In-Store Dwell Time: 9 KPIs to Master

Published by Ronny Max on

What Are InStore Time Metrics?

Here’s a secret in Location Analytics. You cannot optimize the physical store without Time Metrics!

Retailers often use Time-Based metrics in evaluating payroll costs. With the emergence of People Tracking Technologies, our attention turned to customer engagement.

Time Metrics have value beyond Video Analytics, WiFi Tracking, and other technologies.

As data becomes more accurate, we do more with Time Metrics.

Detection technology introduced the Average Time and Max Time.

Tracking technologies capture the time per Object. And Vision Analytics has the ability to talk in pixels and seconds.

Time measures quality and productivity. In the checkouts, the Time Metrics monitor speed. In customer service, Time defines the quality of the interactions.

Why Time Models?

Time is more prevalent in store operations than you may think. There are three core costs to the physical store. The first is real estate. These costs are part of long-term decisions by the retailers. The second is inventories. From a cost perspective, the decisions about inventories and products are made over a period of months. The last cost is payroll.

The advantage of the labor costs is in their flexibility. This proved beneficial to retailers who must adapt in the short term. It also created sustainability challenges. Many retailers suffer from high turnovers, lack of training, and staff dissatisfaction.

From Activity Based Costing to Customer Engagement

Winning retailers spend resources and research into labor costs and productivity. They value their employees with Activity Based Costing (ABC). The method transcribes payrolls in term of activities.

Take fulfillment. This is an area where most studies about the efficiency and productivity of labor come to play. If an associate costs $15 per hour, with an extra $10 of benefits, then the retailer pays $25 hourly wage. In other words, $25 is the value of the associate’s activities per hour.

Task Management solutions are great for managing the fulfillment and stocking activities. Such technologies improve store operations by setting priorities and assigning responsibilities. They provide accountability and reduction in costs.

In physical stores, we are taking the idea of Time Management on another direction.

In Optimization, we see Time as a function of increasing sales.

Moreover, we focus on profits. Due to the duel usage of Time Metrics as a function of both costs and revenue, we can optimize the store for profits.


The Zen of Heat Maps

Heat Maps help us understand the nuances of location and time. As a visual tool, Heat Maps are usual. We can easily see how and where people stayed in specific zones. The challenge of heat maps is in the accuracy of the data underneath. A phenomenal visual map often turns out as a useless management tool. Thus Heat Maps depend on the quality of the data.

Here is how to think about Heat Maps in the context of Time.

Detection-Based Heat Maps

The technologies of detection gave us location heat maps. Those heat maps provide snapshots of objects, within a period time. Each snapshot puts objects in particular zones in the store. Since the detection depends on time segments, the heat signature can vary. Thus the same zone can be changed from red or green if we view in 15 minutes segments instead of 3 hours.

Device-Based Heat Maps

With device-based tracking, we face new challenges. The heat snapshots depend on signal frequency. Often the signal is detected in the specific second, but when we adjust for the time segments it is no longer valid. For example, the signal is captured in the Milk Aisle but the person already moved to the Meat Area. As a result, the visualization is no longer valid.


Time-Based Heat Maps

Static Time comes from the Stanford University Vision Project. In this heat map, the dots are a function of pixels, regardless of the physical world. We can adjust the heat signatures based on how long the dots are stationary. We called this Static Time.

Static Time has many advantages. It means the location of the object has no dependencies on hardware. And the position of the object relates to the image, not the sensor, nor camera. As a result, the time capture can be accurate in seconds.

The big challenge is technology. But this is changing, fast. And we should be seeing more and more machine-learning solutions.

9 Time-Based Metrics

Time-Based frameworks have been around for a while. They started at the Point of Sale and moved into Customer Engagement.

Below are highlighted time metrics:

1. Transaction Time

The most well-known time metric provides insights into the checkout process. Together with Scan Rate, we can assess the productivity of cashiers. And it helps in learning about Average Basket and Queue Management.

2. Idle Time

Probably the most important, and the least talked about, the metric is Idle Time. The Idle Time is the time between transactions. This is when the human operator is not active. There are various reasons for Ideal Time, and most relate to consumer behaviors.

Idle Time is also one of the most secretive metrics because it provides a competitive advantage. With Idle Time, we can assess operational efficiency. For Walmart, for example, one second shaved from Idle Time is worth millions of dollars.

3. Wait Time

Wait Time is the core metric in Queue Management. Most retailers use Average Wait Time as KPI. But the Average Wait Time has almost no operational meaning. And Time-Based Frameworks such as “Serve 95% in less than 3 Minutes” is far more useful.

4. Service Time (Checkouts)

Service Time has various meanings. In this case, we refer to the length of time it takes the customer to check out. This metric includes both the Wait Time (queues) and Transaction Time (in front of the cashier).

5. Service Time (Sales Floor)

The concept of service time also has a sales function. The service time includes a variety of interactions on the sales floor. We can test for service at the counter, at a sales point (i.e. fitting room), and on the sales floor.

6. Dwell Time

As the output from detection technologies, Dwell Time is the time spent in a specific zone. Many retailers and providers use the Average Dwell Time as the measure of customer engagement.

7. Engage Time

Also known as Linger Time, this metric defines an interaction with a display. Engage Time refers to a more limited area than Dwell Time. And this metric serves as a trigger to service alerts.

8. Stay Time

Stay Time is a function of the individual behavior of tracked objects. Dwell Time (zone) and Engage Time (display) are time metrics of a specific location.  Stay Time tracks the behaviors of staff and customers.

In other words, Dwell Time and Engage Time to refer to behaviors of people in a location. But Stay Time quantifies individual behaviors.

9. Static Time

Static Time takes the concept of Stay Time further. In Vision Analytics (see above), the tracked dots are image-based. They are not dependent on the physical world.

Bringing It All Together

In the optimization process, time metrics play an important role. This is especially true in testing for customer engagement. If you recall the InStore Scene, the customer had three points of engagement. The time spends at each point was a function of the product (fresh, frozen, and coffin) and steps in the journey.

As Time Metrics become more accurate, we can achieve more in our ability to optimize the store for profit.

Ronny Max

Ronny Max is a recognized author, expert, and speaker. She is the Principal of Silicon Waves, which provides C-Level Consulting on In-Store Optimization, Location Analytics, and People Tracking Technologies. In 2015, Ronny served as the retail expert in Stanford University Vision Project. In 2017, she founded Behavior Analytics Academy.