%201.svg)
Ecommerce Evaluation Results
July 17, 2018
Foreword
We want to address whether or not these ecommerce frameworks provide not only the basic functionality, but are also capable of serving these growing trends. An ecommerce framework or application that is stable, reliable, consistent, and simple could make the difference in improving customer experience, driving sales opportunities, increasing organisational agility, and the overall functionality and services offered through it. What functionality do I need in my ecommerce application so that it can respond to these trends? We know there’s more out there than enterprise ecommerce solutions (the Big Five). We want to be at the frontlines, the moment before the CIO makes that fateful decision that will determine the next steps of his/her organisation’s future. Our goal with the evaluation is to illustrate opportunities that go ignored!
Why do we care?
The world of software development is changing, improving, and adapting to new conditions and approaches that have challenged our previously held paradigms of software development. Having witnessed years of software development and its changes, we have come together to stand at the frontiers of software development in order to open those doors and inspire our clients towards future-oriented solutions and an out-of-the-box approach, to realise a world beyond the monolith.
We want to present clients and customers a perspective that doesn’t rely solely on enterprise solutions, or on a generic solution just because it’s “easier”. What we want to show clients is an opportunistic approach that better reflects business intent than a large, complex solution involving years of modification and customisation before it can actually be used. Our idea is: why not develop a solution based on the highest possible development standards while offering B2B functionality, catering to your business users’ needs while avoiding unnecessary application clutter and performance issues?
You can still have it all without sacrificing your ecommerce software architecture
At Sly AG we develop software using microservice architecture. While service-oriented architecture, offered by the mainstream Big Five, might be the preferred approach for some, an alternative path exists: one that reduces complexity and cost, improves maintainability and resilience, and results in a solution that is not as heavyweight but equally powerful.
Reviewing results
From our analysis and experience with the material these ecommerce applications provided, and our review of the functions they offer, we created the following diagrams to illustrate our findings.
Using Pearson’s correlation, we identified the z-score points of those ecommerce applications and plotted them on the quadrant. This allowed us to normalise the distribution and to understand the distance between one point and another, and their relatedness. Pearson’s correlation allows us to standardise the variables and their distribution between -1 and +1. However, because of the various functionalities reported (and those that aren’t reported by the ecommerce provider), the distribution shows some outliers. The distance between the variables is quite large, which could be due to the amount of information available for rating each solution. Some have extensive documentation and functionality descriptions (Oracle, Elastic Path, and Hybris), whereas others not so much (Squarespace, Shopify, Moltin).
With millions of dollars invested in software architecture and development, the Big Five have reached a state of great complexity while offering high functionality to go along with it. Our findings consistently show that the Big Five have held solid ground and built great traction and momentum for their solutions. With the impending rise of microservice approaches, we are finding that few of them are taking the leap and modifying their current architecture to suit the changing paradigm. The Big Five are mostly SOA-oriented, with the purpose of leveraging the technical power built into their layers. They are partly microservice-oriented as well, having single-piece functionalities that can, technically, be deployed in isolation from the rest; however, their tight coupling makes it difficult to trace causes and build updates for those pieces.
The Big Five are moving towards greater system autonomy by encouraging the use of microservices and even opening the environment for developers to build them (for example, Hybris YaaS, a marketplace providing microservice applications for Hybris solutions). Others have made big leaps in developing microservices from the get-go, like Elastic Path, which (for better or worse) leverages microservice-oriented functionality despite a single database. Unfortunately, not all of the functionality can be turned off.
What about the others? The others are closer to our preferred solution: microservice-based. We rate them as having high functionality because they are flexible enough to be developed accordingly. Opposite the Big Five are solutions unknown to many software developers: commercetools, React Commerce, Unchained, and Cezerin.
Our analysis and ratings showed that high-cost players traditionally offer most functionality out of the box, ready for use. The challenge is to deactivate all functionality that is irrelevant to your business; otherwise, those functions may keep operating in the background and make unnecessary calls that inevitably produce exceptions, since there is no data to retrieve. After many hours of consultation and development work, those functionalities may be decoupled from the monolithic structure and thereby deactivated (be careful that others are not affected by this decoupling). At times, decoupling may not even be necessary, as the Big Five may provide a toggle that lets you deactivate and activate functionality as needed. Nevertheless, those lines of code are still there, and stripping them (after toggling them off) might harm other functionalities that depend on them. Oftentimes it’s difficult to discern where functionalities begin and end; the dependencies can be too great and the risk too high when deactivating components that are unnecessary for your business.
Out-of-the-box solutions might sound promising at first glance, but we know all too well the cost of slimming them down and simplifying processes so that customers’ needs are at the forefront, rather than functionalities that customers typically don’t need.
Is it cheaper in the long run to have an out-of-the-box solution, or is it more expensive to customise it and strip it down to your requirements?
At the opposite end are those that can provide higher functionality at lower cost. This is because of their flexible development, which allows functionality to be integrated in the same way, but geared to the needs of the customer. React Commerce, Cezerin, Nemesis, Broadleaf, and Unchained provide very few out-of-the-box functions, and they are nowhere near as comprehensive as Oracle Commerce, Intershop, and Hybris. This is intentional, because a bloated enterprise solution might not end up meeting your needs: it’s easier to create a tailored solution than it is to strip one down and then customise it for your needs.
