Since Elkjøp implemented Quinyx for Workforce Management, they have managed to reduce administration for their managers resulting in 11 000 hours of time savings for the store managers on a yearly basis. This equals 50 hours in admin time per store and year.
They are now turning to AI Optimization to further improve their store efficiency. After a successful Proof-of-Value with Quinyx, Elkjop has now decided to do a wider test of AI-powered workforce management within their organization.
Morten Kirk Schwartzmann, the Nordic Productivity Manager of Elkjøp explains:
“The implementation of Quinyx workforce management has been smooth. We have successfully been tracking against our goal to reduce administration, that is a goal we did set up when deciding to implement a new tool for scheduling, time & attendance and employee communication. As a next step, we further want to optimize the way we man and staff our stores and to do that we want to try AI as an enabling technology”.
Managing staff, forecasting future demand and creating efficient schedules for a global retail chain is a complex business, and to give readers an understanding of the challenges we can divide them into three buckets:
- Business demands - to create a good schedule you first need to understand the demand in terms of customers visiting the stores and their behavior. Then you need to understand the number of headcounts needed, and what competence is requested (Sales, Support, Cashier, etc), and how long-time certain tasks consume (labor standards).
- Labor Laws - retailers operating in several countries need to take local labor laws into consideration and ensure that schedules don't violate any regulations. This could be the minimum amount of rest between shifts, maximum working hours per week, etc. If these Labor Laws are being compromised, penalties could be the result.
- Employee happiness - we also need to listen to our employees and their personal needs. They might request to leave early on Thursdays to pick up their kids from school, or to work together with their favorite colleague on weekends.
Take all of these parameters into consideration, and multiply them with 10, 100, or 1000 employees and you can quickly imagine the difficulties to create optimal schedules.
The road ahead with AI and data-driven scheduling
“We have recently carried out a data exercise together with the folks at Quinyx, and with the use of AI we could see that we have the potential to further improve our forecasting and manning.“
“As you can imagine, we neither want too many or too few people in the store. If we have too many people on the schedule, our salary costs will be higher than expected. And if we have too few, it will hurt our customer experience and sales negatively.”
“Finding the right balance is the holy grail, and without proper data - it is close to impossible to decide if we should add more headcounts or not. In the data exercise, we could see periods of understaffing and overstaffing. For example, we could see 20% overstaffing at some points, and 30% understaffing during certain time slots. These variations we want to minimize.”
“With the use of AI, we believe there is room to improve the forecasting accuracy, maybe up to 50%. And this is what we are now about to pilot together with Quinyx.”
Elkjøp is the largest consumer electronics retailer in the Nordic countries, with over 410 stores (including franchises) across six countries and approx. 11,000 employees. The company is owned by British group Dixons Carphone PLC.