The concept of commerce as a “platform of platforms” stems from the need to provide a platform of internet-level commerce services to support trade on any other platform. With a commerce platform-as-a-service (PaaS), companies can focus on expanding their core IP and building differentiating features instead of building commerce services such as product management, order management, shopping cart and checkout.
Popular examples of platforms of platforms that deliver internet-level services in areas outside of commerce are AWS, Okta, and Stripe. With these platforms, companies can focus on their core IP instead of managing technical infrastructure and services.
Through AWS, Amazon offers infrastructure-as-a-service (IaaS) for computing, storage and networking, while Okta offers identity-as-a-service (IDaaS) that companies use to offer simple and secure sign-on. The Stripe platform is a payment service that plugs into other platforms like Shopify, so Shopify and small businesses using the platform can accept payments. (My company also uses Stripe to power the payment product within our commerce platform, making Stripe a payment platform of platforms for us.)
The point is: You don’t build a payment gateway to start accepting payments, just like you don’t build a commerce platform to start conducting trade. Instead, you plug into a platform of platforms to perform essential activities that are not unique to your business.
In addition to being convenient, a platform of platforms is also flexible. Services can be used independently and plugged into any business model. Further, companies using the services don’t have to manage the underlying infrastructure as this is managed by the platform. But, until recently, there has been no platform of commerce services like there has been a platform for web services, identity services, and payment services.
Commerce as a domain in this platform of platforms paradigm got missed because building a platform of platforms is hard. Consider Stripe, a company that started in 2009 and has just started gaining major traction. The platform did not evolve over the course of a couple of years; it took over a decade’s worth of work from visionaries and engineering talent that believed in that vision.
Compare Stripe that focuses strictly on payments with a commerce platform that must focus on product management, inventory management, pricing, promotions, subscriptions, loyalty and customer experience. This is a significant undertaking that many software providers only take on pieces of. A few examples include Salsify for product management, ReCharge for subscriptions and Bloomreach for experience.
To date, companies have had to use third-party services like this alongside custom-built services and monolithic platforms.
Take Roblox, an online game platform that achieves monetization by letting users create and sell games to other users. When it launched in 2005, there were third-party e-commerce platforms available but they were not platforms of platforms; they were monolithic platforms and not flexible and adaptable to cutting-edge business cases like their own. Just like they needed the AWS platform to support internet infrastructure, Roblox needed a commerce platform to support commerce infrastructure. They developed a homegrown e-commerce platform, but other companies might rather focus on other lines of business than to divert resources to develop their own.
Even if Roblox had wanted to look elsewhere, no service provider was up to the challenge — but this is starting to change. There is now demand for a commerce platform of platforms, much like there was demand for a web services platform of platforms when AWS launched its IT infrastructure services in 2006.
New commerce software platforms and initiatives have appeared over the past few years (including my own) attempting to build scalable commerce services. The issue is that many are too focused on the headless commerce trend and missing out on the big picture. Headless commerce directs the frontend customer experience, but not internet-level services, which is what’s needed as 77% of software engineers, including those at Roblox, move to service-oriented software architectures.
To fill the commerce platform of platforms gap and support a variety of use cases, a platform must be both headless and modular (i.e. service-oriented). In addition to supporting common use cases for D2C, B2C and B2B commerce, a platform must support emerging use cases on gaming platforms, digital publishing platforms and consumer-to-consumer (C2C) platforms. In short, the commerce platform of platform must be agnostic to the desired use case or business model.
In addition to meeting requirements for flexibility, the platform must also scale. For example, with traditional commerce platforms, open source e-commerce software, and custom-built e-commerce solutions, large enterprises and high-growth brands must plan months in advance to support holiday shopping that occurs online. And, even after this planning, e-commerce sites from Costco, Home Depot, H&M and Nordstrom Rack still go down, costing up to $11 million in sales.
With cloud-native services that a true commerce platform of platforms is composed of, scalability is no longer an issue, leaving time for developers and engineers to build business-differentiating features, not e-commerce infrastructure.
Overall, this is a very new space with many new entrants and choosing a commerce platform of platforms could quickly get confusing. A simple way to test a platform is to answer these questions:
Choosing a commerce platform that checks these boxes will serve you well as you build new e-commerce features, expand existing features or migrate from legacy commerce technology.
This article was originally published on Forbes.
Board of Directors @ fabric. Previously @ Google, Amazon, Staples, eBay, and Groupon.