• Fintech Brainfood
  • Posts
  • Fintech 🧠 Food - September 5th 2021 - Amazon & Affirm partner, PayPal Stock trading, and the Fintech provider tradeoffs.

Fintech 🧠 Food - September 5th 2021 - Amazon & Affirm partner, PayPal Stock trading, and the Fintech provider tradeoffs.

Hey everyone 👋, thanks for coming back to Brainfood, where I take the week's biggest events and try to get under the skin of what's happening in Fintech. If you're reading this and haven't signed up, join the 7,644 others by clicking below, and to the regular readers, thank you. 🙏

Hello 👋

This past week Y-Combinator held its demo day for the Summer 21 batch of companies. Usually, I like to give a quick summary of those for you Brainfood readers. This batch, however, had 69 (yes, sixty-nine) Fintech Companies at demo day. So it’s going to take a little while to get through. So watch this space 👀PS. Can’t stop thinking about loot.

PPS. This is a long email this week, if Gmail clips it, visit the link here

Weekly Rant 📣

Fintech Providers - Making sense of the noise.

In the past half-decade, new providers have taken parts of the Banking-as-a-Service stack and brought it up to modern software standards, but how they've done that differs.

We can picture the stack as below, at an abstract level 

(There's tons of nuance missing here, of course, but it's a helpful frame for a conversation).

Then when you map providers against this map, you'll see things get messy.  

  • Some providers fit in multiple categories

  • Some providers are vertically integrated

  • Some providers own a horizontal layer

  • Some providers do just a smaller piece of capability.

  • And now, some providers give developers the primitives and lego bricks to build themselves.

(To every provider I missed here please don’t @ me. Also for the sake of fitting logos on a diagram, I know some providers here do more than the categories they’re in, some arguably less. I guess that’s kind of the point. It’s very hard to do apples to apples in Fintech provider land.) Update 6th September: Six providers have reached out complaining about this infographic. And you wonder why Fintech buyers find this stuff hard? Maybe we should open source this stuff and let them duke it out?

There’s so much nuance in what each of these companies actually offers. Amount is much more vertically integrated, as is Solaris. A company like Alloy is much more horizontal, and may often partner with many of the other names you see in the diagram. There’s almost no true apples-to-apples comparison in Fintech provider land.

Two providers might look like they do "compliance," but what they actually do under the hood can differ dramatically, as can how they do it.

Some of these folks have full banking licenses, some have payment licenses and existing partnerships, some just provide software. They’re all the right answer for someone, some of the time.

But the trend is, the broader the offering, the less changeable (or more opinionated) the offering is. If it does more for you, it’s harder for you to make it work the way you want it to (if you have a bespoke use case).

None of these approaches are wrong, but they present their customers with a trade-off.

Opinionated vs. Build and assemble yourself.  

Last week, I spoke to the CTO of a consumer Fintech who lamented the provider landscape (especially as they are building a remittance app, which involved a lot of cross-border activity). This CTO had worked as an engineer in countless consumer Fintech companies and encountered the same problem repeatedly. 

The choice Fintech builders face is between a highly opinionated provider who gets you to market quickly or building yourself to solve your specific problem.

An opinionated provider makes choices on your behalf to help you get to market quickly. The canonical example is Stripe, which hides mountains of payments complexity behind a single API. Stripe is making choices about how to handle errors or how analytics work. Customers have almost zero internal build to do to be able to accept payments. The same is true for providers in compliance-as-a-service, fraud-as-a-service, and more who own either a horizontal or vertical layer in the BaaS stack.

To borrow from Packy M's "APIs all the way down," when something needs to get done, but it's not worth our time doing it, hiring a 3rd party API provider like Stripe makes sense.

An opinionated provider not only improves time to market, but they're also 100% dedicated to that thing you don't want to waste your time on. They help companies reduce their R&D spend and prevent them from maintaining code every time the payments world changes. Oh, and payments and finance are regulated; not having to deal with all of that complexity is a massive accelerator.

