Discover more from Batch Processing
Stripe, Amazon, and The Compounding Nature of Digital Infrastructure
Stripe hasn’t won yet. It’s still day one at Amazon. What the Amazon-Stripe partnership tells us about how digital infrastructure providers grow.
Earlier this week, Stripe announced an expanded partnership with Amazon.
Under the new agreement, Stripe will become a strategic payments partner for Amazon in the US, Europe, and Canada, processing a significant portion of Amazon’s total payments volume across its businesses, including Prime, Audible, Kindle, Amazon Pay, Buy With Prime, and more.
Stripe will expand its use of Amazon Web Services, Stripe’s long-standing cloud infrastructure provider, to run and grow its business while reliably serving millions of internet companies.
In this Batch, I’ll dive into what Stripe and Amazon announced, and what it tells us about the digital payments and cloud services markets.
The Amazon-Stripe Announcement(s)
Let’s start by all admitting that we really have no idea how much of Amazon’s total payments volume (TPV) Stripe will process. Obviously, things can change in the future, but what we can definitively say right now is that Stripe may process some payments for a few services in some regions in which Amazon operates. There are some important caveats to the announcement:
Amazon’s core retail business is not included in the list of services Stripe will be used for.
Stripe is eligible to process “a significant portion of” Amazon’s own subscription products (e.g., Prime, Audible, Kindle). “Subscription services, including Amazon Prime, generated almost $32 billion in revenue in 2021.” That’s impressive but it’s a far cry from the $746B in global retail sales Amazon is expected to do in 2023.
Under the new partnership, Stripe is also eligible to process payments for Amazon’s digital wallets (e.g., Amazon Pay, Buy with Prime) that can be used on non-Amazon websites. Amazon has been expanding Buy with Prime access recently, bringing it out of invite-only mode, and I expect to see both Buy with Prime and Amazon Pay as supported wallets on Stripe shortly (like Adyen does).
Stripe is not Amazon’s sole payment provider.
This seems fairly obvious given caveat #1 but Amazon will continue to use its own in-house payments infrastructure and third-party providers like Adyen, which powers payments for Amazon Japan.
We also know this simply because Amazon offers PayPal’s Venmo digital wallet as a checkout option (among others) in the US and Stripe does not offer any PayPal checkout solutions.
And then there’s Amazon’s Payment Service Provider program, which allows merchants selling on Amazon Marketplace to use one of the dozens of payment service providers that have met Amazon’s standards for identity verification and fraud reduction. Stripe is not one of them.
The partnership is not global.
The announcement states that “Amazon started using Stripe in 2017 to accelerate market expansion in Asia and Europe,” and that the expanded partnership will cover the US, Europe, and Canada. But as Payment Dive’s Lynne Marek points out, the announcement “made no mention of its work for Amazon in Asia.” We do know that Stripe helped Amazon rapidly add BLIK, a payment method preferred by ~60% of Polish buyers.
It seems like Amazon was experimenting with several partners around this time (2017-2020) to see how they could gain access to new markets and payment methods faster. Amazon even bought a payments company—PayFort—in 2017 to bring localized payment processing in-house for the Middle East and North Africa (e.g. Egypt, Saudi Arabia, Kuwait). There was also the aforementioned Adyen deal to power Amazon’s payments in Japan in 2020.
Surely this partnership will help increase Stripe’s total payments volume (TPV) and revenue as they pursue profitability, but that’s kind of beside the point. The announcement itself is more important for Stripe’s long-term prospects than the economics of this specific deal because most people (and importantly buyers of payments at companies big and small) will likely not catch these caveats and will assume Stripe is Amazon’s sole payments service provider. “If Stripe is good enough for Amazon, it’s good enough for me!” And Amazon’s full-throated participation in this announcement was vital. Let’s take a look at what happened last time Stripe shared that they were powering some payments for Amazon. From Ashlee Vance’s 2017 profile of Stripe in Bloomberg Businessweek:
One way to justify [Stripe’s valuation]: Stripe’s new partnership with Amazon.com Inc., the largest and most sought-after customer on the internet. Over the past couple of weeks, Stripe began handling a large, though undisclosed, portion of Amazon’s transactions. Neither company will address the scope of the deal — which was only revealed by Stripe’s addition of Amazon’s logo to its website — but it could help Stripe greatly increase its transaction volume. (Amazon had no comment.)
