Credit Life Magazine

Part 2 of a three-part series on the “Future of Ecommerce Search”

Artificial intelligence. Chatbots. Voice search. Virtual reality. Self-driving cars. Technology in 2017 and beyond sure promises to be exciting for consumers. For tech-centric businesses, these trends can be game-changing for those who adapt and overwhelming for those who struggle to keep up. As with all tech waves, disruption will follow close at heel and winners and losers will be anointed when the dust settles.

In this three-part series, we take a look at one niche in particular that is poised for change — product search in Ecommerce. That is, the process by which consumers discover and purchase products online, and the digital interfaces that they use to express intent. In part 1, we look at how the Ecommerce search experience is likely to evolve for consumers in the coming years. In part 2, we look at the technology that will enable these changes. In part 3, the final installment, we take a look at how these changes will affect the ecosystem of Ecommerce businesses, especially online retailers and the companies that support them.

In part 1, we took a look at how consumer expectations will evolve in the next few years, and drive companies to build natural human-centric sales experiences. This week, let's examine the tech that’s actually going to drive much of this change.

  1. NLP Chat APIs

The first task in bridging the gap between humans and machines is to get machines to understand what humans are really saying. Consider the sentence “I’m looking for blue shoes for my son”. A human reader will understand that the sentence conveys the following:

Goal: PurchaseCustomer: Parent
Target: Male child
Category: Footwear
Color: Blue

The rising tide of NLP APIs (Natural Language Processing APIs) will help digital applications attain this type of understanding. By helping extract intent from conversational text, they will help open the floodgates for more natural human experiences, the kind that we covered in part 1 of this series.

2. Ecommerce Knowledge APIs

Understanding language and intent alone is not sufficient, though.

The next step is to build the machine equivalent of an informed salesperson, i.e., a system that has intricate knowledge of all the products on sale, and the ability to identify specific products that meet the customer’s requirements.

Consider the following product:

Product link

A quick glance at the product page can help a human build a table of facts such as:

Category: HenleysGender: Men
Brand: Neonysweets
Sleeve: Full
Material 1: Cotton
Material 2: Spandex
Buttons: Four
Wash Type: Machine

And infer further attributes, using external knowledge, such as:

Weather: WinterCruelty-free: Yes (Silk Cotton)
Fit: Regular
Durability: Medium

This is the sort of information that an experienced salesperson would (or should) know. Chat APIs will need such information behind the scenes to be of real utility to consumers.

Currently though, most data that powers Ecommerce stores is far too limited and messy to play nicely with this goal.

A range of domain-specific APIs, that bring to the table both knowledge and intelligence, are likely to emerge to tackle this need. These APIs will have deep retail awareness, in addition to human-like understanding of linguistic structure.

Ecommerce retailers will turn to these APIs to bring and keep their databases up to par.

3.Voice and Visual Interfaces

Once the machine equivalent of an expert salesperson has been built, all that’s left is to discover the best way to allow consumers to engage with this salesperson. This interaction will be facilitated through a range of interfaces.

Seen this week at #NRF17

It’s no secret that voice is one of the front-runners in this race. Advancements in voice recognition technology have already made Alexa, Siri, Cortana and “okay Google” part of the day-to-day lives of many consumers.

Demand for mobile-integrations and stand-alone voice devices is only just heating up. Holiday sales for Amazon Echo in 2016 were up 9x compared to the year before, and the trajectory looks set to rise.

We’ve still seen only the tip of the iceberg when it comes to the transformative potential that voice can have.

Five years from now, we might very well look back at present-day versions of these products as toy prototypes compared to what lies in wait.

In particular, we are very optimistic about the role that the addition of visual interfaces to voice-first devices could play in facilitating image or video centric but voice-driven retail experiences.

This hybrid approach could be the game-changer that expands the usage of these voice devices beyond the select niche use-cases that they are currently restricted to.

4. Messaging Apps / Chatbots

Sometimes, you may want to have a more passive shopping experience. Here’s where text messaging through chatbots comes in. Messaging apps are already part of our day-to-day lives, and we use them to have very real human interactions, even without seeing or speaking to the person we’ve interacted with. It’s only logical to assume that this increasingly natural medium of communication will broaden to offer a new kind of shopping experience.

The potential of chat is already apparent for all to see. If the evolution of WeChat in China is anything to go by, then these apps are poised to reach far beyond the basic use-case of sending text messages to contacts. Already, millions of WeChat users use the app to browse fashion and make payments. It’s likely that this trend will amplify and spread worldwide.

Add a chatbot AI component that brings knowledgeable machines into the equation, and you have yourself a recipe for change and growth.


5. Virtual Reality, IoT, Wearables, Augmented Reality and More

VR, AR, home automation and wearable devices each promise to deliver innovative experiences in their own unique ways. For instance, VR could finally help deliver the rich in-store visual experience that retailers have long aspired to replicate through digital channels. This early in the game, it is difficult to definitively predict which technologies thrive and which one fades, especially since they are each at different stages of development and commoditization. What we can say for sure is that since all of them will rely on underlying Chat and Knowledge APIs in an attempt to deliver natural experiences, there is a lot to be gained for companies that can service these trends.


6. Assistive and Predictive Technology

All of these developments will increase the quantity and improve the quality of the data collected by retailers. This will in-turn provide a boost to push-based shopping, that is, shopping experiences that are initiated by machines and not by humans.

Devices will get better at knowing what consumers want, even before consumers have expressed the urge.

Imagine, for instance, receiving an alert from a virtual salesperson that says “it’s grandma’s 80th birthday tomorrow — do you want to send her a greeting card”?

Analytics and advertising companies will, as always, be at the party to help Ecommerce companies capture, wade through and monetize this latest deluge of data.

These technology leaps are likely to shake up the Ecommerce industry and unseat many of the incumbent leaders.

Some companies will get it right, some will fall behind the curve and some will overstep the boundaries of what is acceptable. In the last and final part (coming next week), we explore how these trends will take shape, and how this will affect both small and large businesses.

Like what you read? Before you go ..

Sign-up for a demo on our website: www.semantics3.comWe have several Ecommerce Knowledge, Data and AI APIs in private beta at the moment. If you want early access to what we have in store, schedule a call with us.
Semantics3 is hiring AI engineers in Bengaluru. If you’re interested, email us at
If you have thoughts on these trends that you’d like to share, drop me a note.

Semantics3 operates the world’s largest Ecommerce product database. We’re a trusted and reliable provider of ready-to-use structured Ecommerce product pricing and metadata, with coverage on all of the top 800 internet retailers.

Written by Govind Chandrasekhar and the Semantics3 Team in Bengaluru, Singapore, and San Francisco