Analysis Of Schedule Effectiveness Of Belk Company
Introduction
Starting with a first store opened in 1888, Belk Inc. became the largest family-owned and -operated department store retailer in the United States. Recently, the company launched human capital management (HCM) initiative in order to reduce time spent by their employees on activities not related to sales. As a part of this initiative, Belk implemented workforce management software solution to automate scheduling system. Instead of traditional manual scheduling, optimal schedules were now generated by the software based on historical sales and complex algorithms. Despite huge investment, the new scheduling system was not as efficient as the firm expected. More than 60% of schedules generated by the software were edited by store managers in spite of high “Schedule effectiveness”. Therefore, managing such a large volume of edits was still extremely resource consuming and required needless duplication of efforts. Consequently, before further expansion of automated scheduling system, Belk needs to reconsider its implementation process. First, the software should be improved and better customized to account for different aspects of scheduling in Belk’s department stores. Then, Belk should better understand stores characteristics and use different metrics to assess the effectiveness of its digital scheduling. Finally, Belk has to involve its employees into the implementation process in order to prevent resistance.
Opportunities and challenges
In is clear that an automated scheduling system offers plenty of opportunities to improve services and customer experience that allow Belk to effectively compete with discount stores and online retailers. However, replacing almost entirely manual scheduling system by an automated one is not without challenges due to employees’ resistance and technical constraints.
Matching demand and labour availability
First of all, automated scheduling system improves Belk’s ability to match store labour with consumer demand that is crucial for its performance. With a sufficient amount of labour, store employees can provide shoppers with an appropriate assistance as well as execute properly operational activities . Previous Belk’s scheduling system was suspected to be based on arrangements and negotiations between store employees and local managers. These arrangements could result in a significant mismatch between demand and labour availability due to unproductive deviance or favoritism. The right amount of labour available in the store is beneficial for both consumer and retailer. Even though overstaffing may have positive effect on sales because of better customer service, it significantly increases labour costs and reduces retailer’s profits. However, understaffing may lead to consumer dissatisfaction. Prior research showed that 33% of shoppers encountered a problem to find a salesperson in the store and, on average, 6 % of shoppers are lost because of the lack of service. In addition, dissatisfied consumer may switch to competitors or express their dissatisfaction on social media resulting in negative world-of-mouth effect that undermines retailer’s competitive position. Moreover, understaffing increases employee dissatisfaction due to stress and overtime hours and, in such way, further deteriorates store performance. Thus, an automated scheduling system, with expected sales forecasted at 15-minute intervals, can significantly reduce labour-demand mismatch and potentially improve store performance. However, as Bass, Director of Workforce Management Implementation, pointed out, sales benefits of such system were difficult to quantify since various factors can contribute simultaneously to the increase in sales. Nevertheless, Mani et al. (2015) find out that by reducing understaffing, retailer can increase its sales by 8. 56%, profitability by 7. 02% and conversion rate by 1. 95% during peak hours.
Increasing flexibility
Furthermore, the automated scheduling system allows Belk to increase flexibility of its labour resources. For instance, the system can increase schedule effectiveness by transforming traditional seven-hour shifts into more flexible part-time shifts. This change allows Belk to increase or reduce its capacity during the day in order to meet the demand. Such flexibility is particularly beneficial in retail industry where seasonality has large impact on the sales. However, higher number of flexible resources in the store can increase the effort necessary to coordinate them.
In addition, optimal schedules generated by software could require an employee to move between different departments in order to improve efficiency. However, switching employees may also result in service deterioration. Even though software considers employee performance and aptitude to generate optimal schedules, flexible workers can be less productive or less engaged in the activities than full-time workers. Moreover, if the task involves interaction with customers, less experience employee may seriously harm store’s reputation. Kesavan et al. (2014) showed that flexible labour resources impact positively sales and profitability. However, due to inverted U-shaped relationship between flexible labour and sales and profitability, it is possible to overuse flexible labour resources and deteriorate store’s performance.
Facing competition
Retail industry has always been extremely competitive. Nowadays it is even more challenging for department stores, such as Belk, to compete with increasing number of discount retailers and e-commerce platforms. However, provision of in-person outstanding customer service and lower labour cost can help Belk to face competition from online and discount retailers. Workforce management solution, if implemented correctly, can reduce the time employees spend on non-sales-related tasks and, in such way, increase labour availability in stores. In addition, sales associates can spend more time to improve their skills and capabilities. Therefore, employees can provide customers with better assistance and improve their shopping experience. Moreover, if the software functioned properly, Belk could reduce its labour cost, that are particularly high in retail industry, and reduce prices. Wages accounted for 13% of industry revenue in 2013 and this cost was expected to grow in the future. Belk’s traditional manual scheduling system required at least one full time scheduler in 2013 in each of its 300 stores. With schedules automatically generated by the software, Belk can reduce number of employees or fully dedicate them to customer services. Moreover, Belk can obtain additional payroll savings by eliminating existing manual technology abuses such as time variances in employee meals or punching system.
