Leveraging artificial intelligence to make e-commerce platforms more accessible to consumers

Virtual Assistants have been there for a good time now and have become more popular with Apple, Microsoft and Google introducing their virtual assistants namely, Siri, Cortana, Google Assistant, on their respective mobile platforms.

The usage of smartphones is more prevalent now than ever before. By 2018, over a third of the world’s population is projected to own a smartphone, an estimated total of almost 2.53 billion smartphone users in the world.

Representational Image. Thinkstock

Representational Image. Thinkstock

According to the Internet and Mobile Association of India and market research firm IMRB International, the number of Internet users in India is expected to reach 450-465 million this year. India has over 300 million smartphone users which has surpassed the US to become the second largest smartphone market in the world. This is good news for the e-commerce companies.

By one estimate, the Gross Merchandise Value (GMV) sold by e-commerce companies in India is expected to grow to around $80 billion by year 2020. To cater to customers from various demographics, e-commerce companies are focusing on enhancing their customer touch-points.

Use of Virtual Assistants in e-commerce apps
Virtual Assistants are increasingly making inroads into our interactions with Information Systems. However, their usage, has not caught up with the e-commerce apps despite its huge potential to make their platforms more user-friendly and thus, increasing their customer base.

In the current state, the process of searching for a product is quite cumbersome. It requires users to enter the right keywords, play around with multiple filters to be able to finally find a specific product which the user would be interested in buying. The use of Virtual Assistants can bring ease to user experience and significantly cut down the time it takes to move from search to cart, ie find and select a product without having to worry about the classifications, tags and filters related to it.

Imagine if the user is able to put in a search with a query like this

‘Show me mobile phones with RAM greater than 2 GB, with rear camera greater than 8 MP and ignore XYZ brand of telecom industry

While we wonder if this is for real, The North Face is already using a virtual assistant for its store as well as online shopping platforms. Also known as the Expert Personal Shopper, the app lets customers find particular articles of clothing by asking the underlying IBM Watson questions about their shopping needs and travel plans. The technology seeks to transform the filter-dependent e-commerce into a more dialogue-based customer driven shopping experience.

The North Face's Expert Shopping Platform

The North Face's Expert Personal Shopper

For instance, if a person wants to buy a jacket for his upcoming trekking trip to Kashmir in the month of October, the app can quickly search over the internet about the weather conditions including temperature, rain/fog and wind conditions in Kashmir in October. It can then suggest the customer that October would have mild rains as well as would be a little windy. In such a scenario, the recommended product is a Triclimate hiking jacket which has a breathable fabric, is rainproof as well as windproof. It can also suggest the right pair of hiking boots which would be usable in such weather conditions and terrains.

With support to process the user request entered in natural language (as used by people when they talk to each other), Virtual Assistants can cut down the number of steps required to filter out the various features/brands and bring up the right set of products the user is interested in. With additional use of Speech-to-Text and Text-to-Speech interface, the user experience can be further enhanced and significantly cut down the need for the user to be aware of the user workflow for the apps.

Technology Stack
The Artificial Intelligence technology stack has matured enough to enable the e-commerce companies to bring features like these on their platforms. It is no longer the domain which is only for the likes of ‘Google’ and ‘Microsoft’.

The following diagram provides a high-level view of the functional blocks required to build a fully functional intelligent bot which can understand the user and provide what they are looking for.

Virtual Assistant functional building blocks

Virtual Assistant functional building blocks

The following is a sample tech stack that can be used to build these functional blocks. The overall landscape is much more diverse and there are many more commercial and open source products available to fulfill our need.

Technology stack for functional blocks

Technology stack for functional blocks

Now is the time
The e-commerce landscape is exploding with the world players joining the battle with home grown startups. With each day, we see new schemes, sales tactics from these players to increase their customer base.

The use of Artificial Intelligence can bring a huge advantage to the early movers. We expect to see lot more activity from the competitors especially the startups who are usually proactive in introducing radical changes in their existing business model.

The article has been jointly written by Deepak Saini (SME - Artificial Intelligence), Anuj Saini (Solutions Architect), Pratiksha Sharma (SME - Search Technologies) at Sapient Consulting.


Published Date: Sep 08, 2017 03:34 pm | Updated Date: Sep 08, 2017 03:34 pm



Also See