Analytics for all

How to Analyze Purchasing Habits to Reach the Right Customers with Watson Analytics

Blog Home > How to Analyze Purchasing Habits to Reach the Right Customers with Watson Analytics

How to Analyze Purchasing Habits to Reach the Right Customers with Watson Analytics

Determining the effectiveness of an upcoming campaign can be a difficult task, especially if you’re focusing on a specific target audience. Luckily, Watson Analytics features some amazing predictive analytics capabilities that can make these types of efforts much more manageable.

The detailed example we’ll walk through will show you the types of campaign validation you can expect to achieve. To follow along and experience the results for yourself, first sign up for Watson Analytics here (it’s free) and then download the example’s sample data set.

Let’s get started. I’m a marketer for a national grocery chain. I have data about customer purchasing habits over a period of four weeks. This data includes how much they spent on groceries each of the four weeks and how they responded to coupons over the same period.

I upload the data and review the data quality report to ensure I’m working with clean, accurate data.



To explore customer buying behavior, I use “Predict” to see which variables affect how customers spend their money.  The predictive target I choose is “Amount Spend.”

I have access to all of the different fields in data and they are the inputs Watson Analytics uses for building the strongest model. It looks like the key factor that determines “Amount Spend” is actually who customers are shopping for. But I can dig deeper to uncover more predictive information.


I explore a decision tree, which is paired with the top decision rules explained in plain language, to see what went into the decision. I get validation for a specific strategy: Women with families respond more favorably to coupons that have the greatest discount. Although this is not a big surprise, it does enable me to make very specific offers that will have a greater chance of yielding better results. Watson Analytics has shown me how to most effectively reach specific customers with the right offer.



I can explore this data even more, using the insights I’ve discovered to further validate my strategy.



If you haven’t tried Watson Analytics yet, now’s a great time to start by signing up for the free edition.

To see even more of this story in action, view this video:

Leave a Comment

Leave a Reply

Your email address will not be published.Required fields are marked *

More Analytics for all Stories

Analytics for all

Timothy Walker

Kick out the jams—the Analytics University design jams!

The IBM Analytics design team will lead Analytics University design jams and related sessions in Stockholm on September 17–20 and Miami on October 2–5.

Analytics for all

Tim Walker

10 can’t-miss sessions at IBM Analytics University 2018

Recently we invited you to “take yourself back to school” at IBM Analytics University 2018. Now it’s time to figure out your class schedule!

Analytics for all

Gregory Verlinden

Advanced analytics meets real-world results at Analytics University in Stockholm

As my colleagues and I work with IBM’s European clients, we get to witness the business impacts of advanced analytics and data science every day.