The Fintech Tax

AWS is fueling (and profiting from) a technology arms race.

Editor’s note: I’m thrilled to be publishing a guest essay from Tom Johnson, Director of Go-to-Market Strategy at Amount.


A quote from a story about Chime’s last funding round, which pushed the neobank’s value to a lofty $14.5 billion, caught my eye:

We’re more like a consumer software company than a bank … It’s more a transaction-based, processing-based business model that is highly predicable, highly recurring and highly profitable.

It’s smart messaging. You see it a lot in fintech. We’re not a bank. We don’t have those useless and expensive branches weighing us down. We’re a software company that happens to deliver banking services. Value us appropriately please.

Here’s the former head of Marcus, by Goldman Sachs (emphasis mine):

we are trying to build a consumer digital financial services platform. On their saving side, we are giving them deposits, which have significantly higher interest rate, online savings accounts. No fees, $1 minimum opening account, and significantly higher interest rate than they get from their traditional banks. And on the lending side, we are giving them loans with 3% to 5% lower interest rate than on their credit cards, and again, no fees, and both in a very transparent and convenient fashion. And we can do that because we don’t have any legacy costs

Translation: we don’t have branches and the legacy infrastructure of a traditional bank. We’re a software company.

The problem with this argument is that there are hidden costs associated with delivering digital financial services too. Costs that are much more pernicious.

Imagine if banks rented (rather than owned) branches and that their rent was based on a per-customer visit price and the very act of using those branches to run their business made it technically impossible to ever stop.

Enter Amazon Web Services.

A Tax on the Internet

Amazon Web Services (AWS) generated $10.22 billion for Amazon in the first quarter of 2020, a 33% increase in revenue on an annualized basis. In that quarter, AWS accounted for 77% of Amazon’s total profit. In other words, AWS is the profitability engine driving one of the world’s most valuable companies.

And it started, essentially, as a footnote (emphasis mine):

Chris and I wrote a short paper describing a vision for Amazon infrastructure that was completely standardized, completely automated, and relied extensively on web services for things like storage. We drew on the work of a number of other folks internally who had been thinking and writing (and sometimes even coding) in the storage services space, and we combined it with our own thinking and experience in infrastructure. Near the end of it, we mentioned the possibility of selling virtual servers as a service.

In 2003, Amazon was entirely focused on fixing its own computing infrastructure to better support its growing e-commerce business. It wasn’t planning on getting into the cloud computing business, until a couple of engineers tossed that idea out there in a paper for Jeff Bezos.

But that was all it took. Within a year, Amazon had developed a compelling vision for this new service:

The paper laid out the expanded AWS mission: “to enable developers and companies to use Web services to build sophisticated and scalable applications”…

“We tried to imagine a student in a dorm room who would have at his or her disposal the same infrastructure as the largest companies in the world … We thought it was a great playing-field leveler for startups and smaller companies to have the same cost structure as big companies.”

That vision has become reality. Amazon has built a software development and hosting platform that empowers everyone, from the college student in their dorm room to Netflix, to deliver cloud-based software services.

It’s worth unpacking exactly how they did it because it illustrates the power and peril of using AWS. There are three steps in AWS’s strategy:

  1. Invest the money necessary to build and maintain a world-class cloud computing infrastructure. AWS never becomes AWS without Jeff Bezos committing to build the best cloud computing infrastructure that money can buy. That’s not a trivial decision. Many enterprise software companies that I’ve worked with over the last 30 years have balked at it. It’s incredibly difficult. Luckily

    The cost to build AWS was justified because the first and best customer is Amazon’s e-commerce business.

    Remember, the idea for AWS originated from a proposal for how Amazon could improve its own infrastructure. In other words, Amazon was already committed to building a world-class cloud computing infrastructure before the public cloud computing market (as we know it today) even existed. And as long as Amazon remains the world’s dominant e-commerce retailer, it will always be committed to maintaining that infrastructure.

  2. Build a robust and ever-increasing set of developer-friendly tools. AWS isn’t just for hosting software (although it started there). Today, AWS is a full-fledged software development platform. It offers developers a suite of advanced tools that can significantly shorten the product development lifecycle, allowing them to push new products and services out to their end customers faster and at a lower cost.

    AWS views these software development tools as a strategic priority, which we know by looking at the pace at which they are being made available to AWS clients. In 2014, AWS released 30 new ones. In 2019, it was 175.

