Monday 12 March 2018

Chatbot and e-commerce, a winning combination
Fanny Ducrocq in the Journal du Net

How can a chatbot add value to an e-commerce website? We review a lever that could well become essential in coming years.

Chatbots are everywhere these days, presented at the International CES with the Netatmo Smart Home Bot in particular, and the market is booming: we estimate that by 2025, this market could reach a staggering 1.23 billion dollars. One of the last key issues for bots is their inclusion in e-commerce websites.

But what exactly is a chatbot? As you already know, a chatbot gives customers the opportunity to chat in real time, either via a messaging application such as Facebook Messenger or via widgets directly present on the trader's website, as is already the case with Oui.SNCF.

A chatbot is used to guide users in two ways:

  • On the basis of rules defined in a decision tree integrated into the bot, telling it what to answer to such and such a question
  • And by using the bot's artificial intelligence to interpret and analyse the customer's question by progressively learning from discussions to provide increasingly precise answers.

A chatbot can capitalise on these two elements, which can be both useful and complementary for a retail website.
In the first case, a customer looking for a certain product can indeed use the bot to filter products and find the "ideal" product. The Levi's website really proves this by steering its customers throughout their shopping experience to help them find the jeans that match their requirements.

In the second, two cosmetics giants have launched bots via Facebook Messenger: L'Oréal and Sephora. These bots learn all about their customers' beauty routines so as to suggest products in keeping with their routines and requirements.

Consumers have clearly adopted these interactions with chatbots: 47% of consumers say they would buy objects from a chatbot, and 37% of consumers prefer to buy items via a company's Facebook page rather than via its website. (HubSpot, 2017).

A bot can also help to relieve the workload of a retail website's customer service by providing answers to basic questions, and only passing on unidentified or more complex questions to a human operator. This enables the customer service to provide a better service, to take the time to gain a better understanding of problems, and secure customer information passed on by the bot. This mode of operation is in keeping with the desires of customers: 55% of consumers are interested in interacting with a company by using messaging applications to solve a problem. (HubSpot, 2017).

With regard to customer behaviour and data, bots broaden the horizons of e-retailers to better understand and secure the loyalty of their customers. Tools such as Facebook Messenger, Dialogflow by Google, Azure Bot by Microsoft, Amazon Lex or IBM Watson can be used to create conversion-related stats, and to isolate certain forms of data to analyse customer behaviour and pinpoint the most popular product and service-related topics 'discussed' with the bot. All of this put together ensures a comprehensive user experience to:

  • Provide suggestions concerning products in connections with customer purchases or searches, in view of personalising their experience
  • Shorten the purchasing tunnel, and finalise orders more rapidly
  • Manage customers' orders: alter information, schedule a delivery, etc.
  • Inform a customer when a product is unavailable
  • Suggest products on the basis of filters, to be totally in line with customer requirements
  • Provide an after-sales follow-up service

Also, seeing as the link between e-commerce and chatbots is only the tip of the iceberg, we can expect further developments to make the experience even more immersive. With voice assistants such as Apple's Siri, Microsoft's Cortana, Google Assistant and Amazon's Alexa we are moving towards bots capable of having real verbal dialogues with users.

  • Share
  • Facebook
  • Twitter
  • Linkedin
Contact us

SQLI PARIS - Headquarters

SQLI Paris
166 rue Jules Guesde
92300 Levallois-Perret
France
Tel : +33 (0)1 85 64 20 20

TO 2018 232me
32 agences sqli fdv