“It’s only natural that retailers will face an element of internal scepticism about using technology for decision making.”
An impromptu wrestling match is taking place in retailer boardrooms, something of a brains vs brawn match-up, writes Michael Feindt Founder Blue Yonder a JDA Company.
In the red corner stand the reigning champions and crowd favourites: a powerful combination of experience, human pride, hunches and intuition. The newcomers in the blue corner, a mysterious tag team made up of data-driven artificial intelligence (AI) and machine learning (ML), are being booed as they enter the ring. Despite the hostility, AI and ML are capable of outsmarting their opponents and taking over as the heavyweight decision-making champions of the world.
For all the conflicting noises you hear about AI, at the end of the day it is all about using data to make decisions with a level of certainty that humans aren’t able to reach. No matter how old or traditional the sector, how knowledgeable the human or how dependable you think your company’s operations may be, it will continually be a case of “opportunities missed” for retailers who continue to ignore AI.
Being completely data-driven is now the fine margin between success and failure. For some, it’s a big psychological barrier to let machines take over and make decisions for them, but they must do so if they want to maintain a competitive advantage.
Data-Driven Automation is The Future
Having founded a company that provides data-driven solutions across a range of industries, it’s clear retail and the supply chain are the natural next frontiers for AI and ML to break through. This is principally due to the scope of competition and the small margins in which every operator is competing to gain an advantage. Of course, an element of mathematics is already used in everyday retail; decision-makers just need to move on and start trusting algorithms to put their numbers to use.
It’s impossible to anticipate everything that will happen tomorrow concerning consumer preferences or sales levels, but you can make predictions in terms of probability distributions, thus also quantifying how certain you are that something may happen. These predictions can lead to very important decisions but it’s vital to remember that with the uncertainty that derives from human-based analysis, it’s impossible to quantify some aspects that are pivotal to the successful operations of a retailer.
For example, a supermarket chain needs around 20 million decisions to be made every single day: the price of everything in its stores, stock levels, promotions, seasonal impacts, consumer trends and more. There are so many factors that need to be addressed, and when humans make predictions, they’re based on limited data and the notion of possibilities. AI, on the other hand, is able to turn those possibilities into the provision of a probability. It will be able to predict the probability of any possible sales number of each product, and subsequentially the optimal decision of how much stock you need, how much to sell it for and where in the store to place it.
Removing uncertainty is essential — and fortunately, these machines are able to adapt to each unique business, monitoring trend curves and cost structures while consistently improving on their accuracy.
Getting Over The Pride Barrier
It’s only natural that retailers will face an element of internal scepticism about using technology for decision making. However, there are three key areas they can focus on to overcome the perceived barriers. The first is underlining the point that AI and ML will augment strategy, rather than replace it. Next, they must take the time to identify and share examples of other companies already putting AI and ML to use. Finally, they should also demonstrate how AI can remove guesswork from the equation: after all, nobody has the brain power to make 20 million decisions a day.
Intelligence In a Time of Uncertainty
The retail industry is rife with uncertainty, but AI can make predictions about some absolutely crucial areas, from stock replenishment and pricing to consumer trends and the influence of changes in the weather. AI can make these drivers not just manageable but also predictable.
While it may appear brazen to remove humans from the equation, it’s not a straight-up employment redundancy: this is an element of human decision-making being replaced with tools that can help businesses act in a smarter way. Simply put, retailers can give themselves 20 million fewer decisions to make every day.
Adopting AI and removing traditional human verdicts will require some to swallow their pride, but ultimately it could make the all-important difference. In an increasingly competitive industry, AI decision making could provide the edge that allows a retailer to pin their rival to the canvas and emerge from the ring with the world title belt around their waist.