Opportunity of IT expansion
Once automated scheduling system works smoothly, with data gathered by the software, Belk will have an opportunity to further expand workforce management solution into other aspects of employee’s relationships with the firm. For instance, a software can be used to develop dynamic trainings, incentive schemes and performance evaluation tools. Since the market for HCM services was growing extremely fast, with new companies entering the market, the price for IT services could decrease in future allowing Belk to efficiently introduce more digital solutions in its operational activities.
Dealing with employees’ reaction
As mentioned in the case, workforce management software meets some resistance at store-level, in terms of complaints and edits, that undermine the initial goal of the project to improve effectiveness and employee satisfaction. It is clear that after years of traditional labour scheduling practices, based on one-to-one communication with local store management, a centralized and automated schedule system imposed top down may be perceived as a threat by local managers and store workers.
In order to preserve local managers autonomy and authority, Belk’s leadership decided to give them a privilege to edit the schedule generated by the software. Nevertheless, an extremely large volume of edits resulted in a lot of needless duplicated efforts, especially in large stores, without reducing misallocation of Belk’s resources. In addition, store’s employees complained about the system that profoundly modified their traditional shifts. Kottler (1995) pointed out that a change in organization is particularly difficult to accommodate by employees when they do not understand why a change is necessary, how it will affect their job and their input to the change implementation is not considered. Therefore, complains and large volume of edits that Belk encountered from store employees can be a result of lack of knowledge about automated scheduling system and its features. Without deep understanding of the change, externally imposed floating hours, volatile shifts and change of departments can demotivate employees and hurt store performance.
Technical and Scheduling constraints
Schedules generated by the software are not 100% effective due to some scheduling constraints. For instance, since the demand was forecasted at 15 minute-intervals, it could not be perfectly matched with labour supply. Moreover, efficient scheduling could also require extremely short length shifts. Mani et al. (2015) showed how reducing shift length allows to reduce understaffing and increases sales and profitability. However, reduced shift length can harm worker welfare. In addition, government or union regulations often impose minimum number of hours per shift that limit flexibility. In addition, the software implementation was characterized by some technical challenges. The system generated a lot of software bugs. Moreover, system could not account for shift rotation for workers and asymmetrical impact of underallocation and overallocation of labour. Therefore, software features have to be improved before further expansion of digital scheduling.
Understanding the edits In order to measure how successful is the optimal scheduling generated by the software, Belk used “Schedule effectiveness” metric that compares the amount of labour scheduled to the amount of labour required to support the forecasted demand. As discussed earlier, an extremely high number of automated schedules were edited by local managers. Despite average schedule effectiveness of 63% among all divisions, 69% of schedules were edited on average. Belk launched the edit survey in order to understand why edits were occurring. This study revealed that “improving effectiveness” was first edit reason for large and medium-sized stores and second among small stores. However, if mangers edited schedules only to improve schedule effectiveness, the correlation between “Schedule Effectiveness” and “% of Schedules Edited” would be negative among Belk’s stores. This is the case for Northern and Western divisions, however, the correlation is positive for Southern division. Therefore, the overall correlation among these two variables is extremely low. Thus, in Southern division the number of schedules edited increases with the effectiveness of optimal schedules. These patterns are counterintuitive, and their understanding requires a deeper analysis of stores characteristics. For instance, Southern division could include particularly large stores with hundreds of employees comparing to other divisions. Therefore, even if the schedule is effective, in order to account for shift rotation and prevent one employee from working too many nights or weekends, a high number of edits would be required. However, the information on stores’ size is not provided in the case. Consequently, Belk has to better analyze store’s characteristics among each division in order to improve its understanding of the edits. It is also conceivable that high volume of edits is explained by technical issues of the software. As mentioned previously, the software generated system bugs that had to be edited by local managers. In addition, managers had to edit automated schedules in order to account for shift rotation or asymmetrical impact of understaffing and overstaffing on store performance.
