Artificial intelligence in retail: It's the right move
Technology has long since advanced to the point where it can make decisions better than people can, and yet grocery managers are still happy to use their own experience to drive decision-making, sacrificing speed, efficiency and savings.
By optimizing key strategic areas of pricing and replenishment, and automating decisions using machine learning, retailers can combine the speed of their decisions with their KPIs (margins, volumes, mark downs). Yet still retailers are not currently marrying the two in a responsive and effective way.
As part of Blue Yonder’s recent survey of 750 grocery retailers across the globe, we asked some probing questions about decision making and customer service. The research revealed that grocery retailers believe robotics, machine learning and artificial intelligence will be some of the key game changers for the industry.
Regarding AI and machine learning: 77 percent of UK grocery retail directors said these technologies will make an impact, with 23 percent saying they already have. It was similar story for robotics with 69 percent of directors in the UK saying it will change the future of the industry, and 31 percent feeling that robotics is changing it right now.
Nearly 85 percent of UK directors said automation would help them to make faster decisions, yet in spite of this, half of grocery retail managers said that "gut feeling" still plays a major part in their decision-making process. Thirty-one percent of directors in the UK feel there are currently too many decisions to be made manually, with the same number stating that relying on gut feel is slowing them down. This suggests that while the majority of grocery retail directors know the advantages that these game changing technologies offer, they are choosing not to adopt them yet.
Despite the human brain’s natural talent for pattern recognition, and no matter how seemingly reliable "gut feeling" may appear to be, machine learning can consistently outperform established methods by taking into account multiplicity of factors that humans would not be able to. This includes things like weather affecting purchasing decisions, the level of interest in certain products on certain days of the week, and many other small factors that influence consumers purchasing habits.
The only way to create a truly accurate and holistic view of what an organization needs in order to streamline its processes (and cut costs) is to use machine learning. To choose just one example, sophisticated algorithms can take into consideration things like weather, promotions and events, which a basic spreadsheet-based system can’t factor in.
In this current time of ever shrinking margins in the grocery retail sector, new technologies offer to give retailers a crucial, tangible edge over competition. Getting replenishment wrong leads to either stock wastage or failing to provide the right product at the right time. Getting pricing wrong results in consumers looking elsewhere for better deals, more specifically, looking at which competitors are offering a better deal.
By adding in machine learning to decision-making, businesses become much more efficient. In automating replenishment, one German retailer delivered results of:
- €20m less waste from accurate replenishment projections
- €5m capital savings due to pricing optimization
- €5m efficiency savings related to streamlining processes
- 20 percent less out of stock because the right products were in the right quantity at the right time
- Out of stock issues were reduced from six percent to 0.5 percent because of how predictive tools allowed them to balance between overstocking and understocking
This is particularly important when it comes to fresh produce by nature of their limited shelf life. With fresh produce, the window in which to course-correct if an item doesn’t sell is measured in hours or days rather than months, and so the algorithms that calculate the acceptable price changes for individual items need to be incredibly agile, doubly so for popular products tied to key retail dates such as Valentine’s Day or Easter.
McKinsey and Blue Yonder recently released a report entitled "The secret to smarter fresh-food replenishment? Machine learning." In the report, it said that 40 percent of grocery revenue is driven by fresh -- get this right and the rest will follow. So for grocery retailers it is a critical barrier to success in terms of margins and customer satisfaction finding the right balance between the best freshness, availability and waste.
It is also worth considering the less obvious advantages that these new technologies offer. Removing the gut feel decision-making process also saves store managers a significant amount of time. Often it comes down to managers to forecast demand of product, orders, scheduling, labour and pricing. This results in hours spent every day on admin tasks trying to do predictive analytics in their heads.
By automating these processes, employees can return to focusing on their core capability. Store managers get back to ensuring they are delivering the best possible customer service. While it will still require monitoring, automation is simply taking the bulk of the monotonous decisions away, freeing up staff to do something that machines can’t do -- innovate and make the strategic decisions.
Where retailers once relied on making decisions manually, there has always been a limit to the number of decisions they are able to make to acceptable levels of accuracy and efficiency, the scale of which automated decision making systems have brought to light. Retailers need to stop looking at data as an abstract object and view it as a lever to create marginal gains through having the best price, best choice and best availability instead -- a proposition that has been impossible to achieve until now.
Retailers who are still discussing the benefits of machine learning and AI tech are already behind the curve -- it is here, changing the industry today. It is a technology that has the potential to significantly change how retailers and consumers interact and purchase goods.
In the short term, 2017 will be about how businesses use these new technologies to their full potential in order to make marginal gains and to reduce decision-making pressures on their staff. We will see AI being used more extensively in business to solve problems for the customer by making better decisions on a daily basis.
Prof. Michael Feindt is the brain behind Blue Yonder. His NeuroBayes algorithm was developed during his years at CERN. Michael is also a professor at the Karlsruhe Institute of Technology.
Published under license from ITProPortal.com, a Future plc Publication. All rights reserved.