A case study: My learnings from finding product-market fit at Revolut 

In this article, Evgenia Suvorova, a Product Owner at Revolut, draws from her experience to outline the Product-Market fit framework that serves as a working tool for the Revolut product team.

9 min read
Share on

Launching a product is like embarking on an odyssey, filled with highs, lows, and uncharted territories. As a product owner at Revolut, I had the unique opportunity to navigate this tumultuous journey, from a nebulous idea to a credit product with dozens of thousands of customers across borders. However, behind the scenes of a seemingly protected environment within a tech giant, I encountered challenges that any business founder could relate to — the foremost being the elusive quest for product-market fit. 

At the time, I had completed my product discovery, defined value propositions, and launched the Minimum Viable Product (MVP) with real customers on board. Excitement surged, fueled by love for the product, the team, and positive feedback from testers. The instinct to go big, invest in marketing, and conquer the globe was strong. Yet, a lingering concern tempered the enthusiasm—the performance gap between the launch countries (Country A and Country B) made me question the precise timing for scaling in each locale. 

In product development, the term "Product-Market Fit" stands as the ultimate measure of success. This milestone is achieved when your company sells a product that effectively solves a problem for a specific market segment in a repeatable manner. 

I, of course, referred to this term too, but in quite an abstract way — I found it hard to actually decompose it or explain why exactly we did or did not have a product-market fit. That led to opinion-based dynamics and anecdotal evidence dominating our discussions. It was time to bring structure to our decision-making process and adapt existing concepts to the unique specifics of credit products. 

Given that my product was the unconventional one in our credit products family, loaded with hypotheses about the market, value props, and what people were willing to pay for, I took a gamble. I was tasked with turning our product-market fit quest into a system, using my product as the litmus test. The framework, later extended to credit classics like credit cards and personal loans, might not be universally applicable, but it worked for my peers and me, potentially inspiring other credit product managers.

I began by defining key questions for validation during the MVP stage: 

● Is the target market present? 

● Is the problem we’re solving significant enough for customers to pay for it?

● Is the model repeatable and eventually profitable for us as a business? 

To answer these questions, I established a set of metrics that allowed continuous monitoring of product-market fit and the impact of changes made to the product. 

Our quest commenced with a fundamental question: Is the market truly as we envision it? Validation began with a comprehensive survey and internal customer segmentation, unraveling the intricate layers of our existing product audience. The Sean Ellis test, a litmus test of users' emotional connection, provided a pivotal gauge for our target market's authenticity. 

Internal customer data segmentation (Age, Gender, Family status, Income, Probability of default, Product usage profile, Transactional activity) played a pivotal role. The "very disappointed" cohort from the Sean Ellis test, coupled with insights into segments' profitability, provided a nuanced understanding of our target group and refined the BNPL product's value proposition. 

Assumptions: Broadly defining the target group across Country A and B as convenience- and affordability-driven customers, lifestyle spenders, millennials, and low-medium risk, medium+ income individuals. 

Learnings: Survey and segmentation in Country B shattered preconceptions. Our core audience emerged as a liquidity-driven group of medium/medium-income, medium-risk customers, primarily millennials. While convenience-driven customers still found value, the segment was smaller than anticipated. This revelation prompted a gradual realignment of our value proposition, targeting, and risk criteria to cater to the most profitable customer group.

Moving beyond market validation, the focus shifted to the heart of the matter: Are we effectively solving the problem for our target customers? The Net Promoter Score (NPS) became our sentinel, with the question "How likely are you to recommend this product?" acting as the compass for customer satisfaction. 

Customer feedback collected along with NPS was one of the most useful findings. It highlighted challenges in comprehension, leading to UI enhancements and core feature adjustments. Gradually, Convenience, Flexibility, and Transparency emerged as the bedrock of perceived value proposition in both countries. 

Assumptions: Expecting our Global MVP to perform uniformly across Countries A and B, envisioning consistent satisfaction levels. 

Learnings: NPS scores painted divergent satisfaction landscapes in Countries A and B. While Country A basked in scores consistently within the 75-80% range, Country B witnessed less enthusiasm, averaging around 60%. The real treasure lay in the accompanying feedback, unveiling nuances in value propositions and feature comprehension. 

