In retail, empty shelves turn customers away. As much as 24% of Amazon’s on-line revenue can be attributed to customers who couldn’t purchase their items because a local retailer was out of stock. The IoT can improve this.
As part of an ongoing project, we at Thingsquare have developed a way to use wireless laser sensors to detect if a shelf is out of stock and immediately notify nearby staff. The staff then quickly can restock the missing items.
So how useful is it? To quantify the value, the system is currently being deployed in retail stores in Europe to measure the impact on shelf item quantities, restocking speed, and, ultimately, on the bottom line.
24% of Amazon’s revenue can be attributed to customers who experienced out-of-stock at a local retailer
Source
The Problem
Retail employees spend much of their time walking back and forth as they notice the need to refill shelves that are are running low on items. Studies show that 5-10% of products in grocery stores are either unavailable on shelves, or out of stock (source).
Employees cost money so making sure that they spend their time in a productive way is important. If employees can spend more time with customers, and less running back and forth through the store, their time will be more productive in terms of contributing to sales.
Missing shelf items may cause customers to take their shopping elsewhere, such as to on-line retailers. Estimates are that as much as 24% of Amazon’s revenue come from customers who have not found what they were looking for in local retailers (source).
The Solution
The solution consists of two parts that are visible to users:
- Wireless battery-driven laser sensors, attached at the back of each shelf
- A smartphone app, installed by store employees on their phones
The battery-driven, wireless laser sensor, here shown with an iPhone X for scale.
The sensors are attached to back of each shelf and use a laser beam to measure the distance to the closest item on the shelf. The laser beam pulsates with very short pulses so as to not be dangerous. Because employees are instructed to always move items to the outer edge of the shelf, the sensors are over time able to estimate the amount of items on their shelf.
The sensors send their measurements through a wireless IoT network that is created inside each store. The network covers the entire store and its range can be extended with range extenders.
The sensors are powered by a battery that lasts for approximately 2 years. Future versions of the system may use solar cells to achieve a never-ending life
The employees install a smartphone app that, during work-hours, informs when shelves need to be restocked via push notifications.
Smartphone App
In addition to the user-visible elements, the solution is powered by software running on a cloud backend to securely handle the data from the sensors and to take care of delivering push notifications to the app, as well as the full Thingsquare IoT stack that makes sure that the sensors can operate even in very large stores.
Technical Details
Software | The Thingsquare IoT stack Custom mobile app Custom backend application |
---|---|
Hardware | Custom sensor hardware |
Wireless IoT technology | 868 / 915 MHz |
Network size | 100+ sensors in each deployment |
Wireless range | 100+ meters |
Deployment type | Indoor |
Device power | Battery powered |
Quantifying the Value
Quantifying the effect that the solution brings is essential to understanding to the value that the IoT brings to a business.
To quantify the value that the system brings to the business, we do a two-step process:
Step 1: Measure the Baseline
The first step is to measure the current state of affairs – the baseline. We do this by deploying only the sensors, without equipping the staff with the app, and use the data we gather as a measure of how shelf stocking is today. In general, this information is unknown previous to this measurement.
Only when we have data on the baseline can we quantify the effect that our IoT solution brings.
Step 2: Measure the Effect of the IoT Solution
We now know the baseline data, and now give the employees the app, while keeping the sensors in place. The data we now collect shows the effect of the solution. We should see that quantity of items on each shelf goes up, and that the total time that all shelves in the store are empty goes down. This data can then be used in quantifying the effect on the total sales, which is also affected by many other factors, such as the time of month and the time of year.
When the baseline and the effect have been measured, we have determined the value that the IoT has delivered.