Easy ways to get the answers you need.
Or call us at:
What's in your basket? Predictive analytics for retail
Delaney Turner 270002T14M email@example.com | | Tags:  predictive_analytics retail spss
0 Comments | 2,371 Visits |
This is the first in a four-part series exploring the performance challenges in retail as stores and chains of all sizes begin their recovery from the turmoil of the past year. In today's installment, we look at the ways predictive analytics can help retailers better anticipate what their customers expect from their shopping experience, thus enabling them to better serve the people coming through their doors.
In retail, knowing what your customer will want is a big plus. Cat owners want to buy cat food, not dog biscuits. And parents with grown children don’t look for discounts on diapers, but they may want to pick up wine or a prepared meal for dinner.
In other words, if you can figure out what a shopper wants to buy and have it available or lead them to it, that’s a good shopping experience. And a good shopping experience for customers means a more profitable experience for retail chains. So a cat owner can come into your store and find the right cat food – and maybe buy some pet toys too.
What’s in your shopping basket?
That’s the payoff when predictive analytics meets “market basket” insight. To start, retailers amass vast stores of data from their transaction and point-of-sale systems. There is a lot of good information in the data: about customer behavior, what sells and what doesn’t, and other kinds of buying patterns. The problem is finding it. It’s nearly impossible to mine this data manually.
But the task can be done using business analytics software. Analytics helps you parse and make sense of all that data. Predictive analytics goes a step further. It doesn’t just find information, it uses algorithms to turn past transactions into future insight.
How predictive analytics works
A solution such as IBM SPSS Market Basket Analysis helps you build predictive models to improve the shopping experience. Here’s how it works: The simplest kind of market basket analysis is to take transaction data and use algorithms to detect combinations of products that tend to be bought together (such as skin care and hair care).
From here, you can stack these products together or use a promotional display offering a discount if people buy both.
This is undifferentiated analysis. Where it gets interesting is when you combine this kind of analytics with other customer data – such as demographics (age, zip code), behavior and attitudes (satisfaction scores), based on things like loyalty programs.
Then you can use predictive models to determine which combination of products or offers works best for which group of customers. A simple example: association rules might tell you that younger career women tend to buy wine and chocolate. Or middle-income men who buy beer and pizza are also more likely to buy items like potato chips.
Responding to the "smarter consumer"
Predicting consumer behavior
The value: you ensure your product offers and promotions match shopper preferences and behavior. By linking purchases to shoppers, you can take this even further: tailoring offers to specific customers and driving more personalized campaigns.
That’s predictive analytics in a nutshell. It’s a powerful tool that can change the game for retailers. So they move from a reactive approach to a more predictive, proactive model – and have the right products in the store at the right time.
And, most important, each customer is more likely to find what he or she is looking for.