The Brickstream System enables a closed-loop learning cycle, so you can continuously capture and analyse information about the service level provided to your customers and the allocation and performance of front-end service colleagues and resources.
Retailers typically realise business value from this management information with respect to the following business problems:
- Front-end wage cost reduction
- An effective central queue management performance measure
- Real-time proactive service management
- Productivity/efficiency increase
- Resource optimisation and allocation – number and mix of checkouts
- Improved customer check-out experience
Airport Service Management
Data has shown that by properly allocating colleagues and resources to match customer demand at the front-end, cost savings can be achieved while at the same time realizing improvements in customer service levels. Typical cost savings are in the range of ten to twenty percent of front-end labour costs with an ROI payback period of less than six months.
Most retailers current Management Information (MI) on queue management performance is flawed, being collected from the sporadic and inaccurate input of queue length data by front-end operators or mystery shopper data. Furthermore, store management are ill equipped to deal with front-end breakdowns in a proactive manner.
So how do you design a system that addresses the expectations of customers, the expectations of store colleagues, the characteristics of your checkout process, and the long-term strategy of your organisation?
- Accurately and continuously count traffic coming into your stores, for a core understanding of your sales opportunity
- Measure queue and service/transaction times for each customer in the queue (Structured, unstructured)
- Use predictive monitoring to proactively respond to rushes and lulls in checkout lines by opening cashier stations or shifting labour among lower priority tasks
- Use real-time and historical reports to improve staff allocations to reduce checkout lines and improve customer to employee ratios in key areas
Brickstream uses a unique combination of hardware, patented software, and customised databases to produce actionable intelligence enabling the best service strategies for your organisation across the entire chain. Brickstream runs no data manipulation or assumption based mathematics on the data; the data delivered is an accurate and true-to-life picture of the actual store activity.
The following paper contains an overview of the Brickstream queue management solution including the following topics:
- BehaviorIQ Analytics
- Return-On-Investment
- Data Collection Platform
- Brickstream Portal
- In-store Dashboard
- Reference Configuration
- BehaviorIQTM Analytics
Table 1 below contains a list of metrics provided by the Brickstream queue measurement solution. These metrics require no inputs from external systems.
| Term |
Description |
| Arrivals/Departures |
Total number of customers who arrived/departed at/from the store. |
| Queue Length |
Average, minimum, and maximum number of customers waiting in a queue. This data can be further aggregated to provide average queue length across all open lanes and total queue length across the entire front-end. |
| Queue Wait-time |
Number of seconds a customer spends waiting in the queue. Wait-times are only computed for those customers that enter from the back of the queue and then exit the front of the queue, i.e. enter the payment area. This metric is based on actual wait-times and is not statistically generated from queue length. |
| Lanes Open |
Number of lanes open for service. A lane is considered open for service when a cashier is present at the register. |
| Lanes Idle |
Number of lanes open that had no customer activity during a specified time period. |
| Lanes Out-of-Compliance |
Total number or percentage of open lanes considered out-of-compliance during a specified time period. Definitions for out-of-compliance can be based on queue length or queue wait-time. |
| Utilisation |
Percentage of time that a service point is occupied. |
| Target Lanes Open |
Total number of open lanes required to meet a customer service objective model based on a store level queue length or customer wait-time objective. |
| Open Lanes Variance |
Specifies the total number of lanes for which the front-end is overstaffed or understaffed during a specific time period. |
| Forecasted Expense Impact |
A projection of the additional cost increase required or the cost savings obtained when meeting the Target Lanes Open objective. |
All metrics can be aggregated by lane type (full, express, self-checkout, etc.) as well as time-of-day, day-of-week, and other dimensions. The Brickstream solution also supports store-to-store comparisons. Brickstream provides the capability to integrate transaction data in order to support analysis of the impact of service levels on conversion rate, basket size, and other sales metrics.
Passenger Satisfaction
Return-On-Investment - Cost Savings
While there are obvious benefits to customer satisfaction, which may result in an increase in sales and customer loyalty, customer satisfaction is difficult to measure and is impacted by many variables. As a result, a primary focus is on optimisation and the hard benefit of cost reduction on a daily operating basis, through changes to front-end labour investment and allocation.
Figure 1 below is an example graph of the Average Lanes Open, Total Queue Length, and Target Lanes Open collected on each Friday over an eight week time period. The metrics are the average value for each 30 minute time period (Configurable).
The graph demonstrates two common problems with respect to front-end staffing, which can be seen in the area between the Average Lanes Open and the Target Lanes Open metrics. The difference is graphed as the Open Lanes Variance metric. If front-end staffing appropriately matched front-end customer demand, the two lines would remain close together and variance would drift around 0. However, in this store, the front-end is typically overstaffed in the morning and the evening hours and understaffed from 4:00 pm until 5:30 pm.
In order to compute the potential savings that could be realized if the front-end labour matched the target model, the Open Lanes Variance and Target Lanes Open are summed at each half hour time interval. The sum of Target Lanes Open is 263 hours. All values are rounded to the largest integer, since full lanes must be opened. The sum of Open Lane Variance is 35, which yields a potential cost savings of 35 / 263 or 13%. This calculation assumes additional labour has been applied between 4:00 pm and 5:30 pm to address the staffing issue.
The following are a list of assumptions used to compute a potential yearly cost savings.
- Twenty standard lanes
- Average hourly labour cost: £7 per front-end colleague
- Average total colleague hours per day: 192 hours
- Average daily wage expense: £1,344
- Estimated savings of 10% or £134
- Yearly (monthly) cost savings of £48,910 (£4,076) per store
- Based on a typical installation cost of £20,000, this is a breakeven return on investment timeline of 4.9 months.
