by rattattart

Couple of weeks ago we embarked on our annual journey to the various pow-wows of frantic networking known as industry conferences.

Industry conferences can be dreary affairs — if you don’t know what you’re doing. Behind the facade of small talk and booth shwag, there’s a tonne of insights not immediately apparent:

At Semantics3, we typically focus on a couple of conferences a year — we’re a retail product data start-up, so retail conferences (especially e-commerce) tend to be really productive for us.

The best part about these conferences is that being a data-centric company, almost everyone attending could potentially benefit from the data.

Its also a great time for self-discovery, a chance for us to reconnect with customers, engage new prospects, and most importantly, to study new trends in the space.

Conferences can be great like that.

This year we decided to check out 2 major conferences — GS1Connect and IRCE 2016. Both remarkably different experiences, but equally interesting.

For this post, we’re going to look at IRCE.

There were quite a few e-commerce trends that came out at us at IRCE, but one particular trend stood out:


Traditionally, the online shopping cycle for consumers has been along the lines of:

  1. Go to favorite online shopping website
  2. Consumer not sure what s/he wants and starts browsing
  3. Hopes to find something that piques interest
  4. Nothing found
  5. Leaves site
  6. Retailer loses revenue :(

However, the goal for the retailer is to have the consumer:

  1. Go to the retailer’s website
  2. On front page, get shown a personalized curation of products, some of which the consumer adds to cart instantly
  3. Search for a product, finds exactly what s/he’s looking for and adds to cart
  4. Consumer gets recommended additional relevant items
  5. Adds more to cart and checks out
  6. Retailer profits :D

In the olden days of… just a few years ago, online shopping had always been about price (e.g., Amazon).

Now retailers — both online and brick-and-mortar, have become well aware of the benefits of tailoring to their customers needs in order to improve engagement and higher conversion.

More so than ever, online shoppers care that products shown should be exactly what they were looking for: trendy, curated and conveniently arrive in their mailbox rather quickly.

If you haven’t heard, the new generation of shoppers doesn’t demand more choices, they demand choices that fit their unique profiles (sound similar to a previous blog post we wrote?).

And the best way to meet this demand? Personalization.

Personalization comes in several shapes and sizes

The practice of personalizing shopping experiences isn’t particularly new.

Companies like Barilliance, RichRelevance, Reflektion, Certona, and more have been offering e-commerce personalization platforms and services for a while.

It’s just that these companies and retailers themselves are getting better and better at optimizing what products, suggestions, and offers should be shown to shoppers at a segment-by-segment or a customer-by-customer level.

Additionally, any part of the shopping experience can be personalized. Here are some that you may have already seen, but not necessarily noticed:

Customized front pages:

Credit: John Wright at Shopi-SEO

Product recommendations on product pages:

Tailored offers via e-mails:

Credit: Ryan Beard at Client Heartbeat

Personalized search results:

Other channels that can be personalized: carts, mobile, display ads, push notifications, text messages

The Sexy Stuff Happening

What goes into these personalization and recommendation engines? Algorithms, machine learning, artificial intelligence, predictive analytics, social listening, segmentation and other sexy buzzwords (and spying on your browser history).

However, the real truth is that without product data, none of this would happen.

Machine learning, AI, analytics, etc. make up just the processing portion of what goes into to make recommendation platforms and engines. But what good are all these processes if there were no inputs?

Rich product metadata provides the key inputs necessary for e-commerce personalization. And the more data, the better.

The basis of recommendation and personalization is around personality, preference, taste, and emotions, and these characteristics are tied to the features of products and the purchase as the whole. Such features include color, size, weight, functionality, connectivity, category, speed, comfort, availability, shipping, etc.

Pricing also matters not only because it changes perceptions, but also because shoppers still have budgets. Additionally, as image recognition engines become more powerful and precise, images make an impact on personalization as well.

The problem here, though, is that retailers and their personalization partners often times are deficient in product metadata needed for highly effective personalization.

E-commerce Data Solutions Save the Day

Personalized shopping can be pretty lucrative. We work with many companies who have built sustainable business models on serving up personalized choices to their customers.

Semantics3 offers the data needed to exploit the current state of retail personalization and product recommendations — and help you build a truly unique business model.

We augment the product catalogs that feed into personalization platform, with tools that function — even as simply as just by providing UPCs, urls, or product titles.

With rich product metadata and fresh pricing, there’s no limit to what you can build!

Like what you read? Let us know!

Interested in our data solutions? Book a call with us today!

Lovingly built in San Francisco by Calvin Chang and the Semantics3 Team.