Australian retailers must embrace the opportunities being offered by artificial intelligence (AI) technology or risk being left behind by more nimble competitors.

Evolving at an accelerating rate, AI has the potential to streamline supply-chain management and reduce operational inefficiencies. At a time when retailers are faced with rapidly changing market conditions, such benefits can have a positive impact on revenue and significantly boost the bottom line.

One of the biggest changes faced by the sector is the rapid shift from traditional bricks-and-mortar retail activity to online sales and omnichannel – some experts have suggested COVID has condensed five years of ecommerce growth into a single year. While the rate of growth is expected to slow, ecommerce is expected to grow faster than the broader economy for the next three years at least.

To succeed in this marketplace, retailers must become adept at using data (both internal and external) to forecast demand and ensuring inventory levels reflect customer tastes and trends. Traditional methods – relying on historical data and spreadsheets to plan for the future – are no longer a successful strategy.

Sector research shows that, on average, retailers are losing up to 12% of revenues due to inadequate inventory management, leading to stockouts and markdowns. This figure, scary enough as it is, does not take into account the lifetime value of a lost customer who grows frustrated and is permanently lost to a competitor. Better use of data and technology is the key.

Spreadsheet reliance

Traditionally, many retailers have relied on spreadsheets to support their forecasting and inventory management efforts. Containing details of variables such as previous sales and current stock levels, these spreadsheets are used to create forecasts that are vital for future resource allocation and planning.

In recent years, advancements in AI have taken these efforts to the next level, making some of the technology that has made some of the world’s largest retailers so successful available to all businesses. Capable of ingesting data from a wide range of sources and forecasting future demand with greater accuracy than traditional efforts, these platforms improve stock management and efficiency. The end result for retailers are a reduction in both stock-outs and dead stock, reducing lost revenue and wasted capital. As such, supply chains can be optimised, resulting in better use of capital, freeing up retailers to invest time and resources into other components of their business.

Indeed, industry experience has shown that, properly implemented, these tools can improve market forecasting accuracy by between 10 and 30%. These platforms can also assist with everything from inventory and pricing optimisation to simulating the impact of marketing promotions.

Many of the tools can be readily integrated into existing supply chain systems. This means they can easily access real-time data and use this to further improve their forecasting capabilities. It also means benefits can be enjoyed without requiring significant additional work for the retailer.

Putting AI to work

Although not yet widely used for supply-chain management and demand forecasting, AI is being put to use in other ways within many retailers. Some, for example, are using AI-powered customer service bots to deal with customer queries lodged via websites.

While this does add value, it’s far from the end game for retailers. Provided with the right sources of data, AI can deliver value in myriad other ways.

The problem, however, is that many retailers lack the high-quality data that is needed by these AI tools and using poor quality data risks compromising the outputs that are delivered. Garbage in unfortunately means garbage out.

As well as sourcing suitable data, retailers also have to make the decision between building their own AI platforms or buying the capabilities from a third-party vendor. While larger retailers may have the resources needed to commission a consulting firm to build a solution, many smaller firms do not.

For these firms, it makes more sense to purchase off-the-shelf tools and configure them for their particular requirements. As well as lowering the cost, it also means they can be put in place and become functional much more quickly.

When the decision has been made to embrace AI, the steps that should be followed include:

  • Set goals: AI can deliver benefits in a range of different ways, so be clear from the outset exactly what you are trying to achieve.
  • Determine a budget: It’s also important to understand the funding available to support the planned project. Estimate the results that are likely and how much should be invested to achieve them.
  • Start small with high Impact:  Rather than trying to re-engineer the way the entire business operates, begin with a small project. This might involve using AI to assist with a low-risk process still with a high impact pay off. Proving the technology in this way early in this journey may help to convince any skeptical colleagues while demonstrating clear ROI.
  • Constantly evolve: The capabilities of AI tools are improving all the time and are iterative by nature, so monitor your progress and make changes and further investments over time.

By taking the time to understand the opportunities offered by AI tools, Australian retailers can be in a much better position to predict and respond to changes in market demand. Making necessary investments now could lead to significant growth in the future.

Eamonn Barrett is partner at Remi AI.