Commerce’s Original Sin

In 1996, the hit song of the year was Wannabe by the Spice Girls. Grunge was still (kinda) a thing and flannel was everywhere. Mozilla was the most popular browser. Amazon was a year into selling books. Ecommerce was just getting started and companies were putting in new architectures left, right, and center, all built around a novel concept: making the catalog available “on the Web” and selling.

Twenty-seven years later, the Spice Girls have moved on. Grunge is out. Flannel is still hanging on (thankfully; it’s good stuff). Amazon seems to be doing ok. But the commerce catalog? It’s still there, livin’ in the 90’s.

That’s a problem.

Actually, for today’s generation of commerce, that’s THE problem.

Let’s start at the beginning. When eCommerce first started, our vision for what it would become was constrained. “Ecommerce” meant a website. One website. The digital store. It was taking a digital representation of our goods and presenting that to the site. Naturally, when conceiving of this, we fell back on current thinking of the day. Our catalogs were database structures, fixed in their hierarchies and constructs, like most databases. At the time, we had no notion for the fluidity and pace of the Internet. We thought it would act like our physical stores, just “on the Net.”

We took this approach and placed the database at the center of our commerce architectures, whether packaged applications or home-grown systems. Of course, because commerce revolved around the goods we were selling, and the catalog is where we kept the goods, we wrapped everything around the catalog…pricing, promotions, display, inventory, cart. The catalog’s tentacles stretched everywhere and touched everything.

We hit limitations in this structure, of course, as commerce matured. And when we hit these issues, we adapted. Changing the catalog was too hard due to all the ripple effects, so we adapted around it, most especially in pricing and promotions. In effect, all of our “new” promotions types (and eCommerce loves endless lists of possible promotion types) were really work-arounds to the catalog. Because we could not change the catalog, we overrode it elsewhere.

Yet, if we are brutally honest with ourselves, the endless cycle of replatforming in our industry, stems from the fact that we have been fighting symptoms and not the root cause.

Not enough merchandising flexibility? Replatform.

Can’t run multiple sites on a single instance? Replatform.

Not enough database scalability? Replatform.

The list goes on and on.

Every one of these symptoms stem from catalog inflexibility. It’s the original sin of the commerce platform and all platforms have it.

Going a level deeper, the root of the problem is that platform catalogs made two fundamental errors.

The first, the elements of the catalog were fused, as in chemically and mechanically bonded. For commerce, there are three aspects of a product–a trinity–that make it suitable for selling online. The description of the product–imagery and text. The pricing of the product. And the relationships between products and associations, or hierarchies. When these are locked together in immutable relationships, it constrains flexibility. And we got frustrated with lack of merchandising control and replatformed...

The second is that in traditional platforms, all other functions were hard-coded to the catalog. So in effect, changing the catalog structure meant impacting promotions, and search, and cart, and the front end, and inventory. And we got frustrated with lack of innovation and replatformed…

A future blog will look at what a modern catalog would mean if it were to be recreated today (spoiler alert: it has been). In the meantime, 2023’s song of the summer – Fast Car by Luke Combs–goes out to the commerce catalog.

Fun fact: Luke was six when it was invented.