Enhancing Retail Analytics with a Unified Customer Profile

Jason McNellis | December 22, 2015

Today I want to briefly discuss a few ways location intelligence could be used to bolster a unified customer profile. As usual, we have two goals in building out a customer profile:

  1. Provide more personalized and engaging experiences to our customers
  2. Create a richer data asset for analysis (to enhance tactical and strategic insight creation)

Let’s imagine a retailer who has placed beacons at every entry point to their store. This retailer also has a voice of customer (VOC) system that catches feedback on some small portion of transactions and they have a loyalty card that captures about 70 percent of spend.

Fragmentation of customer data is common for retailers. The loyalty card data is in an on-premises data mart, the VOC responses are stored in the cloud, and the beacon data is stored by a second vendor and sends batched reports. If these three systems were isolated and never communicated they would still provide insights:

  • The beacons measure store traffic which is a denominator of one executive KPI: purchase conversion.
  • The customer surveys go directly into another executive KPI: customer satisfaction. The survey also determines one component of store managers’ bonuses.
  • The loyalty card provides a rich data set that is used for predictive models to improve direct offers. It also used to measure the retention KPI.

While each data stream provides value in isolation, think of the value unleashed if the three data sources were combined and constantly in sync? A short list could include:

  • Wait time could be derived and its relationship to satisfaction quantified. This relationship could be used as an input to staffing models.
  • New KPI’s could be derived like seconds per dollar shopped, a metric that has been shown to be highly predictive of store sales growth.
  • Customer survey insights could be sharpened by matching exactly to items purchased. Conversion rates could be increased (also lowering response bias) by personalizing questions based only on those departments shopped.
  • In store display changes, rather than just be evaluated on department sales, could also be emulated on market basket compositions, dwell time and changes to customer experience.

As the examples above illustrate, combining customer data sources allow research into the drivers of KPI’s allowing for more specific and more justifiable recommendations. Over the last decade there has been a clear mandate for turning insights into action. Knowing the value for a given KPI or how it is trending is no longer sufficient. When KPI’s go down executives will want to know why and what can be done to reverse them. And increasingly they will want to hear data-driven, quantified answers.

Building a unified customer profile clearly leads to a richer data set for analyses. By linking multiple data sources at the customer level and keeping them synched, the analytics team is empowered to better support executives because:

  • Trade-offs can be analyzed when multiple views of the customer are brought together—like sales, satisfaction, and browse time in the example above.
  • It simplifies testing new hypotheses because the data is already in sync—promoting responsiveness. The often quoted 80 percent of analysis is spent on data preparation is dramatically decreased increasing analytic output with the same staff.
  • One version of the truth can be accessed from multiple tools—from predictive modeling to business intelligence, to campaign management—can all simultaneously access the same set of customer profiles.

Are you looking to leverage location data to fuel analyses or customer experience? Leave a comment.

retail-case-study

Share This
Jason McNellis
Jason McNellis