    And while the first few generations of AWS tools focused mainly on core infrastructure capabilities (database management, elastic search, etc.), tools introduced in recent years have moved further up the technology stack, directly addressing challenges faced by business executives. Today, AWS provides developers with tools for training machine learning algorithmsbuilding chatbots, and detecting online payments fraud.

    These tools are incredibly attractive to software developers, which is the problem. They entice developers (and thus the companies they are building for) to build solutions using these tools. However, solutions built with these tools can’t be run in other cloud environments so the developers end up tightly coupling their technology to AWS, making it practically impossible to switch to a competitor like Microsoft Azure. This vendor lock-in inhibits healthy competition, which is key to driving down costs over time.

  3. Scale up volumes and enjoy the profits.The reason that Amazon and Google and Facebook are so incredibly profitable isn’t because of the products they sell. Retail and advertising aren’t exactly new ideas. It’s the way in which they deliver those products. The magical thing about the internet is that it allows for companies to deliver products to customers at almost zero marginal cost. That means that the more you scale up (and grow beyond your fixed costs), the more profit you make. This is why neobanks like to think of themselves as a software companies:

    Yes, I realize this view into profitability is overly simplified and that SaaS products have some marginal costs. This is just to illustrate the larger point.

    But, as Ben Thompson notes:

    it’s so much easier and cheaper to get started with AWS that the idea of buying your own server infrastructure — an expense that consumed the majority of venture capital in the dot-com bubble era — is preposterous.

    And so companies leverage AWS to lower their fixed costs, but end up unknowingly surrendering that magic ‘zero marginal cost’ upside to AWS:

    Meanwhile, AWS locks its customers into its infrastructure for the long term by encouraging the use of its add-on software development tools, which has the wonderful side benefit (to AWS) of further expanding its margins:

    And, as Chamath Palihapitiya points out, that upside for AWS persists indefinitely, as the overall volume of transactions taking place on the internet continues to grow:

    AWS is a tax on the compute economy. So whether you care about mobile apps, consumer apps, IoT, SaaS etc, more companies than not will be using AWS vs building their own infrastructure … If you believe that over time the software industry is a multi, deca-trillion industry, then ask yourself how valuable a company would be who taxes the majority of that industry? 1%, 2%, 5% — it doesn’t matter because the numbers are so huge — the revenues, profits, profit margins etc.

    Fueling (and profiting off) the Fintech Arms Race

    It’s not controversial to say that technology has become the dominant competitive differentiator in financial services. The big four U.S. banks had technology budgets ranging from $8 billion to $11 billion in 2019. That same year, U.S. fintech companies raised a record $59.8 billion in funding from investors.

    The goals for this money vary based on the companies that are spending it:

    • For banks like Truist and KeyBank, the priority is transforming their organizations from siloed, branch-driven service providers to nimble, digital players.

    • For B2C fintech companies like Chime and Affirm, the priority is continuing to scale their existing businesses up while adding additional products and services to capture a greater percentage of their customers’ attention and wallets.

    • For legacy technology vendors like FIS and Experian, the priority is modernizing their technology stacks (much of which is made up of various acquisitions) into a modern set of software services that will make them the vendor of choice for banks and larger B2C fintech companies.

    • For Banking-as-a-Service (BaaS) companies like Bloom Credit and Alloy, the priority is continuing to scale and effectively compete for the business of banks and B2C fintech companies with their more focused, API-driven software services.

    And sitting in the center of this fintech ecosystem? AWS.

    It facilitates fintech companies’ ability to rapidly build, launch, and scale new products. It facilitates legacy technology vendors’ modernization efforts, as they transition their products from on-prem and private cloud to the public cloud. And it facilitates banks’ digital transformation initiatives, as those organizations attempt to become more agile and data-driven.

    As AWS enables each party in the ecosystem to accelerate their technology innovation cycles, it increases the pressure on the overall ecosystem to move even faster. This pressure to ship leads developers in the ecosystem to rely more heavily on AWS software development tools, more tightly coupling them to AWS.

    The result is an ecosystem that becomes, over time, more and more dependent on AWS, which AWS profits off enormously as more volume running through its infrastructure translates, almost completely into more profit.

    Don’t Forget About Big Tech

    As if the challenges enumerated above weren’t enough, we also need to consider the competitive impact of non-bank technology companies like Google, Facebook, and yes even Amazon.

    In some cases, these companies are are partnering with and/or selling to banks (competing with legacy technology vendors and BaaS providers). In other cases, these companies are competing directly for end financial services customers (against banks and B2C fintech companies). In all cases, the key distinction between these large technology companies and everyone else in the ecosystem is that these technology companies own their own computing infrastructures. So these companies, unlike everyone else, stand to profit off the magic of zero marginal costs.

