Products that scale

Paul Graham wrote a very good piece called 'do things that don't scale'. It recommends start-ups to go to extraordinary lengths to find, acquire, woo and retain customers; things like free consulting, dedicated customer service, physically chasing and onboarding customers, and many other laborious things that lead to the coveted product-market fit before scaling.

The advice to do things that don't scale, however, is meant for early-stage startups; ambitious, scrappy teams that are still looking for what makes their customers tick.

For a product manager joining an established or scaling organization, the goal should be exactly the opposite one: to do products that scale, and product that stack.

So what is a product that scales/stacks? It's a product that:

• 🚀 is viral enough to sustain its own growth (scales)
• 🌱 is financially viable, either directly or through cross-subsidization, in the long run (scales)
• 🤝 has strong technical synergies with the existing offering (stacks)
• 🔁 improves on something that the company is already doing very well (stacks)

In May 2021 there have been some incredible product launches in the fintech space: Airwallex, the global business account, launched own acquiring in Hong Kong; Revolut, the UK-based neobank, launched a browser extension called Shopper, and Curve, the passthrough wallet, launched a partnership with Cardlytics for customer rewards.

All of these are very good products - but do they scale and stack?

Revolut Shopper: stack, no scale

One of the many products Revolut launched in its relentless quest to become a super-app is Revolut Shopper. It is a browser extension that generates single-use, burn-after-using virtual debit cards. It auto-fills your e-commerce forms with this virtual card where possible, and (potentially) gives you an additional discount code to use - reducing friction and increasing safety during checkout.

This is a product that stacks, but doesn't scale.

It stacks, because it brilliantly leverages Revolut's penchant for trying out new things to take control (and therefore stress) away from the user and into their own domain. It has very strong synergies with their existing virtual cards offering, and of course improves on what's already a successful card program by driving additional volume.

But it doesn't scale, because it's neither viral not sustainable. It is not viral, because there is no way for a non-Revolut user to be 'lured' into this product without already having an account: the product can't sustain its own growth, instead becoming an additional service on top of the idiosyncratic set of features Revolut offers. It does not serve as a conversion funnel into the main revenue product (the card program), instead working as another feature most users won't care about.

Its biggest weakness, however, is sustainability. Contrary to popular belief, virtual cards do cost money to issue: for every new virtual card, a fee needs to be paid to the card scheme that issues it; these fees are not negligible, making single-use virtual cards ultimately unviable for purchases (and interchange schedules) that are not business-sized. Revolut either got a deal from the card scheme to have this cost waived for a limited time, or decided to eat the cost in the drive for more customers and more transactions. Neither of these options, obviously, makes for a sustainable product: it only works as long as someone else (either a profitable customer, the scheme or some venture capital) is footing the bill.

This doesn't make Shopper a bad product - it's actually a really cool idea, and might be a successful one: it just doesn't deliver as a scalable product, but only as a stackable one.

Curve + Cardlytics: scale, no stack

Curve is one of my favorite fintech companies: it is a wallet that acts as a bridge between your purchases and your existing bank, giving any boring debit card brilliant superpowers like spending insights and tokenization. Recently, it launched a partnership where a third party called Cardlytics gets (hopefully) anonymized transaction data from Curve's users - to drive market research and insights. Curve's users, in turn, get significant cashbacks (from 5 to 20%) paid out when they shop at certain merchants in the UK with Curve.

This product clearly has significant scale: it has the virality to drive additional customers into Curve, by enticing them with cashbacks that are way higher than even the wildest discounts seen in the industry (even if just at selected merchants): this is a product that is viral enough to drive and sustain growth. And it most likely cost little for Curve to partner with Cardlytics - since the real transaction counterparty is the users' data and spending habits, that Cardlytics will in turn sell to their corporate customers in exchange for more cashbacks and discounts, making it inherently sustainable.

I'm just not sure it stacks as well as it scales.

This is a third party integration, where customers agree to share data with Cardlytics (and its partners): from a technical synergy perspective, nothing (except maybe an exclusivity agreement) is stopping any other fintech to do exactly the same integration with Cardlytics. Unlike the existing rewards on Curve, this is a much smaller moat.

As for the improvement on what the company already does as a core service; well, Curve already has a cashback program, Curve Cash, which has much lower rates (1% to 3%). While it's clear how Cash and Rewards co-exist in the company's offering, it remains to be seen if customers understand the difference between the large Rewards offered by Cardlytics in exchange for user data, and the small Cash(back) Curve already does. The risk of self-cannibalization from the product that Curve actually doesn't own is quite significant.

Airwallex Acquiring: scale + stack

Airwallex, an APAC fintech that offers bank accounts, corporate cards, forex transfers and more to businesses that work across borders and currencies, recently launched an acquiring platform in Hong Kong.

This allows business merchants to issue a simple web link to their suppliers, customers, or partners. By clicking on that link, these parties can pay with a few clicks using (for now) VISA or Mastercard cards anywhere in the world, at domestic rates for Hong Kong, intra-regional rates for Asia and ANZ, and international rates for all other regions.

Out of the three products presented, this is by far the least flashy - and in fact it's the only one that requires a bit of industry knowledge to understand the use case. But when we consider the core business of Airwallex - to let businesses pay and get paid easily and rapidly across borders - it becomes clear that this is a product that both scales, and stacks.

It stacks, because it works hand in hand with Airwallex's existing offering of bank accounts, international bank transfers and card issuance. In fact, it is basically the missing piece that allows customers to never leave Airwallex's platform. By doing so, it improves on something the company is doing well, by closing the loop between spending money (the card), holding money (the bank account), and receiving money (the acquiring platform).

More importantly, it scales as well: customers now need to maintain one less supplier, and in fact can now both pay (using the virtual cards) and get paid (using the virtual acquiring links) inside the Airwallex platform. The potential for virality is huge: when a supplier receives an Airwallex link to pay, they become a prospect that gets exposed to the platform. And of course, it is inherently sustainable, as there is significant money to be made in Acquiring - with or without issuing sinergies - with gross margins often exceeding 1-1.5% of the transaction amount for smaller customers.

This is the iconic product that both stacks on top of the existing product, and scales significantly alongside it.

Stack and Scale

It's always hard to look at products you haven't worked on and try to understand the ideas and decisions (and often random turns) that lead to the features being launched. That said, looking at Revolut Shopper, Curve Rewards and Airwallex Acquiring within the framework of products that stack and products that scale definitely helps in understanding where these products are headed for.

Stacking and scaling are often underrated in an industry that is (at the moment) flush with cash. But I do try to integrate this perspective in my day-to-day work, and always ask myself:

• 🚀 will the product draw customers in?
• 🌱 will it be financially sustainable?
• 🤝 does it fit with what we currently do?
• 🔁 does it improve on our core offering?

Brilliant products don't need to always answer all four questions with a resounding yes. But in the rare cases where your product both scales and stacks, you can almost be guaranteed to have something great in your hands.