Retail Insights Using Estimote Stickers
While ecommerce has been able to collect data on every click and purchase a customer makes, brick and mortar stores are relatively blind to what their customers are doing. To close that gap, my team set out to find a way to measure customer interaction with products while in the store.
In a previous blog, I discussed beacons: devices that use Bluetooth to estimate your proximity to its location and provide contextual experiences in relevant areas, like a push notification with a promotion when a user enters a store. Recently my team worked with a different type of beacon called the Estimote Sticker, which provides location context for physical objects, rather than venues. A sticker, which is much smaller than a regular beacon, can broadcast data about an object’s location, motion, and environment to a mobile device within range.
We generally only think about installing a handful of beacons on walls and fixtures within a given space, but what if we installed hundreds of nearables on the products themselves?
Stickers as Product Interaction Sensors
My team expects that retailers will begin to use multiple devices and sensors to visualize the customer journey through the store and improve the in-store experience, similar to how retailers optimize their ecommerce site using conversion funnels. Many companies, like Google, Shopify, and Adobe, have cloud-based ecommerce platforms that are customizable and make it easy to create engaging and personalized online experiences. Meanwhile, there is still a huge disconnect between the physical and digital world.
By placing Estimote Stickers with motion-sensing technology on products, we can analyze how customers engage with the products in a retail store. Using this information, retailers will have a better understanding of customer product preferences so they can adjust merchandise, store layout, and customer services accordingly.
We created a product engagement demonstration using Estimote stickers, an iOS device, Google’s Cloud Platform (GCP), and a Shopify development store.
We configured the stickers to send data during motion events. An iOS app retrieves that data and sends it to a GCP Pub/Sub topic. Pub/Sub was the natural choice for processing these data streams from the device as it is fast, scalable, and decouples event processing from the code handling user-facing requests. The data packet sent to GCP includes the beacon’s universal unique identifier (UUID), the motion start and end time, the motion duration, the X, Y, and Z-axis acceleration value, and the iOS device name. This data is saved inside a SQL data storage table to track past engagements.
For retailers to visualize their store information, we developed a dashboard that is hosted on GCP’s App Engine. The dashboard uses a combination of WebSocket connections to view product engagements in real-time and REST calls to view top engagements, configure beacon-product associations, and retrieve products from Shopify’s API. The dashboard can be accessed via URL on a browser or the iOS app, which in addition to collecting sensor data contains a web view.
Below are screenshots that show how easy it is to assign an Estimote sticker to a product from a Shopify store.
The sticker would be placed inside an envelope or something that is easily attachable to a product. On the Beacon Configuration page, an associate can move the selected sticker to see it listed in the “Unassigned Beacons” table. We designed it this way because UUIDs are not visible on the sticker and some stickers may have the same color or image (like a bicycle), so there is no way to easily identify them other than through movement. When the associate selects the edit button, he can select one of the many products available on the Shopify store. After the sticker is assigned, the associate physically attaches the sticker to the correct product.
The Product Engagement page of the dashboard allows an associate to view recent interactions caused by product or sticker movement. The “Top Items” section shows top product engagements for the day, week, month, and year. To reassign the sticker, return to the Beacon Configuration page and select the edit icon.
Using this dashboard, retailers can see which products customers prefer, but this is only the beginning of what can be accomplished with the data. Another product engagement application is to detect which products are being tried on in the fitting room. A smart mirror could change its light settings to match the item style or recommend complimentary accessories.
This demonstration is one example of how retailers can collect and analyze data from sensors to create personalized in-store experiences. In addition to Estimote stickers, my team has integrated other devices, such as RFID readers, cameras using video analytics, and MAC address sniffers, into the platform to learn more about customer behaviors.
Interested in hearing more?
Please contact me or my team members (Sunil Shettigar or Dave Nguyen). To read more about Accenture Labs and our R&D areas, visit our website.