    And don’t discount the fact that AWS’s corporate parent is one of the big tech companies that is directly competing against banks and fintech companies, across a range of product categories. To another company, the fact that one of its divisions may be actively competing with the clients of a different division might be seen as a problem; a conflict of interest to be carefully and fairly navigated. But that’s not Amazon. That’s not Jeff Bezos, whose most famous quote exemplifies a terrifyingly relentless view of competition.

    Amazon has used its cloud computing arm … to copy and integrate software that other tech companies pioneered. It has given an edge to its own services by making them more convenient to use, burying rival offerings and bundling discounts to make its products less expensive. The moves drive customers toward Amazon while those responsible for the software may not see a cent … Some of the companies have a phrase for what Amazon is doing: strip-mining software. By lifting other people’s innovations, trying to poach their engineers and profiting off what they made, Amazon is choking off the growth of would-be competitors and forcing them to reorient how they do business, the companies said.

    No bank, technology vendor, or fintech startup should, for a second, fool themselves into believing that “your margin is my opportunity” won’t, some day, apply to them too.

    So, What Should You Do?

    Shunning the public cloud entirely is not an option, particularly for fintech startups that are working with limited investment capital. You will likely have to work with AWS (or Microsoft Azure or Google Cloud), but the details of how you work with them, the nuances of your cloud strategy, matter a lot.

    So whether you work for a bank, or a fintech company, or a legacy technology vendor, here are a few rules to keep in mind as you develop and execute your strategy:

    1. Focus on your customer, not the technology. Technology is the dominant competitive differentiator in financial services today, but only in so far as it enables companies to deliver differentiated products and services to their end customers. When companies lose sight of this; when technology becomes the end rather than the means to an end; that’s when you see $900 billion wasted on digital transformation and startups burning through their runways. Leveraging AWS to deliver a strong, highly differentiated customer value proposition won’t eliminate the ‘AWS scale tax’, but it will ensure that you are able to maintain a healthy profit margin. Netflix made $1.86 billion in profit in 2019 despite spending nearly $230 million on AWS.

    2. Live in the long tail. If your business model doesn’t make sense at a small scale, it won’t suddenly start making sense at a larger scale if you are using AWS. Fortunately, there are plenty of opportunities to develop strong value propositions and profitable businesses at a small scale. The internet allows for niche service providers to find and serve niche customer segments. Etsy’s ability to thrive in the age of Amazon proves it. Emerging B2C fintech companies focused on specific audiences — StorkCard designed for parents or Hammock designed for landlords — provide a great roadmap to follow. 

    3. Leverage AWS for experimentation. Let’s be honest, the ability to spin up a new cloud instance and build a fully functioning SaaS product in a matter of weeks is incredibly cool. It has value. It just needs to be ring fenced appropriately. One approach would be to utilize AWS as the foundation for an innovation lab, where the software development tools and speed to market that they enable can be used to rapidly prototype and test new ideas. 

    4. Stay technology agnostic in production. A great corollary to the rule above — make sure that any software that makes it out of the lab and into production isn’t tightly coupled to AWS. This will slow down your developers. It will be frustrating. But it is absolutely essential. Maintaining the freedom to move your software between public cloud providers and even your own private cloud environment is necessary to avoid the long-term impacts of vendor lock-in (lack of pricing power, deteriorating performance, etc.) 

    5. Use a hybrid approach for hosting. Assuming that you are able to stay technology agnostic in production, there is a huge advantage for vendors in being able to deploy your services to a variety of cloud environments. You can offer the advantages of all the different cloud computing providers without being tied to the downsides of any one. You can meet your potential customers where they are most comfortable. Some customers may be committed to AWS for other parts of their infrastructure and are comfortable with the cost-benefit mix AWS offers. Other customers may not need all of the bells and whistles that AWS offers and so may prefer a simpler private cloud implementation. 

    6. Hold vendors and partners to your standard. An under-appreciated risk in working with technology vendors and partners is the possibility of (unintentionally) tightly coupling yourself to AWS through those external companies. I have written and responded to many technology RFPs and, while pretty much all of them asked what cloud environment(s) the service provider used, none of them asked for details on exactly how they used them. Gaining this level of visibility into the cloud strategies of your vendors and partners is critical for ensuring that their strategy is aligned with your own.