If you're at an early stage or simply want a capability that "just works," opinionated APIs and providers are the dream. For 80% of what you want to do, they’re likely all you’ll ever need.

There are a few challenges with opinionated provides, though.  

Opinionated providers often cost more than if you built a capability yourself. The more opinionated the provider, the more margin they can justify because they displaced more of your own R&D Capex and Opex spend.

Opinionated providers also, by their nature, work in a certain way. That might be fine if you're part of their core user base, but going back to the example of our CTO building a remittance app. If your fraud-as-a-service provider assumes anything moving into Kenya has a higher probability of fraud, that might break your product.

Interestingly, we're now seeing a middle ground emerge. If we go back to Stripe as an example. The emergence of "Payfac-as-a-Service" is designed to be halfway between building a payments division yourself or outsourcing wholly to a PayFac like Stripe. Companies like Infincept, Finix, or Payrix play in an interesting "we'll make you a PayFac with our more configurable platform" space.

The reality is opinionated vs. build isn't a binary choice; it's a spectrum.

The choice is a trade-off. How important is time to market? How much can you risk vendor lock-in for this capability? And how much do you want to own the economics of that capability?

As every company becomes a fintech company, owning the economics can be appealing. But in the earlier stages for a Fintech company, getting live and showing traction to investors is vital.

The new primitives of finance help.

One way out of the spectrum of opinionated vs. build yourself is to have open-source codebases that contain the individual software primitives your developers need.

As covered previously, projects like Moov and fragment have taken the non-differentiated code built in-house by developers and built projects and communities around those open-source codebases. This code is often "close to the metal," handling deep payment infrastructure or lookups against sanctions lists.

These primitives of finance can be combined as developers see fit and often do one thing exceptionally well. The codebases are battle-tested as some of the biggest companies around the world adopt them at scale and make improvements over time. The primitives don't carry the same level of opinion as the API providers but help avoid some of the self-build issues we saw previously.

The primitives often suit larger companies operating at scale and improving their unit economics or developers looking to something more bespoke than the API providers would allow.

These primitives present their own challenges, though.

If you're a relatively smaller organization, you have to add them to your platform or around other providers you have. This requires custom work and build. This potentially increased Capex, the effort involved, and time to market.

We're also in the early stages of developing fintech primitives. They cover a small % of the Banking-as-a-Service stack and are very US-Centric (so back to our remittance app for Africa, not helpful in the short term).

The Orchestration Challenge

There’s something missing from our picture.

How do all of these providers fit together?

It feels like there’s a box missing from the diagram, that runs alongside the whole stack, orchestrating it, and allowing you to make propositions out of it.

We have a few approaches emerging.

  • Vertically integrated BaaS: Companies like Amount vertically integrate several Fintech specialists to allow banks to roll out lending at pace

  • Loosely coupled BaaS: Fintech companies like Unit and Bond allow broader BaaS use cases with a mix of pre-built partnerships and "roll your own partner bank" approaches.

  • Horizontal BaaS: Providers like Alloy are becoming the default for identity, AML, and compliance as a horizontal (and actively partnering with the more vertically integrated BaaS providers)

When companies are small, they often make design choices that enable speed to market, but that create technical debt or vendor lock-in. So they’re likely to choose an “all-in-one” solution that gives them as much of the stack as possible.

Companies that scale eventually get to replace their technical debt, but they do so by building their own orchestration software.  Every fintech company or non-bank that does Fintech is now wrestling with an orchestration challenge. Imagine you're a global e-commerce platform, and you're live with Neobank like features. Maybe you have partners for the US and, to some extent, Europe. What about other markets where those providers don't exist? 

Maybe loosely coupled BaaS is the way there, or maybe we need an alternative. If everyone is building the same orchestration code, maybe we need that, and just that.

