It may seem that artificial intelligence has reached a critical mass, but it has not. In fact, it is only beginning to have an impact in some sectors, including retail. But, according to findings compiled by KPMG, retail AI has room to grow, and much more. And by 2027, artificial intelligence in retail will skyrocket to $ 19.9 billion from around $ 7.3 billion in planned spending in 2022, according to meticulous research.
All of this, and only half of retail professionals believe they have scratched the surface of what is possible when technology meets in-person shopping. So why the lag time despite the potential of AI? Blame it on the confusion around AI in general.
What is and what is not AI
Many people don’t understand AI conceptually. This leaves them less likely to invest in emerging technology, even if they see that it works for e-commerce. Or they feel that AI is limited to robots storing shelves.
However, AI is simpler than many retailers imagine.
In essence, AI algorithms are mere “if-then” statements. As long as the outcome parameters are set, the schedule collects, evaluates, and uses the data appropriately. And if-then situations happen all the time in retail.
An example of “if-then” in retail
Let’s say the manager of a grocery store stands in the checkout lines. When there are more than three late customers, the manager opens a new record to make the customers happier.
In other words, there are hundreds, if not more, issues that crop up in the retail scene that need to be addressed by managers to keep customers happy and the process to run smoothly.
With AI, you could eliminate the need for the manager to hang around to keep things moving. Instead, the store’s cameras or sensors could get the job done. That way, managers can take care of other matters during business hours and beyond. At the same time, the data collected by the cameras could go through more if-then statements.
Yes the store is busy every day at 3:00 pm and customers are waiting angrily in line, so we need more cashiers at 3:00 pm every day.
Data-driven decision making
Let’s take the situation one step further. The AI sensor could store incoming data and measure the average customer wait time. Those averages could help the manager know when employees were most needed to take care of overloaded pay lines. When do most cleaning situations occur?
There is virtually no limit to the doors that AI software can open.
In Australia, IA fashion stands are measure customers’ body language and mood to make clothing suggestions. At Starbucks, artificial intelligence is used to track the best-selling beers and customize special offers. What special offers are most requested and received by customers?
Other retail companies are improving their warehouse management, assigning machines “heavy duty” tasks in warehouses. Some stores do a deep analysis by seeing if customers spend more when they turn right or left when entering the store. By knowing which island your customers spend the most time on, you can better plan spending on exhibits on those islands, as well as promotional coupons.
Preparing for AI in Retail: Obstacles Retailers Face
One thing is for sure: AI can be a powerful retail tool. However, it is not without its obstacles. Fortunately, most of the obstacles to adopting AI technologies can be overcome by asking (and answering) a few questions.
1. Why do we want to use AI?
This may seem like a trick question. Is not. It is ethical. Retailers must be clear about why they want AI and their answers must make sense. Case in point: if they are using AI to improve customer retail experience, – great. On the other hand, if they’re driving sales through AI-powered fear generation, that’s inappropriate.
Everyone needs to have a moral foundation regarding the use of AI. Its potential for good is enormous. But when used for the wrong reasons, it can cause great harm. Therefore, the correct answer to this question should focus on service and security.
2. Which of our processes could benefit from AI?
You can’t explore all the possibilities of AI if you don’t understand where its bottlenecks are. Think back to when you were 15 years old and you worked at a fast food place. I changed the sign on the marquee regularly. How did I know what to write? Someone from the company would fax a company-wide memorandum to my franchise owner. Then my manager would review the fax and deliver it to me. It’s not exactly an optimized system, is it?
With artificial intelligence, a person can program a digital sign to activate or even schedule the sign to change based on anything from time of day to weather conditions to on-site sales.
When you are considering implementing AI in your retail store, start by thinking about what your algorithms would look like in an analog way; get help if you need it, don’t miss this opportunity. For example, where do you routinely collect and disseminate information? Those are probably areas that could be sped up if you hand them over to AI.
3. Who should help us implement our AI solutions?
When considering this question, did you automatically think, “an AI expert or an IT person?” That is what most retailers assume, but it is not true. The best person to help you with your AI applications is an operational efficiency expert. This type of professional will strive to understand your business processes and your store, in order to design a satisfactory artificial intelligence solution for your if-then statements.
You’ll know you’ve found the right partner when you get all kinds of additional questions.
These questions are likely to include inquiries about what type of information you currently collect, what digital processes you could automate, and how you intend to use the data collected by AI to make improvements to your retail systems that are already in place.
It’s hard to know how far AI will go in retail. However, it is clear that the way consumers shop and the way stores go about their daily business will change. So even if you’ve been slow to embrace AI retail solutions, now is the time to put your doubts aside and jump on the bandwagon.
Image credit: markus spiske; unpack thank you!