly all other areas of their lives.
  • Machine learning can help — yes, we’re all tired of ‘AI’ applied to everything and anything. There are some problems well suited for today’s machine learning capabilities and others that are not. Ophelos believes (and has convincingly proven) that debt collection is one of the perfect ‘narrow’ problems where AI makes a big impact on performance.
  • Build full-stack — the Ophelos founders could have used their SaaS backgrounds and machine learning knowledge to optimise debt collection performance within enterprise debt collection teams. And, in the future, I believe they will pursue this strategy to have an even greater impact. But, in the short term, they believe the best way to change the debt collection experience for end-users and to improve collections performance for enterprise, is to control the process end-to-end. This requires building a full-stack debt collection agency.

Crafted with love

At Connect we back founders who want to build product-led companies. Product management experience amongst a founding team can be invaluable but it’s not essential. Of more interest to us is the way in which founders think and go about solving customer problems.

  • Crafted: the Ophelos founders started with an obsession for the customer problem but they weren’t prescriptive on the solution. They developed high empathy for their customers and spent time researching and talking with each of their stakeholders. They had opinions on what they thought the solution might be but ultimately went about proving their hypotheses with data. And importantly, each of these tests was approached in a way that allowed them to get to an answer in the fastest possible time. They also naturally defaulted to solving each part of the customer problem with software.
  • With love: as I said earlier, I don’t think anyone grows up wanting to innovate in this market. There is a larger purpose at work here for the Ophelos founders and it’s best demonstrated by their social impact thesis and the fact they’ve committed to becoming a B Corp.


This requires a reframe. Ophelos’s customers may never tell anyone about their delightful user experience and, in the nicest possible way, they probably hope never to use the product ever again. But none of this is to say that Ophelos is unloveable. The important thing for us is: (i) the quality of the customer experience, (ii) the importance of this experience to the customer’s life, and (iii) the size of the relative benefit versus a customer’s available alternatives. Through this lens, we believe Ophelos has huge potential for customer love.

By many

This is the one area of my investment thesis where I’d happily be proven wrong. Fewer people falling into debt arrears would be an objectively good thing (all other things being equal). But this isn’t a reality. Yes, the macroeconomic cycle will drive variations in the annual market size, but the underlying structure will remain the same. Global debt levels have roughly tripled in the last 20 years and 20% of this (c.$50 trillion — pre-Covid) is household debt. The debt collection market within English-speaking countries equates to c.$18bn.

Consumer credit is fundamental to the operation of our economy and society. Individual factors outside of our control will always mean that a large number of people fall into arrears and need debt management support every year. Rather than trying to restrict the size of the debt market, or make it harder for some people to access it, I believe building better debt collection products is a fundamental part of creating a healthier consumer credit industry.

The Wrap...

Call me naive but ultimately I’m an optimist and believe great talent and technology applied to the right problems can change our world for the better. I backed Amon, Paul, and Qingchen because they were passionate about taking on the really important but unsexy problem of building better debt collection experiences for those that need them.

I believe this can create a win-win for the industry:

  • enterprises (i.e those selling products and services) can achieve better collections performance on their overdue arrears (i.e. more money collected + at lower cost + with less brand risk)
  • customers (i.e. people with overdue debt) can experience a better journey and outcome (i.e. more self-serve + more personalisation + more flexibility + more focused human support if and when they need it)

And this is just the beginning of the larger Ophelos vision to create “a world where everyone is financially educated, empowered and equal”.

Thanks to Connect Ventures