Machine Learning
Creative Intelligence, combining the black box with human ingenuity

 

Discover how AI and Machine Learning is redefining ecommerce experiences.

With almost two decades of ecommerce experience behind our technology and people, we've watched the volume of data available to retailers explode in a way nearly impossible to predict.  

The advent of Artificial Intelligence (AI) technology is providing smart automation that is transforming how retail customers manage their shopper experiences.

 
 

FAST FACT

Did you know...

40,000

Search queries are performed per second (on Google alone)

 
 

What is AI? 

Artificial Intelligence is best defined as the ability of a machine to perform the kinds of cognitive functions we typically associate with humans, such as learning, reasoning and perceiving.   

Machine Learning is a subset of Artificial Intelligence and has had the most impact on ecommerce to date. Machine Learning algorithms are applied to large data sets to identify patterns and they constantly adapt to new data to improve efficacy.  

Discovering Intent

Discovering what someone might want.

Making Predictions

Anticipating the likelihood of something happening.

Being Prescriptive

Guiding future decisions and actions taken.

 
 

It is vital to remember that information-- in the sense of raw data-- is not knowledge, that knowledge is not wisdom, and that wisdom is not foresight. But information is the first essential step to all of these.

Arthur C. Clarke,
Best known for the novel and movie 2001: A Space Odyssey

 
 

How Machine Learning benefits ecommerce

With Machine Learning, ecommerce companies, with extensive product catalogues, are better able to optimize conversion by applying automation to highly predictable occurrences that require minimal human judgement.

Benefits include reducing errors, increasing speed of response and freeing up resources by improving efficiency. 

 

Fredhopper Discovery Platform + Machine Learning

We use ML to both automate decisions as well as providing Merchandisers with Actionable Insights that help them apply more effective strategies to meet shopper needs, while also achieving their key commercial goals.

Here are a number of ways we currently apply Machine Learning.

Visual Recs & 'Shop the Look'

Applies fast image processing to identify the style and make-up of a specific product and then uses this to identify visually similar or complementary items (for visual recommendations and 'Shop the Look').

Enabling a shopper to consider all the available alternatives to a selected item, that are visually similar or complementary.

Search Relevance - Ranking Cocktails

Analyzes shopper search behavior and applies weightings to key factors in order to present the most relevant range of items, by priority.

Presents items in order of interest for the shopper.

Personalized Search - Personalized Rankings

Gathers data on the behavior of shoppers, analyzes it to determine affinities, and exposes them as a dynamic data attribute that Merchandisers use to create strategies to target search results.

Exposes shopper preference data for Merchandisers to tailor search results to the current shopper.

Personalized Recommendations - Personal Shopping

Analyzes in real-time both session behavior, as well as historic behavior, to identify the exact areas of interest of the shopper.

Predicts & recommends items that the shopper will be interested in.

Trending Recommendations - Trend Matching

Identifies and generates a digital footprint of the shopper, so that learnings from other like-minded shoppers can be applied.

Predicts & recommends items that the shopper will be interested in.

Hot Mix Recommendations - Collections

Analyzes associations between popular products, to create collections that represent all popular product categories.

Recommends a collection of miscellaneous items, that are strongly associated.

Also liked Recommendations - With the Crowd

Identifies and analyzes items that have strong associations between each other based on like-minded shoppers.

Recommends items that shoppers of the current item also liked/were inspired by.

 
 

ML + Merchandisers

Merchandisers are also directly informed with Machine Learning insights. They can combine these with their own expertize and creativity to define and rank the specific business strategies that will achieve optimal conversion. This enables them to deliver exceptional shopping experiences, at the same time as achieving their own commercial goals, such as promoting high margin items or addressing excess inventory.

By 2019, Artificial Intelligence will alter how 25 percent of Merchants, Marketers, Planners, and Operators work, improving productivity by 30% and KPIs by 10–20%.

 
 

AI without people is not intelligent

Artificial Intelligence in retail is proving itself beyond the hype. It is clear Machine Learning can transform online shopping experiences. However, the application of black box thinking will only get you so far in delivering the relevancy and inspiration that helps you create a real emotional connection with your shoppers. 

This is why the Fredhopper Discovery Platform as been designed as a Platform as a Colleague (PaaC) delivery model.

Meet Freddie.

 

Ready for more?