With validation and satisfaction as pillars, the spotlight shifted to attracting target customers at a reasonable cost. The Cost of Acquiring a Customer (CAC) emerged as the gatekeeper, influencing the product viability equation (LTV/CAC). 

Credit-specific funnel metrics—Application rate %, Approval rate %, Take up rate %, and Probability of Default at origination %—unfolded the narrative. Country-specific disparities not only highlighted approval rate challenges but also unveiled the relevance of the audience attracted in each market. 

Assumptions: Projecting high approval rates and application rates for the BNPL product, underestimating potential discrepancies.

Learnings: Discrepancies indeed surfaced. Approvals, take rates, and application rates fell short of expectations, triggering a series of funnel improvements. While Country A aligned closely with assumptions, Country B witnessed persisting gaps, especially in attracting customers with lower-than-expected credit quality. 

Product usage became the focal point, with metrics like Spend % of limit, Revolving rate %, Early repayment rate % and Operational default rate % dissecting user behavior. The exploration extended to fixed term and revolving credit products, with a keen eye on usage patterns. 

High early repayment rates and low revolving rates became indicators of audience sensitivity to debt and interest payments. Country-specific behaviors underscored the need for tailored approaches, especially in Country B. 

Assumptions: Assuming high average limit utilization for BNPL limits and optimistic activation rates. 

Learnings: Reality unfolded with a polarized audience—25% dormant users in Country A and a staggering 40% in Country B. While some users exhibited Spend >100% of limit in Country A, indicating revenue potential, Country B showcased more conservative usage patterns, affecting revenue generation. 

Celebrating initial success is insufficient; sustaining recurring revenue streams necessitates product stickiness. Metrics like MAP% or DAP % and Close rate % emerged as gatekeepers of churn and retention. 

Assumptions: Assuming a regular retention curve with gradual engagement decay over a 2-year period. 

Learnings: Dormant users surfaced as a challenge, comprising a significant portion. However, active users displayed remarkable consistency, with 40-45% cohorts in Country A and 30% in Country B making repeated purchases consistently. Hence, our biggest objective was to support new joiners with the engagement plan and to initiate the first usage of the product (through a promotion or otherwise) to overcome the first usage barrier and to build a habit. 

Now, at the last step, we’re aggregating all these metrics together in one simple product viability ratio: 

LTV/CAC ratio - is a measure of how much you make from a customer over their whole lifetime relative to how much you spend to get one customer. 

There are some industry benchmarks available for that ratio too. Logically, a ratio of less than 1 means lost revenue on every customer. 3-5 is thought to be a healthy level for growing a business; and the ratio >5 is a green light for additional marketing investments. We took the most ambitious metric of 5 as a reference, to be extra confident before we invest any resources in growth. 

Assumptions: Comparable unit economics in both countries at MVP stage. 

Learnings: LTV/CAC in Country A was way above the threshold, it was actually in double digits, which promised us good returns on our growth investments. 

In Country B, however, the LTV/CAC ratio was substantially below our tolerance, it was actually closer to 1. So we didn’t lose money, but keeping the product as it is would result in us hanging around the breakeven point going forward and being very vulnerable to any changes in the market. So based on the combination of factors, we had to review the product and the business model we used in Country B. 

Growth potential

The product-market fit framework paved the way for testing "growth hypotheses" before scaling. Metrics such as New sign-up growth % and Product penetration % were crucial for assessing future expansion. 

It was important to look at the combination of these metrics as we didn’t only want to rely on the existing growth rates but also assess room for future expansion. 

Setting up targets 

While industry standards and benchmarks provided some guidance, setting product-market fit targets was highly dependent on the specific products and markets. 

Hence, we used a diverse range of sources to come up with our target metrics: 

Product-market fit framework application 

The framework became an integral part of our product routine, with regular meetings to delve into the metrics. Products and markets performing poorly underwent re-evaluation, while successful cases were boosted with increased marketing efforts and plans for new features. 

The biggest lesson I personally took away from this whole process is how tricky the idea of a Global MVP can be. It worked well in Country A but fell short in terms of value proposition in Country B. Even though this is the very reason we built an MVP in the first place, it taught me to be more careful about using a one-size-fits-all approach when launching new products in different locations.

Explore more great product management content by exploring our Content A-Z

Comments

Join the community

Sign up for free to share your thoughts