Launch Legends: Ben Shen on building Cash App from P2P payments to a financial ‘super app’
When Ben Shen joined Cash App in 2016, it was still known as Square Cash — a slick new peer-to-peer payments app. Fast forward to today, and Cash App now offers a range of financial services to nearly 60 million monthly active customers, including a card, savings, investing, tax filing, and more.
In this conversation, Ben, who led Cash App’s banking teams, shares insights on the evolution of the platform.
You came from a finance and strategy background before moving into product. How has your career path shaped your perspective and approach as a product leader today?
I started my career in traditional financial services, first as a derivatives trader at Goldman and then working as a buy-side analyst. The general arc of my career has been moving closer and closer to the ground - going from managing a book of derivatives risk to analyzing how companies and industries function, to ultimately wanting to be a builder and operator, leading me to join Cash App — then called Square Cash — in 2016.
My background in finance and strategy has been meaningfully accretive to my effectiveness as a product builder. For any transaction-based business such as Cash App, understanding customer behavior and unit economics is akin to understanding the holistic business model. Understanding how customers engage with features such as P2P or the Cash App Card, and how adoption of individual features impacts holistic engagement with the broader platform, helps one form an incredibly good pulse on platform growth and business health.
When the Cash App Card was launched in 2017, we looked into how that affected P2P activity. Realizing that adoption of the card was meaningfully accretive to P2P engagement allowed us to lean into driving adoption of the card much more aggressively. Over time, you learn how to build flywheels upon flywheels, getting customers to hire you for an increasingly broad set of customer jobs and expanding customer LTV and your ability to acquire folks onto the platform.
Cash App is arguably the closest thing we have to a financial ‘super app’ in the US today. How did it evolve from a simple P2P product to a full financial ecosystem that includes banking, trading, P2P, etc.?
Cash App never set out to be a ‘super app’ - though the reality is that money is one of the world’s biggest common denominators. In Cash App’s case, the concept was always about driving economic empowerment, using money as the conduit to affect positive change.
It started with P2P payments — moving money quickly using card rails, as opposed to ACH, which is what most platforms relied on. That foundational capability made us competitive and attractive, particularly for customers who valued speed of money. From there, we noticed that many users used instant deposit to move money off the platform, and we introduced the Cash App Card, it similarly provided instant access to funds, allowing the customer to spend anywhere Visa was accepted without paying a fee.
From serving immediate spending needs, like P2P and card transactions, we then expanded into things like savings and investing, which were not as in immediate focus on Maslow’s hierarchy of needs, though were important in helping many of customers further their financial health. The evolution was very organic, and was a combination of meeting customers where they are, while also helping push them towards advancing their economic situation. The approach was ultimately grounded in doing one thing well for customers, followed by a few things, and then expanding to other adjacent needs.
I'm curious, how do you think about cannibalization? As you build out new products, is there a North Star that guides product decisions across teams?
Cannibalization comes up often, though instead of framing products as competing with each other, we’d focus on expanding the overall pie. For a lot of customers, Cash App isn’t their entire financial life, so there is a huge opportunity to grow the portion of their financial activity they conduct on the platform. The framework that ties it all together is inflows — that is, how much money users bring onto and keep on the platform. How can we meaningfully increase total inflows per active user or the average balance per user? If we do that well, we can grow the overall pie and experience fewer trade-offs between individual features, like spending vs. saving vs. investing - even if cannibalization will of course never be zero.
How does that show up in organizational design for the product teams?
Cash App doesn’t have a single team that owns a metric like ‘inflows per active’. Inflows come from many sources — P2P, direct deposit, physical cash deposits, and more — so teams are organized by domain or features that have close connections to one another - whereas metrics like inflows per active are horizontal north star metrics that span the entire product and transcend specific verticals or teams. Everyone understands that even if they’re working on their individual team metrics, it should also ultimately be accretive to north star metrics like inflows, thus contributing to the success of the overall ecosystem.
What's been the most controversial product decision you’ve made?
One which comes to mind is the launch of physical cash deposits in 2021. Internally, a lot of folks didn’t use cash much in their everyday lives, and hence there was debate over whether it was high leverage. However, it became clear that for a significant portion of the customer base, there were many who are still paid in cash or use physical cash in their everyday lives and not having this would have been a meaningful primary banking blocker. There have been several other spirited debates along the way, about features like savings, automatic paycheck distributions and others.
A lot of early-stage startups wonder when it’s right to move from a single product to a full suite. How do you think about the timing for launching new products?
Cash App has historically focused on doing fewer things really well and getting resonance from customers, even if it seems narrow at first - and then expand from there. One approach has been to focus deeply on core, proven customer needs which transcend a singular feature, such as speed of money, liquidity, flexibility / control, and to build out products and features which serve these core needs.
What does the process look like during the pre-product launch or discovery phase?
Cash App aims to operate teams that eat, breathe and sleep their domain, and feel empowered to build and launch products with great built-in distribution. Discovery starts with a clear declaration: What’s the customer problem we’re trying to solve? What’s the business need? At the end of the day, the flywheel is to serve customer needs amazingly well, drive deeper engagement and inflows, translating to a healthy, lasting business which is even better positioned to serve customers.
After the declaration comes discovery: how we might solve it. Sometimes we let our imaginations run and pressure test different things: how the problem is currently being solved, where it falls short, what constraints exist, and whether those constraints apply to us. Eventually, the rubber meets the road: what are we actually going to build and ship? Once you’re in the market and customers are interacting with the feature, that’s when a lot of learning happens. Customers will engage with the product in ways you didn’t expect. For example, when we launched savings, we saw people using it in all sorts of creative ways: saving for short-term goals, splitting bills with roommates, and more. Those insights often spark new ideas and further evolution of features.
