ANALYTICS
SECTION ONE
SECTION TWO
How to Improve In-Store Analytics

By Nona Cusick

As we now see overlapping consumer cohorts with very distinct needs and behaviors shopping the 
same stores, there is a need for grocers to look at behaviors at a more granular level. In-store analytics
are increasingly important and provide grocers with increased capabilities to understand shopper segments and significant shifts in shopper behavior. It’s unwise to assume that Boomers will choose to come to a store and that Millennials will tend to shop online, or that Millennials will tend more to use a mobile app to navigate the store.

Food shopping is a more personal experience, and there is a need and ability for grocers to cater to sometimes dissimilar patterns - whether they occur online or in-store. In addition, competitive pressures from other formats are accelerating: C-stores, Dollar and even Drug stores keep nipping at the basket of grocers and affecting revenue and margins. The adoption of in-store analytics by the sector leaders is putting significant pressure on the late adopters. It takes time to gather the data, draw insights and test and learn how to influence behaviors. As a result, early adopters are gaining a competitive advantage.

We see two emerging points of view in how to improve in-store analytics:

1. The “Text-Book” Gap Analysis Approach: A sober assessment of analytical maturity, defining goals, followed by enacting a program that closes the gap between the two.

2. The Agile “Assumption-Busting” Approach: Looking at a narrow area (a region or a category) and challenge assumptions. The key is to assemble a team of data scientists, process experts and a business leader to develop and crash test a series of hypotheses about consumer behavior. This type of approach is executed by a series of agile sprints that bring more sophistication to the model at the same time as you deliver business value. It’s a less “sober” approach, but allows leaders to drive business value while building organizational maturity around data.

Regardless of the approach, we tend to view analytical maturity as a series of four tiers: Data Tier, Data Exploration, Descriptive Analytics and, finally, Prescriptive Analytics and Optimization. Each of these tiers brings together insights about shoppers, products and stores. Here is a look at each:
 
  • The Data Tier has the ability to collect, store and deliver structured and unstructured data from internal and external sources. The data required in this tier should address higher level business needs. For example, the need for an assortment solution may only require product or store data, or may be more comprehensive and tie shopper information to the analysis and, thus, require consumer profile data.

  • Data Exploration corresponds to the need to visualize, dice and slice data to verify hypotheses. For example, what has been the sales trend for salty snacks across our northwest stores for the past three summers? What are consumers saying about this category?

  • Descriptive Analytics correspond to the bulk of the analytical work: price, assortment, display location, shelf configuration, category adjacencies, etc.

  • Prescriptive Analytics correspond to all analytics that result in an actionable recommendation. Think of descriptive as finding out “what’s there” and prescriptive as “what to do about it.”

One aspect to bear in mind is how to execute these strategies. It is generally not economically efficient to develop an internal advanced analytics capability. The most cost effective and impactful approach is the use of managed platforms of analytics-as-a-service: External consultants build and manage the data tier on behalf of the retailer. The exploration layer is made available 24/7 and descriptive and prescriptive analytics can be leveraged as needed.


Nona Cusick is SVP of Consumer Products, Retail and Distribution, at Capgemini, a consultancy.

Click on the LinkedIn logo to join the new Shopper Technology Institute Discussion Group
SECTION THREE