So at some point in 2017, Stripe secured Amazon—the largest retailer on the planet—as a customer and announced it by ::checks notes:: publishing a .svg file to their home page (?!), and when Amazon was asked about the partnership, they said “no comment.” Yikes! We can assume this was a mistake or that Stripe was, understandably, eager to share the news with the world in an effort to convince larger customers that Stripe was no longer just for startups. From Vance’s piece again:
Stripe is also trying to make deals with Target Corp., Under Armour Inc., and other merchants to snag money available outside the startup scene, partnerships made more possible by the trust Amazon is showing.
That strange episode makes this week’s announcement all the more special for Stripe since they now have AWS’s CEO and VP of Payments on the record saying nice things about Stripe’s reliability, global reach, and “commitment to putting users first.”—characteristics important to enterprise customers.
Stripe Hasn’t Won Yet
Stripe famously started out as a service for developers who needed a quick way to get started with payments without talking to another human being. But that’s changed. Packy McCormick at Not Boring has done the best job explaining the compounding nature of focusing on small customers early and growing with them over time in his piece Stripe: The Internet's Most Undervalued Company:
Stripe similarly uses time horizons as a competitive advantage. It began by serving an overserved [sic] segment of the market -- engineers at startups -- with a product that traded features for simplicity and speed. And it’s grown with them. Like Slack and Snap, Stripe takes advantage of the compounding effects of young users. At an increasing rate, startups become big companies, and young people become decision makers. While incumbents and other competitors focus upmarket, on the most lucrative opportunity in the present, Stripe focuses on compounding over time.
Stripe’s CEO, Patrick Collison, confirms this strategy. From Ashlee Vance’s Bloomberg Businessweek profile again:
Stripe continues to attract startups. It intends to be behind the next Uber or Airbnb, to cash in on its meteoric growth. ‘If you think about the broad trajectory of the internet, most of the breakout successes are still to come,’ Patrick says.
So, from the early days, Stripe has correctly understood that most of the revenue it will get from customers comes in the out-years and it should do everything it can to keep those customers. The strategy seems to be working and plucky startups that started using Stripe a decade ago are now processing billions of dollars a year. In 2021, 60% of tech companies that went public processed at least some of their payments volume using Stripe, and I imagine a large portion of Stripe’s current payment volume comes from customers it acquired early that have now become massive (e.g., Instacart, Shopify). But as these companies grew, their needs became more demanding and complex. Companies like Lyft (First Data) and Postmates (Adyen) added additional payment providers in an effort to lower their payment processing costs. DoorDash and Instacart both selected Marqeta as their card issuing provider. In response, Stripe ramped up its rapid product velocity starting in 2018 with the release of Billing, Terminal, and Issuing.
The 2018-2022 time period is critical to understand Stripe’s enterprise aspirations. During this period, it appears that Stripe’s net-revenue growth rate began to fall and stabilize at (a still impressive) 50-70% YoY.
Stripe was motivated to win new deals big enough to meaningfully add to their revenue AND wait for growing customers to contribute serious revenue. Stripe’s CEO Patrick Collison stated this two-pronged approach plainly in his 2019 Stripe Sessions keynote:
Our strategy is very deliberately to serve both ends of the continuum (startups and enterprises), and every point in between.
