The learning computer is about to tell consumers what they want
Over the last few years retailers have coveted Big Data, this data is often extensive, wide ranging and unstructured, encompassing information relating to customer spending habits, trends and influences.
However having this data is one thing, being able to analyse and make sense of the data is another, being able to use the data to predict accurately what the customer actually wants can create a whole new level of customer – retailer experience.
This is an exciting new world of cognitive commerce whereby retailers can use systems which develop expertise as it learns from the customers it interacts with.
IBM have been developing a cognitive computer system called Watson for several years, Watson is a learning system that can analyse huge amounts of unstructured data to reveal useful insights.
IBM actively encourage developers to find new uses for their learning system and these are starting to emerge in retail, with some really innovative API’s that has introduced Watson to the world of commence and its abundance of data.
Great practical use of IBM Watson – Cognitive scale
Cognitive Scale has developed a cloud based cognitive system that is built on Watson, machine learning, NLP and advanced analytics, these combine to form guided commerce which integrates with the retailers existing commerce solution.
This highly intelligent system builds a shoppers cognitive profile so that it can generate personalised recommendations based on price, colour, size, historical shopping trends, it combines this information with what it knows about the shopper’s social media patterns, their location and even the local weather to predict what the shopper is most likely to want to buy
Watson has limitless uses in retail, for example in fashion, being able to analyse social media, geographical trends and weather conditions could mean that fashion retailers are more prepared as they could predict what the customer is more likely to purchase. For the consumer Watson can help customers find the right shirt for them by filtering out the styles, colours and patterns that Watson has learnt that they don’t like and only shows them shirts that it can accurately predict they will like.
If a customer dislikes a type of shoe, then it will not show the shopper that type of shoe again, if it has learnt that certain colours won’t go together then it won’t show these items together or if it knows that certain brands sizes are smaller than others the system can work out what would fit the customer.
All of these factors have multiple benefits to both consumer and retailer, for example around a 3rd of online clothing and apparel purchases are returned, the main reason for this is that shoppers are not sure if the item will suit or fit them, so they buy a few sizes and styles, then return the ones that don’t fit or they don’t like, this is an inconvenience for the consumer and a huge cost to the retailer, using a solution such as cognitive scale the consumer would purchase the product best suited to them in one transaction, is less inclined to by others “just to try on” the result of which could mean significant cost savings to the retailer.
By James Pepper, Technical Services Director, Vista Retail Support