It also possible that the key measurement used by Belk to evaluate the effectiveness of scheduling is incomplete. Thus, automated schedules can be less effective on other dimensions that could explain high volume of edits. Other metrics can also help Belk to understand how managers could improve effectiveness of a schedule generated by the machine. Therefore, Belk should consider other methods to track the performance of schedules. Some examples are discussed in the following lines. For instance, even if sufficient amount of labour amount is deployed in the store to support the demand, the store can be under-or over deployed by skill. Therefore, efficient scheduling has to consider employee capability and competency in order to ensure that each employee with specific skills is scheduled at the right place. To evaluate how efficient the schedule generated by the software is on this dimension, Belk can measure “ability or grade efficiency” based on gradients in abilities and competences of different teams. One of the major issues within the software is that it treats overstaffing and understaffing as identical. Consequently, Belk can weight over-deployment and under-deployment differently when evaluating schedule effectiveness in order to account for significantly different outcomes. In addition, “Schedule Effectiveness” measure compares forecasted labour requirement to the amount of labour scheduled by the system. However, even with sales forecasted at 15-minute intervals, there might be some errors in predictions. Therefore, Belk can also evaluate how efficient the schedule is based on past sales and compare both estimates.
Action plan
After years of entirely manual scheduling system, anchored in company’s culture, it is a challenging task to move to entirely automated one. However, workforce management software solution allows Belk to significantly improve previous scheduling system through standardization and automatization of scheduling processes with no favouritism or unproductive deviance. By reducing time spent on non-sales related tasks and better alignment sales floor coverage with customer demand, employees have a better opportunity to provide customer with an outstanding service experience and increase sales. Therefore, effective management of store labour increases Belk’s ability to effectively compete on the market. However, with an extremely high number of edits to the schedules generated by the software, automated scheduling system does not necessary reduce misallocation of resources and duplication of needless efforts. Therefore, before further pursing the implementation, Belk needs to better understand and manage the edits. First, Belk needs to understand what drives the edits. This requires a comparison of edit survey results with stores’ characteristics, among different divisions. Belk should compare volume of edits with store’s size, number of full-time and part-time employees, relationship of employees with management, etc. This allows Belk to better understand the reality of store-level implementation. Then, Belk should use other metrics to evaluate the effectiveness of schedules generated by the software. This analysis can help to understand why schedules that are highly ranked on “Schedule Effectiveness” metric are still edited by local managers and how managers can improve their effectiveness.
Furthermore, Belk should further invest into improvement and customization of the software. Bugs in the system should be eliminated and software should be tailored to account for different aspects of retail business such as shift rotation or asymmetrical outcome of understaffing and overstaffing. Even when automated scheduling system will work smoothly, some manual edits would always be necessary. Therefore, edit functionality in the software should be improved, so that local managers can edit schedules in simple and automated way. For instance, the software should be able to automatically allocate the shift to another available employee when one shift is modified manually. This functionality will be particularly useful to account for shift rotation in large stores with hundreds of employees. Finally, for the change effort to be successful, it is also important to properly lead the change implementation and, in such way, prevent employee’s resistance to change. Kottler (1995), highlighted that change process goes through different phases and usually requires an important amount of time. Based on the analysis of these phases some recommendations can be given to Belk’s leadership in order to improve the implementation process:
First, Belk’s leadership needs to create a great sense of urgency for the change. This requires examining the market and competitive realities in order to identify major opportunities or potential crises. Then, leadership has to discuss this information with their employees. For instance, Belk can better communicate to their employees about increasing competition in retail industry or declining competitive position to create a strong enough sense of urgency for the change.
Afterwards, leadership needs to develop and communicate through different channels their vision of the future. This vision needs to be easy to communicate and appealing to employees. For instance, the leadership can discuss with employee how an automated scheduling system is expected improve their daily routines and consumer’s experience.
In addition, Belk should encourage its employees, not only managers, to contribute to change implementation process and give their feedback and ideas for potential improvement.
Finally, the automated scheduling needs to be rooted in corporate culture by showing employees how it helped to improve the performance and making sure that next generation of top management adhere the new approach. It is important to highlight that Belk is undergoing an important transformation. As Belk’s Senior Vice President for Staffing and Productivity, Eric Bass, underlined “Belk is trying to move from employee-centric vision, where employees could get the schedule they wanted, to consumer-centric system where we understand when consumers are coming to the stores and provide them with outstanding service. ” It clear that such change may encounter some resistance since it profoundly modifies established routines and practices. In addition, implementation and tailoring of technology can generate a lot of technical challenges. Therefore, it is normal that such transformation takes time to be entirely completed and anchored in company’s culture.
