I attended some of the sessions this week at Modern Commerce Day (presented by CommerceTools). There were interesting sessions from companies such as how Lego has modernized their ecommerce site using serverless patterns, Express.com on migrating to a new architecture during COVID (online sales were their sole revenue stream due to physical stores being closed), but I wanted to focus on the session by Mike Lowndes (Sr Director Analyst at the Commerce Technologies and Experience Team at Gartner) talking about Composable Commerce.
I found myself agreeing with pretty well everything he said.
First, Mike introduced the transition of e-commerce to Digital Commerce (website, to mobile first, to “multi-experience” where there are many customer touchpoints both online and in the physical world). The Express example above is a perfect case study of this pattern. They have fully embraced the unification of all surfaces of interaction. Shipping due to a product being out of stock in a store or from an online purchase is the same shipping and tracking, etc. It is one business.
Mike also summarized industry megatrends of SaaS, cloud technologies (AWS, GCP, Azure, etc), frontend, business agility, agile methodologies, DevOps, etc. A key point was that “build or buy” is the wrong way of looking at things. Today it is “build and buy”. If you want to differentiate yourself, think about what in your experience needs to be different to your competitors. This is an all-of-business decision, not just a website feature review. You don’t want to pay for building features that are not differentiators – it is much cheaper and safer to use an existing solution. You only want to consider building code if it gives you an edge. (And if you don’t need or cannot afford differentiation, then stick with an existing ecommerce platform.)
This of course is the heart of Composable Commerce. Pick a suit of best-of-breed SaaS services that match your business needs and use them together. If you need special capabilities in part of the solution then build that out yourself.
Commerce capabilities (carts, product catalog, etc) are becoming commoditized. Differentiation is more and more in the customer experience, and hence the rise of DXP (Digital Experience Platforms). Webstores are a dime a dozen. Experience orchestration, including personalization, is on the rise.
It is interesting to me in terms of architecture that there are well known patterns for backend componentization: web APIs, webhooks, eventing with publisher/subscriber queues, and so forth. Within a single system (Mike used the term “Packaged Business Capability”) you have microservices supported by technologies such as Kubernetes or serverless programming models. Of course Jamstack and MACH got mentions.
Mike then touched upon “Application Composition Technologies”. Another session at the conference was a good example of this, a discussion on work between VueStorefront and Bloomreach. They announced a new “coming soon” feature of allowing marketers to compose pages using Bloomreach that contained VueStorefront components. This allows developers to build components, while leaving page composition to marketers.
I am seeing this model more and more often, and I like it. It addresses issues such as how to do A/B testing. As soon as you need developers to assist with A/B testing, it will probably be too slow (developers are a precious resource). It is better to let your marketers do final page composition and A/B testing of different layouts. This is a much more logical separation of responsibilities between developers and marketers.
I do think the frontend side of the story from an architectural perspective is less mature than the backend. How to incorporate a new loyalty program into your frontend? There can be one API defined by a SaaS service, but there is not one frontend integration pattern. Most likely you are going to need some degree of coding here from your developers, at least for now.
Mike ended up with his presentation that he believes true Composable Commerce is not here today, but he expects it to improve over the coming years. He listed numerous benefits to businesses adopting the model.
Mike of course made a lot of other points during his presentation. I cannot say that I learnt anything new personally. But it all made sense to me and helped position the Composable Commerce story. If Shopify, BigCommerce, Salesforce, etc meet your needs then you probably don’t need a Composable Commerce story. But the more your business grows in sophistication and moves towards Digital Commerce, the more relevant the Composable Commerce story becomes.