This strategy meant going after larger, enterprise customers, which brought Stripe into direct competition with the best-run payments company on the planet: Adyen. Adyen doesn’t serve smaller merchants but they have the same “earned secret” as Stripe (volume under the curve compounds if you are patient) and their customers usually started from a much larger base. Each year 80% of Adyen's payments volume growth comes from its existing customers. Around this same time, Adyen was scaling its partnership with Uber, signing deals with McDonalds and Subway some of the largest quick service restaurants in the world, and luring eBay away from PayPal. So, all of a sudden, Stripe, a payments service provider that had spent the better part of a decade excelling at acquiring small customers by fine-tuning their API documentation found itself losing customers to Adyen because it didn’t have a good enough integration with Microsoft Dynamics, a leading Enterprise Resource Planning tool.
Enterprise customers care about things that startups don’t (or at least not to the same degree): reliability, global reach, and costs. They also care deeply about social proof. An up-and-coming startup might take a risk on a smaller payments company but an enterprise-grade customer will want to know that other large customers are using a provider before signing on with them. Social proof was one of Adyen’s early advantages. The Adyen founders had already built a payments company before (and sold it to the Royal Bank of Scotland) so customers could trust that they knew what they were doing. So, in addition to the new products Stripe launched during this period, they also put a lot of effort into establishing social proof signals that would lead enterprise customers to trust them:
They added a cast of enterprise-grade executives and board members:
Diane Greene, co-founder of VMWare and former CEO of Google Cloud joined the board of directors.
Dhivya Suryadevara, former CFO at General Motors became Stripe’s CFO.
Mike Clayville, former Vice President-Worldwide Commercial Sales and Business Development at Amazon Web Services became Stripe’s CRO.
Christa Davies, EVP and CFO at London-based Aon and former CFO for Microsoft’s Platform and Services Division, joined the board in 2021.
Mark Carney, former Governor of the Bank of England and the Bank of Canada, joined the company’s board of directors
They signed and advertised enterprise-grade customers:
Ford signed a 5-year deal to use Stripe for their corporate and dealership e-commerce and payments experience
Atlassian plans to use Stripe for global billing and payments across all of its cloud product offerings
Salesforce now offers Commerce Cloud Payments, an out-of-the-box payment solution powered by Stripe (you can use other payment providers like Adyen and Global Payments with Salesforce Commerce Cloud, but they’re not pre-integrated).
They more than doubled the countries where they offered local merchant acquiring services from 20 in 2017 to 46 in 2021.
This brings us back to the importance of the Amazon-Stripe announcement. Stripe gets to continue touting its relationship with the most enterprising enterprise customer imaginable as they continue to move upmarket and battle Adyen. If you take Stripe’s multi-decade vision seriously, then investing in a relationship with Amazon, even if it’s not Amazon’s full payments volume and even if it’s with thin margins (no idea if that’s the case) is worth it because it allows them to acquire and keep large customers in the future, many of which are still to come.
It’s Still Day One at Amazon
What of Amazon’s motivations to announce a partnership with Stripe now, when they had previously been reticent? The answer is likely similar to Stripe’s, both in terms of the long-term nature of the bet and the competitive dynamics. First, Amazon Web Services (AWS) gets to lock in a fast-growing enterprise customer of their own in Stripe as they fend off Microsoft’s Azure and Google Cloud Platform (GPC) in the Cloud Wars.
I say “lock-in” because other tech companies with huge and growing cloud expenses (e.g., Pinterest, Snap) have struck deals with cloud infrastructure providers in recent years that commit them to hundreds of millions of dollars in guaranteed annual payments in exchange for per-event pricing discounts. Other AWS customers, like Spotify and GitLab, are being lured away by GCP to save money. These companies (like payments companies) care about profitability and I expect that trend to continue as the technology sector suffers through a recession in 2023 and 2024. I’d be very surprised if Stripe didn’t use the threat of migrating to Azure or GCP as leverage to lock in lower rates and get Amazon to agree to participate in their announcement.