I want to talk about things that haven’t worked, too. Has anything surprised you in a negative way?
If every product you launch is a hit, you’re probably not pushing the boundaries enough. There are some products that have tremendous resonance and others, where the data showed otherwise. For example, we’ve built the ability for customers to send each other money in many different forms such as stocks, bitcoin, and gift cards - though most people still preferred sending each other money in a simple fiat USD format.
At the end of the day, it’s important to get things in the market quickly rather than be building in a cave for too long, based on assumptions on how customers will ultimately engage. Allowing customers to vote with their actions helps inform what we double down on, what to keep as-is, and what to sunset.
Looking at the broader FinTech space, a lot of companies seem stuck on a single dominant product line. Experiments don’t often grow into full ecosystems. How would you diagnose what's gone wrong? Or the failure modes that you see?
One common failure mode is starting with the mindset of: ‘We want this business model, so let’s build that’. That approach isn't customer-centric enough and often manifests an approach which doesn't have a high chance of succeeding. If you decide to be a neobank for X demographic and then just copy-paste features, like card, direct deposit, savings, it's a wild swing on whether those will resonate.
Instead, Cash App has always tried to approach things by thinking deeply about the customer's needs. With the Cash App Card, there was both the adjacent job around instant access to money, along with helping people feel better about their money - given how visible a card is as a representation of one’s financial situation. With savings, it wasn’t just ‘let’s offer a savings account’ it was ‘how do we make it easier for our customers to save?’ When customers are already engaging with your platform at each step of the way, it's easier to expand into chapter two.
Do you think there's a connection to poor adoption? I see a lot of companies launch a new line and get some % adoption rate but it rarely becomes a breakout success. When you don't have customers' needs in mind, where does it break?
Having a large customer base and distribution is a big advantage, though it’s not enough. If you just cross-sell something unrelated to how people already use your platform, adoption becomes a wild swing. People value speed, control, liquidity, peace of mind. The better approach is to anchor it on these higher-level customer needs and the utility it offers. When you build around those, you have a better shot. Cross-selling something just because it’s good for your revenue rarely works unless it also clearly serves a customer need. It’s fine to build a suite of products, though how you build each part matters. How you build and launch individual product lines shouldn’t just be about checking boxes, and the approach should ideally be more customer-centric and creative.
Cash App is more than just a financial tool, it feels like a cultural movement. How did you do that?
The simplicity and elegance of the design are definitely a part of it — those little moments of delight when you use the app. Though another major part has been the team’s creative approach to engaging with communities. From the early days, Cash App’s design and marketing teams were a huge driving force - whether engaging with the community through the Cash App Twitter account with a distinct voice - which at the time was rare for a financial platform, or being early on advertising on podcasts, gaming platforms, and many other creative approaches.
This combination, mixed with the magic of the product, led to organic resonance from a cultural perspective. Folks across industries, like entertainment and music, reached out to us because the brand resonated with them, and they wanted to use Cash App in their own way to connect with their own communities. Over time, other fintech platforms have aimed to replicate some of these strategies, though the way it unfolded for Cash App was quite novel and organic.
My understanding is that Cash App has strong resonance among US teens. How has this come about
As people go through life, their financial needs get more complex. Teens, being earlier in that journey, have simpler needs, though they still want a clean, intuitive experience. Whether it’s getting money from their parents, sending money to friends, or making basic purchases, Cash App’s product ethos — simplicity and elegance — works well for them. Cash App has built tools that empower teens and their parents. Teens can handle everyday transactions, while parents feel confident that their kids are learning responsible habits. By starting early, Cash App is helping set them up to feel confident and in control as they move into adulthood and face more complex financial decisions like earning income, saving, or investing.
Any hot takes on common product best practices you strongly disagree with?
One thing I feel strongly about is to be really clear on whether something is being run as a light-weight test which won’t be rolled out at meaningful scale, vs. a full-fledged MVP with the aim of a generally available launch. I’ve disagreed with the approach of framing something which we want to roll out to 100% as a ‘test’, simply to lower the bar on required quality and conviction.
What are the qualities you see in the best PMs you’ve worked with?
First, an ownership mentality and treating the product like it’s one’s proverbial neighborhood store. You care about every detail, big or small. Second, intellectual curiosity. It’s great to have deep industry and domain knowledge, especially in regulated areas, but I also look for people with a beginner’s mindset - people who can approach new problems with fresh eyes and creative thinking, and draw patterns between even seemingly disparate things.
And how do you test for that in interviews?
I dig into what they’ve built — how they went from idea to execution. What was the ‘why’ behind the product? Was it customer insight or a business need, And how did they think about how to build it? Did they think about how to creatively leverage the platform’s unique strengths / advantages beyond a cookie-cutter approach?
Last question: How do you see the PM role evolving in the next 5-10 years with the rise of AI tools?
There is an increasing blending of capabilities and roles across the development organization. One can now draft product docs, come up with design mocks, build dashboards, and build prototypes more easily. As such, there’s a need to embrace blurrier cross-disciplinary lines, while still finding ways to build collaboratively and respect each discipline’s ability to bring their superpowers to the table. The best PM is someone who can flex their superpowers, while also being able to put on a design, data science, design, finance, or other hat. As AI enhances folks’ abilities to flex in a cross-disciplinary way, this will become even more pronounced - though this should ultimately be accretive to the overall team’s output, similar to how modern NBA lineups consist of multiple folks who can handle the ball, shoot the three etc, rather than positions being as specialized as they were in the past.