Defining the Customer Journey on HP.com
Screenshot from HP.com after a customer puts a printer into her cart.
Project Challenge
HP engaged the Yale Center for Customer Insights (YCCI) for a Spring 2018 Discovery Project to learn about the online customer journey for HP personal computers and printers within the commercial division.
Our team was asked to define a "typical" online journey for an HP.com commercial customer and define the online characteristics and behaviors at each stage of the journey, then drive recommendations based on our insights.
We were given three months of click-stream data to analyze. This consisted of about 58.5 million clicks, each click described by some combination of 521 variables.
Teammates: Apoorv Chaturvedi, Christopher Miller, Grace Myers, and Joshua Posner
An early stage of our process: identifying useful variables. For the purpose of this project, we were able to use such a limited number of variables thanks to creative string searches using the “url” variable.
Our Work
We created a broad set of stages for the online journey, then analyzed this journey across distinct customer segments, based on identifiable firmographics, online behavior patterns, and product-related characteristics. We then focused on how the journey might differ for visitors shopping for different product types (e.g., premium versus standard laptops).
The customer journey we found on HP.com was even less linear than we anticipated.
Impact
In working with HP’s click-stream data, we also uncovered a list of opportunities for better management of their database to improve usability for future data science hires.
Since our project's conclusion, HP has expanded its data team to further explore the online journey for a commercial customer, as well as organize their data in manner that reduces future technical debt.
Next Step:
Understand how to reconcile the online click-stream data with ethnographic customer research