Personalization = The Combination of Data Science & Human Judgement


Have subscription boxes had their moment in the sun? You might be inclined to think so – after all, Hitwise reports that September’s overall traffic in that space has dropped 3% compared to a year ago, with meal preparation kits like Blue Apron taking the biggest hit. But there’s one subscription box brand that’s doing quite well for itself. Apparel retailer Stitch Fix has seen its traffic more than double in the same period. It’s growth they credit to their use of data.

Using Data to Create Robust Customer Profiles

As part of their sign-up process, Stitch Fix collects more than 80 distinct data points, including fashion preferences and fit challenges. Once this data is analyzed, human stylists add their input, and users receive boxes of clothes chosen for them to try and buy. Stitch Fix has as their goal the provision of 1 to 1 personal service through the combination of data science and human judgement.

Both components of the equation are important. Data science provides us with an objective analysis of customer behavior. Examining what customers actually buy is more meaningful than customer commentary on what they’d like to buy; human stylists can interpret fashion nuances that the best algorithm cannot. Human judgement coupled with predictive data is a winning combination.

Creating Customer Profiles is an On Going Process

It’s always appropriate to ask your customers for more information so that you can serve them better. Presenting an 80+ question questionnaire the way Stitch Fix does may be a bit overwhelming for your customer; consider a series of smaller surveys instead. In addition to this data set, you’ll want to consider the objective information gathered by your website analytics program. The third element in this equation is human insight: front line sales professionals have valuable insights to contribute to customer profiles.

Customer profiles are not and should not be static. Customers change over time, both demographically and behaviorally. Your data can provide an overview of who your customers were when they started doing business with you, who they are today, and who they are likely to be tomorrow. Behavioral changes can include shopping patterns, device usage and more. eMarketer sum it up thus, “While reading the trends right is key, getting data about customers from many different points is increasingly important if retailers hope to attract, please and keep customers.”

You Don’t Have to be a Data Scientist to Run a Data-Driven Business

Stitch Fix has a reported 75 data scientists on staff, many at the PhD level. Smaller businesses are most likely not going to be able to commit the same level of resources; that being said, it’s absolutely possible to run a data-driven business without being a data scientist. Learning to make the most of the data you already have available in Google Analytics is enough to provide a meaningful competitive advantage, especially when you have the right tools at your disposal.

Personalization = The Combination of Data Science & Human Judgement
Article Name
Personalization = The Combination of Data Science & Human Judgement
Data science will provide an objective analysis of customer behavior. Ask your customers for information so that you can serve them better.

Comments are closed.