Copyright: City of New York

Put retailers (flush with their 2016 shopping season winnings) and e-commerce tech companies together, and you could have happy fireworks

This year (for the first time ever), we’ve decided to attend the annual National Retail Federation’s Convention and Expo (nicknamed “Retail’s Big Show”) over in New York City.

It’s a really interesting year to be in retail. 2016 was a smashing success for many retailers, both brick and online, with many shoppers springing for last minute deals online, outpacing every year since 2005.

Much of that growth came from online, at the expense of brick-and-mortar stores, with Mastercard forecasting a 19% jump in online sales.

Retailers are definitely listening to this data, which makes for very interesting times at the 2017 Big Show. For the first time ever, every retailer is going to be attending with an e-commerce agenda in mind. Flush with the season’s winnings, they’re building out their war chests, ready to take on 2017 with a slew of e-commerce initiatives.

Here’s some themes we think are going to come up again and again in conversations at #NRF17:

The Amazon effect

Every brick-and-mortar retailer out there is trying to figure out how not to lose market share to Amazon.

Retailer’s don’t have an e-commerce problem, they have an Amazon problem.

The reason is that Amazon Prime, and the launch of Amazon’s own logistics network is putting a serious dent in retailer’s market shares. Last time we checked, Amazon just leased 3 aircraft for its Prime Air programme, with a further 15 orders.

Be afraid. Be very afraid.

That’s crazy. Amazon just launched an entire package delivery logistics network, all in the space of 1 year alone. Imagine what their plans are like for 2017.

Logistics are not the only thorn in retailers’ sides. Its technology. Amazon has tech that not a lot of people have in the industry.

Better product search, better product data management, better data curation tools, better customer tracking, better everything.

If retailers are to take on Amazon, they have to start with their technology.

Good AI can help retail

Artificial intelligence is not a magic pill that can solve all of retail’s problems. It’s not search. Having a funny chatbot isn’t going to magically increase sales.


Deploying AI intelligently and to very specific optimization problems can yield tremendous technology improvements for retailers.

If you haven’t already, check out our thoughts on this space:

There are 2 areas that can be vastly improved with good Artificial Intelligence:

Product Categorization

Good product categorization helps customers look up, search and purchase products much more easily. It allows them to discover new products and help surface up previously hidden products that were miscategorized.

Product categorization, until very recently, was an arcane science, mostly debated in various technical conferences. Many retailers chose (wrongly) to expend efforts in manual categorization, with some limited heuristics to automate the process.

However, the deployment of a good AI-based approach to categorization can yield much higher accuracy with a much lower cost base (an automated approach compared to the current manual process).

This allows retailers to quickly onboard, categorize and process product data at scale

Product Matching and Metadata Consolidation

Many retailers struggle with incomplete product metadata. Not having enough product data to display on your product pages has a huge effect on consumer confidence, and can drastically reduce sales volume.

At Semantics3 we use an AI-driven approach to consolidate product metadata across different sources. Our product matching process identifies identical products sold across different retailers, collects the factual product metadata, and consolidates it to a single product record, with all available product metadata tagged to that product.

Using various lookup methods (like UPCs, product URLs, name, keywords, model part numbers and so on), retailers can pull up a fresh, updated, and consolidated product metadata record, ready to be used in generating product pages automatically.

Brands selling directly to consumers

Today’s Shopify stores are tomorrow’s brands, and today’s brands will be tomorrow’s retailers.

Today’s Shopify store is tomorrow’s brand. Today’s brand will be tomorrow’s retailer.

That’s what we think is the overall movement trend for retail, and retailers can choose to play an important part in this revolution.

A major area where retailers can take advantage of this trend is in putting up marketplaces. Retailers still have an advantage that many brands lack; the sales channel through which consumers can purchase items, and the logistics and product fulfillment infrastructure.

While brands can set up their product websites, engage in a direct-to-consumer relationship, marketplaces can serve as an important channel to engage consumers that don’t typically like to shop outside of their favorite retailers.

On the converse side, brands could take advantage of the e-commerce revolution in quite a few ways:

  1. Minimum Advertising Price compliance: Using a price tracking API like Semantics3, brands can set-up automated MAP violation detection systems to help prevent undercutting and preserve brand value.
  2. Consolidated Product Data: Many brands often times have incomplete product data for their own range of products. Using Semantics3 AI-driven product matching and consolidation, brands can create a standardized product metadata record that they can provide to retailers and other sales channels to be automatically uploaded into product pages
  3. Product Analytics: Using the Semantics3 Product APIs and Semantics3 Analytics, brands would be able to get a birds-eye view of their products, as available on retailers’ websites. We collect a variety of signals like sales channel coverage, price change trends, brand scores, as well as sales volume analytics (especially if the brand sells on Shopify)

That being said….

We’re attending NRF17 from January 15 to 17 — if you’re in town, just drop us a note at and we’ll coordinate a time to meet up!

Let us take you out for a coffee, a beer or more. We’d love to meet up, check-in, sync-up and generally talk shop =)

Interested in meeting up? Just email us at and we’ll set up a time to meet!