Just a few short years ago Artificial Intelligence (AI) and Machine Learning (ML) adoption was slow. Now, several Australian players will be ready to deploy the technology this year, with several others launching within the next 12-18 months.

The Australian market is in line with what is happening across the globe. According to the JDA & KPMG Digital Supply Chain Investment Survey, 80 percent of CXOs see AI and ML as the most impactful technology for solving complex supply chain problems, and 75 percent of CXOs believe that cognitive and predictive analytics will have a disruptive impact in the year ahead. According to JDA Software’s 2019 Voice of the Category Manager Survey, 40 percent of retailers and manufacturers have admitted they are behind on leveraging AI and ML and nearly half of retailers and manufacturers plan to invest in customer-driven data science in the next five years.

With the adoption of AI and ML accelerating, there are three major use cases becoming increasingly common in Australia and globally.

Removing the guesswork

AI and ML collect and analyse micro data and patterns within patterns, enabling better decision making along a complex supply chain. One of the use cases of the technology that has been particularly popular in Australia is the ability to plan and forecast more effectively. This removes any guesswork by generating the most accurate forecast possible based on real time data.

By analysing both internal variables such as customer data and external variables such as consumer preferences and behaviours, social media data or weather, retailers can forecast and plan much more effectively. Better planning reduces instances of products being out of stock and ensures supply meets demand, by optimising allocation, replenishment, fulfillment, markdowns and more.

Targeting the ‘segment of one’

Forget personalisation. The next big thing is hyper-personalisation. Moving beyond simple segmentation by age, gender or location, hyper-personalisation analyses data points to provide real-time customisation of the shopper experience for the customer. Think of it as the ‘segment of one’.

AI and ML can tailor the shopper experience to each individual customer, significantly enhancing the customer experience and improving conversion. Examples in practice include customising product development for each customer, such as designing shoes for each individual foot, or optimising the customer experience around very localised shopping behaviours.

And interestingly, hyper-personalisation is more important to customers than price. They would rather spend more for a tailored customer experience.

Reducing the rate of returns

One of the major pressures on profitability for retailers, especially in apparel, is returns. The costs of return shipping, transportation and sorting through returned stock can significantly impact margins. Returns also create a significant inventory management challenge.

One area where AI and ML can address the challenge of returns is through ‘return-conscious pricing’. For items purchased at a lower price point, the rate of returns goes down. AI and ML can inform retailers on how to price products to lower return rates while maximising margin. Other data points which can influence the rate of returns include basket size, the type of product, or purchase frequency. By analysing this data, retailers are armed with better information to take informed action to address the problem. When coupled with hyper personalisation, retailers can significantly lower the rate of returns.

One of the fantastic things about AI and ML is that the time to value is short. What we’re seeing is that retailers are starting small, exploring and testing ideas, with the ability to see a real result within only 4-8 weeks. By being agile and experimenting regularly, retailers can transform their supply chains in a very short time. For retailers contemplating investing in AI and ML to optimise their supply chains, these three use cases are a great place to start.

Girish Rishi is chief executive officer at JDA Software