Figure 2 below shows the impact of the staffing issue from 4:30 pm to 6:00 pm on customer wait-time. An obvious result of reallocating a portion of labour to the afternoon will be a reduction in wait-time and an increase in customer satisfaction.
Airport Front End Services
Staff Reallocation/Optimisation
In many retail environments, front-end managers monitor customer flow and demand at checkout in order to address service breakdowns by opening additional lanes, enacting ‘queue busting’, or invoking other protocols. The problem with this approach is that it is reactive; solving the problem after the breakdown occurs. This is also prone to error, since it is dependent upon staff recognising the service breakdown and takes up valuable time from the front-end management. In addition, this can lead to shop-floor staff being diverted from their duties; such as shelf-filling; leading to increased ‘off-sales’, reduced turn-over and unhappy shoppers.
Data is collected in real-time and can be forwarded to a real-time dashboard application in order to generate proactive alerts prior to the front-end breakdown. A straightforward predictive model can be created from store arrival data and the current front-end status, i.e. number of open lanes.
Figure 3 below is a simple example of the relationship between store arrivals and the impact on future quality of service at the front-end. What the graph illustrates is when the total number of open lanes remains flat and arrivals experiences an increase, the percentage of lanes out-of-compliance significantly increases in the future.
With a proactive alerting system, it is possible to alert store staff to future breakdowns in front-end service, prior to there occurrences. The alerting system is self contained and does not need input from any external systems.
Real-time management results in the following improvements.
- Productivity increases achieved by redirecting front-end staff to other activities when front-end demand decreases
- Improved customer service and satisfaction by proactively addressing and reducing front-end service breakdowns
- Productivity increases achieved by enabling front-end management to focus on higher value activities, other than service monitoring
Self-Service Resource Allocation
With the advent of self-checkout (SC) systems, retailers are faced with a number of questions:
- How many SCO lanes to deploy?
- What is the impact on the service levels of standard checkout lanes?
This understanding is used to continually improve service operations, staffing allocations and store design initiatives.
Based on the metrics delivered from the Brickstream queue measurement solution, it is now possible to make these decisions based on individual store performance or groups of like stores. Performance data can be input into store resource allocation models in order to select the optimal configuration that balances investment in hardware and customer service objectives.
Basic questions with respect to resource allocation can now be answered and better managed based on service level information.
- Are SC systems under utilized because standard lanes are overstaffed?
- Are customs at standard lanes encountering longer wait-times as a result of the replacement of lanes with SC systems?
- Are my customers migrating to SC systems as I expected?
Data Collection System
At the heart of the Brickstream data collection platform is the ClarityTM. Clarity is an intelligent video analytics appliance equipped with stereo-vision and advanced object tracking algorithms. This unique combination of stereo-vision and object tracking allow the Clarity to provide accurate and consistent tracking of people, as they pass through the field of view. The Clarity has the ability to distinguish between adults, children, and shopping trolleys, as well as other objects.
Track and behavioural analytics are embedded within Clarity sensor, in order to automatically transform visual track data into specific business metrics; i.e. the number of customers in queues, individual customer wait-time and detecting the presence of a cashier at a till. For queue measurement, the interaction of object tracks is analysed in order to filter out customers that may just be passing through the queue area or shopping near the product displays.
Each Clarity appliance is fully autonomous; processing, storing and forwarding data every 60 seconds. The data-stream can also be transmitted to an in-store server running real-time applications, typically at a frequency of every 2 seconds, in order to provide instantaneous and predictive information with respect to front-end performance and real-time alerting to assist in real-time front-end management.
The Clarity is a Power-over-Ethernet IP addressable device with an embedded web server, providing both local and remote configuration, support and management. Figure 4, right is an actual snapshot of the Clarity interface. In this picture, the system is showing two queues; monitored via a single Clarity, outlined by the green and blue lines, as well as the status and various statistics with respect to each queue. Data transmitted from the Clarity is stored in a centralised data warehouse at a granularity of five minute time periods.
Brickstream Portal
The Brickstream Portal is accessible anywhere, anytime via the web. The Brickstream Portal provides a web-based front end to the central data warehouse where your data is stored. Regularly used queries can be saved or new ones easily created, allowing an authorised user to zero in on key metrics.
Figure 5 is an example report. The Brickstream Portal is implemented on the Microstrategy Business Intelligence platform. The central data warehouse has been designed for multi-dimensional analysis, facilitating integration with other BI platforms, like Business Objects.
Brickstream Dashboard
The Brickstream Dashboard is a real-time, browser-based monitoring tool used by store managers, to immediately identify and improve customer service issues. The Dashboard uses both real-time data and predictive algorithms to proactively alert managers of customer service issues. This allows stores to correct customer service problems when or before they occur.
Figure 6 is a snapshot of the real-time dashboard application, which provides the following information for each group lanes.
- Action – number of lanes to open or close
- Average queue length across the group
- Number of lanes open
- Percentage of lanes out of compliance
- Percentage of lanes idle
- In this example, a graph is associated with each metric showing the trend over the last 60 minutes. A store level view provides the status of each individual lane and the total store arrivals over the last hour.
Alerts, as well as the application, can be received and viewed on handheld devices. Dashboard applications are built and deployed as Java applets or Flash applications and are easily customisable.
About Brickstream
Brickstream provides next generation business intelligence solutions that enable retailers, financial services institutions, and consumer packaged goods manufacturers to collect and analyse in-store customer behaviour, in real-time. This understanding is used to continually improve service operations, staffing allocations and store design initiatives. The result is a better in-store marketing experience, driving improvements in customer satisfaction, loyalty, and value.
