Personally, my online shopping habits have changed significantly over the last year. I am not alone, though. Mastercard published a study where they found that 68% of South African consumers have turned to shopping more online as a direct result of the COVID-19 pandemic (read about that study here). What stands out is that 63% of respondents stated that they are consciously choosing to support online local small businesses - the “local treasures” that we fear will close.
This move has undoubtedly put stress on retailers, large and small. We, as consumers, are reacting and adapting to the uncertainty of the pandemic by shifting our behaviours. Businesses are being forced to adapt to a digital consumer whose expectations have changed dramatically.
Retail is inherently fast paced, trend focussed, and fluid in nature with an extended logistics chain and relies heavily on making predictions. These predictions will affect how retailers serve customers. Retailers are developing agile and resilient operating models based on cloud, artificial intelligence and machine learning (AI/ML) technologies at a faster rate, according to Google.
AI/ML can deliver value to your business from customer acquisition and retention to logistics to “back office” activities. These include (but are not limited to) providing customers with personalised product recommendations, planning workforce, predicting the success of a product, improving demand planning as well as cash forecasting and automating finance.
With AI/ML, your data no longer provides historical understandings, rather, data serves as the proverbial headlights. Since AI systems are using real-time data, forecasts are updated faster and more accurately. This helps you identify and mitigate risks before they make an irreversible negative impact.
To get the most out of any AI/ML model, there needs to be an upfront effort regarding gathering data. While the dataset does not need to be perfect, it does need to be similar to that which will be used during daily activities. Preparing your data can be a time-consuming and error-prone process. However, if you use tools such as Trifacta, this process will be streamlined to ensure clean and relevant data is made available.