The second reason I think Amazon was interested in announcing its partnership with Stripe is that lots and lots of startups use and emulate Stripe. AWS, like Stripe, understands the compounding nature of digital infrastructure and tries to work with as many companies as possible from day one. As we discussed earlier, early-stage startups eventually grow into large, successful businesses. From Geoff Colvin's November 2022 profile of AWS for Fortune:
Instead of raising millions of dollars to buy servers and build data centers, startups could now get online with a credit card, and pay a monthly bill for just the computing power and storage they used. If their new app was a hit, they could immediately engage all the cloud services that they needed. If it bombed, they weren’t stuck with rooms of junk equipment.
Realizing this, AWS began giving away credits to up-and-coming companies and lowering their pricing repeatedly (for all companies) to lock them in for the long term. Today, they have AWS for Startups, which offers a whole suite of services to promising young companies. A partnership with Stripe could reestablish AWS as the standard web services provider for startups that have been courted more aggressively by GPC in recent years.
The Compounding Nature of Digital Infrastructure
A few years ago, I wrote about cloud providers, like AWS, moving up the stack into the application layer and directly competing with companies like Twilio. As I wrote in Twilio & Tin Plates:
In early 1899, JP Morgan rolled up over 85% of the finished steel manufacturers in the country (producers of metal pipes, wire, and tin plates) into the National Tube Company, then struck deals to source National Tube’s primary steel exclusively from competitors of Carnegie Steel, the largest primary steel manufacturer in the country at the time. This maneuver stole market share from Andrew Carnegie’s core business and he responded by forming a finished steel manufacturer of his own, which could always sell its finished products for less than National Tube. From Charles R. Morris’ The Tycoons:
Given the relatively modest investment required to enter most finished steel businesses, it would always be easier for the primary steel companies to integrate forward into wire, hoops, or tubes. And by using their new finished lines to sop up surplus primary steel capacity, they would have the luxury of selling below cost and killing off independents at will.
In our modern-day web services analogy, AWS is Carnegie Steel, Microsoft Azure and Google Cloud Platform are its competitors, and Twilio is a tin plate manufacturer. Twilio can’t integrate backward into a wholesale web service provider. And if enough customers like Spotify and Apple continue to leave AWS for Google Cloud Platform or Azure, we might see AWS start to go on the offensive.
I was wrong (at least so far). Companies like Twilio, Snowflake, and others have experienced tremendous growth and success in the “application layer” even as cloud infrastructure providers offer directly competitive products. Part of the reason is that the cloud infrastructure market is so big and growing so quickly. From Fierce Telecom’s piece on Amazon’s Q1 2022 earnings:
While rivals are gaining market share little by little, AWS continues to be buoyed by the ongoing growth of the overall market. In the first quarter cloud infrastructure spending by enterprises reached almost $53 billion, a 34% increase over the same quarter in 2021, according to Synergy, and marking the eleventh time in 12 quarters that the year-over-year growth rate has been in the 34% to 40% range.
Although smaller in scale, the digital payments market has a similar dynamic. A 2021 Credit Suisse report forecasted modern payment processors (Adyen, Stripe, and Checkout.com) to grow at a ~40% CAGR through 2026. Categories with billions of dollars of quarterly revenue growing at 30-40% YoY is absolutely insane. The fact that there’s so much revenue to go around means that digital infrastructure providers like Stripe and AWS are better off staying in their lane (fighting to maintain market share and acquiring new customers) rather than trying to move up the stack. And the growth of the overall market will continue to lift them to greater and greater heights.
*Disclosures: I own shares of Adyen, PayPal, and Stripe. I am also dependent on Stripe for revenue as they are the sole payments provider for Substack, which I use to publish Batch Processing.
Cutoff Time is a section of Batch Processing that includes interesting links to news or ideas that caught my eye but decided not to write about in this batch:
On LinkedIn, Michael Liquornik pointed out some important caveats to my take on Checkout.com in my latest piece about payments and profitability: Checkout may have OPEX centralized in the UK that is not earning revenue in that same market.
This batch was powered by: