Eliminating Shopper Fatigue: Making Online Commerce Faster, More Accurate

Say you’re working on a project that calls for something specific, like a “one-quarter-inch slotted magnetic screwdriver.” But when you go to the website of a hardware chain and type that phrase into the search field, zero results appear. You sigh and search again, typing simply “screwdriver.” This time, 652 matches fill your screen. You sigh again, frustrated by the “Goldilocks experience” so common to online searches—too little information or too much information, never resulting in the desired outcome.

No wonder a recent study found more than one-third of online shoppers give up browsing after 10 minutes, exasperated by difficult search tools, an overload of options and a dearth of offerings personalized to them.

An Enterprise NoSQL Database provides the necessary horsepower to accelerate development.

Traditional relational databases, the technology many e-commerce websites were built on, were designed for a different time, when shoppers were content to just find product types, not products with specific attributes. These relational databases aren’t equipped to deal with massive amounts of unstructured data, such as product descriptions from suppliers’ catalogues — which are necessary to make searches more precise.

Indeed, even the customer service reps at many e-commerce sites flip through paper catalogues because their own search engines are too slow and unreliable.

Codify, a provider of product content services, found a solution to this dilemma. The Chicago-based company helps online retailers make their products easier to find on web searches by leveraging MarkLogic’s Enterprise NoSQL technology, a crucial platform for retailers handling the explosion of vast amounts of data in many different forms.

The beauty of NoSQL is that it’s a schema-agnostic data model that ingests data in whatever its current form. Codifyd uses MarkLogic to quickly and reliably merge millions of data points from thousands of suppliers into a product catalogue for each of its clients.

By gathering such fine-tuned information instantaneously, Codifyd recommends products matched to specific attributes in real time, increasing customer trust, loyalty and retention. This more precise information also allows retailers to bundle relevant product offers in a set, improving upselling and increasing the average order size. For example, a retailer can serve up the “one-quarter-inch slotted magnetic screwdriver” the customers searched for as well as a toolkit that contains that particular screwdriver.

NoSQL allows retailers to move quicker in other ways, too. Today, it takes six months to put data for new products on a distributor’s website. Codifyd has been able to reduce that to two weeks since the NoSQL approach allows the data to be ingested in its existing form. For one customer, Codifyd was able to add 660,000 new SKUs to its product catalogue in a single hour, increasing online sales by 5 percent—an eye-popping amount in an industry with razor-thin margins.

And that, as Goldilocks would say, is just right.