Dreamforce 2018 : Artificial intelligence at every level in e-commerce

Product Recommendations, Merchandising Optimization, Customized Display, Easy Product Search … Salesforce leverages features powered by its Einstein Artificial Intelligence technology to enhance the online shopping experience, especially mobile shopping.

Salesforce keeps repeating: its artificial intelligence technology IoT Einstein is not found in a single product but must spread and feed all applications of its ecosystem … Commerce Cloud, its e-commerce platform, n do not escape the rule. This was demonstrated by Gordon Evans, vice president of product marketing for Commerce Cloud, whom we met at Dreamforce, the annual conference of the customer relationship giant currently being held in San Francisco.

As a result of Demandware’s takeover in June 2016, the Commerce Cloud platform is now used by 2750 merchant sites across 50 countries. 350 million buyers connect each month through these sites and this figure is expected to rise to 500 million over the holiday season. In 2017, the amount of goods and services sold on these merchant sites reached $ 21 billion. An amount up 35% over the previous year.


Doped Conversion Rate

Several artificial intelligence features are already integrated into Commerce Cloud. This is the case of the product recommendation (Einstein Product Prediction) which makes it possible to show in suggestions the products that would be most likely to interest a user. With this brick, the outdoor sports brand Black Diamond would have increased by nearly 10% the conversion rate of its commercial site and would have seen an increase of more than 15% of its turnover per visitor.

The Einstein Commerce Insight module allows you to better understand which products are bought together by Internet users to enable the brand or the merchant to optimize online and in-store merchandising actions. Salesforce has also introduced two new search tools powered by Einstein and whose deployment in Commerce Cloud is scheduled for the first quarter of 2018. These are Einstein Search Dictionaries and Einstein Search Prediction.

Smarter Search Engines

The first allows a retailer to automate the addition of the most relevant terms to his search dictionary. “When you’re looking for a winter blue parka on a website, you can use other words to make the same query, like” coat “or” turquoise. “And, on the merchant side, it’s hard to make sure to have indexed all the terms that users could choose, so we can use Einstein to analyze the terms used and automate their addition, “says Gordon Evans. “We believe this feature is particularly important for brands and retailers with an international strategy,” he continues.

The Search Prediction tool allows you to automatically adjust the search suggestion made when a user starts typing his query in the search bar based on a battery of data (customer knowledge, online behavior, geolocation, etc.) Thus, if a user located in the United Kingdom starts his search on a ready-to-wear site with the letter “s”, he may be offered the word “sweater”. On the same site, a user located in California also typing “s” will be offered the word “short”. “These are simple but effective features to get the most relevant products back up as quickly as possible,” said the vice president of product marketing.

The New Demands of m-Commerce

An imperative that is becoming more and more pressing for brands as shopping habits shift to mobile. “58% of the audience of sites that support Commerce Cloud comes from mobile devices,” confirms Gordon Evans . To adapt to these new uses, the UGG boots brand uses Einstein to prevent the user from scrubbing too long to find the shoes they are looking for. Based on a series of information, the tool makes it possible to go back to the top of the list the products deemed most relevant for each user. The integration of this feature would have allowed UGG to increase its turnover per visitor by 6%.

With the ubiquity of social media photos and their key role in the buying act, it is very likely that Salesforce will integrate a turnkey computer vision brick into Commerce Cloud in the near future to allow Internet users a visual search. Today, computer vision technologies are only available as APIs within Salesforce.