Just as we have seen payment orchestration become a category, I think we'll see Fintech provider orchestration become a category. Maybe that comes from the existing BaaS players, but I actually think it in itself is a specialism. For companies who don't want to use a BaaS provider but do have an orchestration problem. Who want to avoid vendor lock-in, potentially go multi-geo, but don't want to build all the primitives themselves. What do they use? 👀



4 Fintech Companies 💸

SaaStrify - Self-driving SaaS procurement

  • SaaStrify handles SaaS subscriptions, renewals, and cost monitoring for SMBs.  SaaStrify uses industry benchmarks and has a "dedicated procurement team" that handles negotiations on behalf of SMBs. 

  • For SMBs, sometimes just keeping track of what SaaS platforms are in the business is hard enough, but getting a better deal is harder. SMBs often don't benefit from group discounts or understanding how to make the most of the more complex providers (like Google or Microsoft). There isn't yet a good alternative to SAP Ariba or Oracle Netsuite, but many of these spend management platforms could evolve that way.

2. Prive - Amazon subscriptions feature for any merchant

  • Prive helps merchants identify what customers are buying on a repeat basis and helps them create a custom subscription offering.  eMarketer data shows subscriptions are 3% of all e-commerce (representing $27bn of GMV). 

  • While other tools technically offer this, you don't see it around the web nearly as much as you do on Amazon. I suspect much of that has to do with the ability to predict what customers would want to subscribe to.   Prive feels like a feature as a product, but if executed right, one heck of a feature for the long tail. With everything e-commerce, "what if Shopify builds it" is a risk, but still. Neat.3. India Gold - Gold storage and collateralized lending

  • India Gold offers a modern, secure and affordable way for its customers to store Gold.  Banks in India would charge hundreds of dollars; India Gold can keep it for as little as $3 per year. Once stored, the Gold can be used as collateral for lending (at 10.2% APR).

  • Across India, families store more than 25,000 tons of Gold, but just 166 tons of that sit in a storage facility. Gold has a high risk of theft but is often not used.  I like this because it combines cultural relevance with the best of the modern internet age. India Gold uses a simple biometric lock for the Gold items and then uses the weight and value of the Gold to lend against. This offline to online combo is like the Fintech lending equivalent of ride-hailing or food delivery. The best of digital and physical combined. Really smart.

4. Bloom Money - Community-Driven Fintech savings 

  • Bloom money aims to help people find extra savings, involve their community or family in saving, and provide reliable remittances.  Bloom Circles are a group savings feature; users create a group and can contribute savings, and later if needed, borrow from that group. This replicates many traditional cultural practices from different societies. 

  • In many communities, money is social. This is especially true for migrant communities who often travel and remit money home.  Led by Nina Mohanty (ex-Starling and Bud), Bloom has an opportunity to replicate the success of the US "digital community banks" like Daylight, Greenwood, and Cheese. 

Things to know 👀

  • Shares of Affirm closed up more than 47% as the partnership with Amazon was announced. This move follows Square buying AfterPay for $29bn and Apple announcing a partnership with Goldman Sachs for BNPL.

  • 🤔 My Analysis: Amazon absolutely dwarfs any competitor with a 40.4% US e-commerce sales market share The nearest competitor Walmart has just over 7%, and eBay just over 4%. Affirm achieved $2.1bn of GMV in Q2 2021. Amazon did $367bn across all of 2020.  If 1% of Amazon customers use BNPL, that's $3.6bn extra GMV in the year (or nearly $1bn per quarter). This one deal could increase Affirm's GMV by at least 50% when live.  No wonder the share price jumped.

  • 🤔 My Analysis: We know merchants love BNPL, but the data is still staggering. Shopify (another Affirm partner) found using Affirm's BNPL increased checkout conversion rates 50% higher, saw 28% fewer cart abandonments, and checkout times that were 27% faster.

  • 🤔 My Analysis: BNPL is the future of lending. Consumers love it because the installments by item make them feel in control, and the lending is in context (it's for that thing I want to buy). Done right, BNPL is in the interests of consumers (because it can genuinely be 0% if the consumer keeps up their payments).  Credit cards are open-ended and don't feel as purposeful as BNPL. Therefore they don't give consumers that same sense of control; actually, the opposite, they feel like a slippery slope into more purchases.  

