In-Store Analytics Hits Prime Time: Part 1

Imagine you’re in a hot air balloon tethered 100 feet in the air over a glassed-ceiling supermarket, providing you a bird’s-eye view. You observe shoppers as they move through the store, see what departments they go to, where they pause to consider a product, what aisles they go down, and which they don’t.

Now you lower your observation platform to hover over the store entrance. You can accurately count the number of people coming into the store and identify unique characteristics and the mood of each shopper.

This is powerful information to a retailer. Knowing how each shopper behaves in the store along with shopper demographics and traffic, can impact every aspect of a business, from marketing to merchandise planning.

This information is invaluable to retailers and brand manufacturers alike. These details can help brands understand the impact of package design, the effectiveness of signage and other merchandising initiatives, and know a product was actually picked up in the store.

Luckily for retailers and brands, in-store shopper analytics is quickly maturing, driven by a growing pipeline of new solutions coming into the market, decreasing cost of existing solutions, and increased understanding. Technology has digitized the in-store environment, providing deep insights into true shopper behavior.

Emerging solutions

Today’s in-store analytics solutions come in many forms, incorporating video, mobile and location-based technologies, among others. Video analytics first appeared on the retail scene around 2008, and early solution providers leveraged new digital cameras and powerful processing to convert video images to data points. While the insights that video analytics can provide are powerful, its high costs have stymied its widespread adoption in consumer goods retail.

Mobile analytic solutions can anonymously detect shoppers’ mobile devices as they enter and move through the store. Properly deployed sensors enable high levels of accuracy and identify which category the shopper is in front of. Mobile analytics are available at a fraction of the cost of video, a new entrants into the space, with task-specific sensors, can deploy a typical supermarket solution for less than $3,000.

Other new location-based technologies are also flowing into the market: one company is using the unique magnetic wave "signature" within each store to provide shopper location, and at least two other companies are embedding "codes" into light fixtures to support shopper location.

Other vendors are taking people-counting solutions to the next level. New solutions provide anonymous demographic data, counting male or female shoppers, estimating age ranges and identifying ethnicity. One solution even reports the shopper’s mood using advanced recognition software and the largest library of facial images. No video is retained to ensure shopper privacy.

Powerful insights

Stop and consider the power of this new data. A retailer with a shopper-mood scorecard revealing the number of shoppers by hour and by day, along with their moods, can now grade service department managers and store managers by the proportion of happy shoppers going through each week. These capabilities create new ways to measure and manage business.

Other solutions leverage kinect-style sensors to provide a three-dimensional view of shopper behavior at a given category within the story. One solution provides directional traffic flow by the category, reports on the number of shoppers who stop and "dwell" in that category, and then provides analytics on what products the shopper picks up, from what shelf, and whether the product is put back on the shelf or in the shopper’s basket.

Think of these new measures as a funnel. At the top are all the shoppers entering the store. Some go to the produce department, reported as a department conversion rate. Others go to an aisle, then to a category, creating similar conversion rates. Still other shoppers dwell in front of a particular category, leading to a dwell conversion rate. Systems can measure that dwell event and report average dwell time. Linking these metrics with purchase data, retailers can measure and report a purchase conversion rate.

That’s just the beginning, though; the learning and use of in-store analytics can go further, and we’ll explore that in Part 2 of this article.