Guest Column: Increasing retail sales by bridging IoT, geospatial and AI
Consumer behavior changes rapidly and is heavily influenced by a person’s mood.
Until now, determining exactly how and why this happens has been difficult for retailers, but modern technology is now able to provide the answers. By utilizing Internet of Things (IoT) technology, retailers can develop a new type of smart store environment that can improve the customer’s mood and increase main retail key performance indicators (KPIs) and metrics.
Geospatial data like the customer’s region of living, trade area characteristics, and the lifestyles of local people is also used in a cross-reference with other stores in different locations to get valuable insights on how the brand's particular customer profile gets affected by the environment.
Together these metrics provide a much deeper understanding of customer expectations and provide greater insight into how stores can improve efficiency and performance.
Thermal comfort and buying behavior
Thermal comfort is a complex set of metrics that are difficult to analyze. One is for the amount of time spent inside the store, or near a particular department, and the other is the average number of items in the basket that is related to the amount of time spent.
To define a human behavior pattern and its dependencies on the human comfort, we bridge IoT sensor data, geospatial data, and using machine learning algorithms for analyzing the data and making actionable insights based on that analysis. Analytics can then be used to help tune in-store parameters in real-time as external environmental factors like the weather outside keep changing.
Lighting and consumer buying behavior
Lighting also plays an important role in influencing the buying decision of the customer in-store. Research has shown consumers stay longer browsing in different departments if those departments are adequately lit. As well as influencing buying decisions, there also has to be a balance between efficient lighting for increasing sales and efficient energy consumption.
Ultimately, adjusting in-store lighting levels either has a significant effect on sales or it doesn’t. If it does, the retailer can set the lighting to the level that was proven to be the most profitable for that particular store. If it doesn’t, the retailer can save on the energy bill by lowering the lighting level to the lowest point in the acceptable range since that wouldn’t influence sales. Either way, the information gained from the process is equally valuable.
Imagine that you have two stores in different locations. The first store is located in the area with 61% of the population living a healthy lifestyle and there are no competitors inside its trade area. As a large percentage of the population within this location lead a healthy lifestyle, it is very likely that sales percentages in the fresh Produce or organic produce department of that store will be significantly higher.
In the second store, we can see that only 28% of people within that area live a healthy lifestyle and there is a vegetable store right across the street. This means that the fresh produce department within that store is likely to have much lower fresh produce sales that the first store. Considering this, it wouldn't be correct to compare the influences of lighting near the fresh produce department on buying behavior among these stores.
The best solution in this case, then, will be to use light level data from IoT sensors on constant base, and to use geospatial data to define similar stores (store clusters) by different parameters like real estate, trade area characteristics, demographic data, and a number of competitors and amplifiers.
By adopting this strategy, a retailer can identify key buying patterns that enable them to make actionable, performance-enhancing changes to their business by combining IoT, geospatial data and machine learning algorithms. However, it is important to realize that analyzing only part of these data points would lead to wrong conclusions. For example, though thermal comfort has a massive influence on human behavior, in order to identify those persuasion points clearly, IoT sensor data and geospatial data must be analyzed together.
Mike Mack is CEO and co-founder of Fract.
Until now, determining exactly how and why this happens has been difficult for retailers, but modern technology is now able to provide the answers. By utilizing Internet of Things (IoT) technology, retailers can develop a new type of smart store environment that can improve the customer’s mood and increase main retail key performance indicators (KPIs) and metrics.
Geospatial data like the customer’s region of living, trade area characteristics, and the lifestyles of local people is also used in a cross-reference with other stores in different locations to get valuable insights on how the brand's particular customer profile gets affected by the environment.
Together these metrics provide a much deeper understanding of customer expectations and provide greater insight into how stores can improve efficiency and performance.
Thermal comfort and buying behavior
Thermal comfort is a complex set of metrics that are difficult to analyze. One is for the amount of time spent inside the store, or near a particular department, and the other is the average number of items in the basket that is related to the amount of time spent.
To define a human behavior pattern and its dependencies on the human comfort, we bridge IoT sensor data, geospatial data, and using machine learning algorithms for analyzing the data and making actionable insights based on that analysis. Analytics can then be used to help tune in-store parameters in real-time as external environmental factors like the weather outside keep changing.
Lighting and consumer buying behavior
Lighting also plays an important role in influencing the buying decision of the customer in-store. Research has shown consumers stay longer browsing in different departments if those departments are adequately lit. As well as influencing buying decisions, there also has to be a balance between efficient lighting for increasing sales and efficient energy consumption.
Ultimately, adjusting in-store lighting levels either has a significant effect on sales or it doesn’t. If it does, the retailer can set the lighting to the level that was proven to be the most profitable for that particular store. If it doesn’t, the retailer can save on the energy bill by lowering the lighting level to the lowest point in the acceptable range since that wouldn’t influence sales. Either way, the information gained from the process is equally valuable.
Imagine that you have two stores in different locations. The first store is located in the area with 61% of the population living a healthy lifestyle and there are no competitors inside its trade area. As a large percentage of the population within this location lead a healthy lifestyle, it is very likely that sales percentages in the fresh Produce or organic produce department of that store will be significantly higher.
In the second store, we can see that only 28% of people within that area live a healthy lifestyle and there is a vegetable store right across the street. This means that the fresh produce department within that store is likely to have much lower fresh produce sales that the first store. Considering this, it wouldn't be correct to compare the influences of lighting near the fresh produce department on buying behavior among these stores.
The best solution in this case, then, will be to use light level data from IoT sensors on constant base, and to use geospatial data to define similar stores (store clusters) by different parameters like real estate, trade area characteristics, demographic data, and a number of competitors and amplifiers.
By adopting this strategy, a retailer can identify key buying patterns that enable them to make actionable, performance-enhancing changes to their business by combining IoT, geospatial data and machine learning algorithms. However, it is important to realize that analyzing only part of these data points would lead to wrong conclusions. For example, though thermal comfort has a massive influence on human behavior, in order to identify those persuasion points clearly, IoT sensor data and geospatial data must be analyzed together.
Mike Mack is CEO and co-founder of Fract.