  • 🤔 My Analysis: BNPL also feels like the next big consumer credit scandal waiting to happen. The lack of friction, combined with the soft credit checks, could also be a slippery slope. A recent study by the UK Citizens Advice bureaux found 1 in 10 BNPL shoppers "are being chased by debt collectors."

  • 🤔 My Analysis: This is a very US-centric story. Amazon has some major European markets that Affirm has no presence in.  If Amazon wanted to do BNPL in Europe, they'd likely partner with someone else.

  • 🤔 My Analysis: Why hasn't anyone built the Open Banking - one-click - BNPL + underwriting app yet?

  • PayPal has hired an industry veteran to lead "Invest at PayPal," a previously unreported division. The move comes amid a retail trading boom as consumers take advantage of apps like Robinhood.

  • 🤔 My Analysis: The Fintech Super Apps are gonna Super App.   Square, Paypal (and now, to some extent, Robinhood if it adds direct deposit) are all rebundling Fintech. It's therefore logical that PayPal would head this way; it already offers Crypto, P2P payments, cards, and BNPL.

  • 🤔 My Analysis: The providers make this possible. PayPal relied on 3rd party providers for its Crypto solution; is it likely to do the same for stock trading?  The Fintech Super Apps are very good at renting and assembling Fintech providers.

  • 🤔 My Analysis: We're living through the great speculation. There's too much money in the system with stimulus, low interest rates, and quantitative easing. Consumers have lots of dry powder for savings after the pandemic and nowhere to make a return.  When the reckoning from this speculation comes, it will be brutal.  And I wonder if the Super Apps like PayPal and Square that added stock trading will be in the firing line the way Robinhood already has been?

  • 🤔 My Analysis: Most banks haven't got the joke and are still trying to use their existing internal machinery to compete with the Super Apps and build in-house. That's like bringing a knife to a gunfight. They need to do something drastic if they want a piece of the action.

Good Reads 📚

  • If 2020 was the year of DeFi and 2021 the year of the NFT, then 2022 will be the Community DAO Crypto isn't just a technology; it's an internet-native economy. Revenue is tracked in ETH. Key decisions are made with community governance. The org structure of DAOs looks like modern-day cooperatives but could rival some of the world's largest corporates. 

  • A Community DAO has several core features. A group chat (e.g., Discord server), a Treasury (shared wallet, in some examples where a group collectively buys NFTs). A token ("DAO shares") and, in some cases, the token is used to gate access to group chat. Governance, where the members vote on decisions enacted by 5 to 10 core contributors. Cash flow, where the DAO generates income or revenue (still limited).

  • 🤔 My Analysis: This is a great read, and Patrick goes into all of the ways DAOs could make money. Including membership fees, creating their own art (NFTs), collecting NFTs, etc.  To me, Community DAOs are a mashup of a membership club, angel syndicate, and business. Except like a cooperative, they don't focus only on a profit motive.

  • 🤔 My Analysis: Seth Godin published "tribes" in 2008, arguing that anyone had the opportunity to start a movement with social networks. With Community DAOs, the internet native generation can create a fund, media company, and membership club all rolled into one.  If Crypto crashes, Community DAO's might not have the funds they do today, but they also won't disappear (just as Bitcoin and Eth didn't die in the last crash). This feels like a trend that's here to stay and worth paying attention to.

  • 🤔 My Analysis: If this all sounds like speculative twaddle, heck, maybe it is. But consider that if you own stock in a company, it's unlikely you can vote in that company's annual general meeting (AGM). If you own DAO tokens, chances are not only can you vote, you're encouraged to do so and often rewarded for contributing. DAOs have near full-time contributors and are drawing in some of the bright young, creative minds.  As Chris Dixon said, what the smartest kids do at the weekends, will be what everyone does for work in 10 years. 

Tweets of the week 🕊

That's all, folks. 👋

Remember, if you're enjoying this content, please do tell all your fintech friends to check it out and hit the subscribe button :)