Argility Technology Group, a Digicloud Africa Google Cloud partner, is helping local retailers transform their operations, grow margins and become more competitive using solutions built on the Google Cloud.
Argility CTO Marko Salic says advanced digital technologies are key to enabling large retailers to overcome today's top challenges. “Local retailers are grappling with supply chain disruptions and increasing supplier costs that erode their margins, so they have to find ways to make their operations more efficient. They also have to overcome the customer loyalty issue – modern consumers switch stores at the click of a button. Retailers know they need to improve customer experience to differentiate in a competitive market, but understanding what that means and actually achieving it can be difficult,” he says.
To help large retailers overcome these challenges, Argility has built software and technology to transform business and increase margins. “A 1% increase in revenue or reduction in costs is significant for retailers turning over billions of rand,” Salic notes. “Our technology is helping retailers achieve improvements of up to 6%.”
Argility solutions help retailers harness predictive and prescriptive analytics to process data at a massive scale, to support better decision-making and make processes more efficient. “It allows large retailers to innovate, become more competitive and improve margins by doing more with the same resources, or the same work with fewer resources,” he says.
Salic outlines Argility solutions designed to make predictive analytics easier for retailers.
“PredictRetail is our AI-powered pricing, inventory, customer and sales analytics platform focused on high-value use cases of machine learning and predictive analytics for retailers and brands. Its PredictPrice module uses historical sales data, complemented by scraping and matching competitor prices, to analyse price elasticity of demand and determine optimal pricing strategies. It allows retailers to identify the sweet spot of what customers will be willing to pay,” he explains.
“The PredictInventory module helps forecast and segment more accurately, so retailers can more accurately predict what will sell in which store on which day, predict earlier when stock will run out and forecast which stock won’t be sold at all. The gains depend on the maturity of business, but for the vast majority that still rely on spreadsheets built up over decades, using machine learning algorithms for inventory prediction can improve accuracy by around 10%. PredictInventory can improve accuracy by 2%-5% for those using specialised inventory tools.”
The PredictCustomer module enables personalisation using a recommendation engine offering products customers are likely to want. Salic says recommendations – crucial for the success of retail giants such as Amazon – can be used to improve sales wherever organisations have customer data. “This could be through online shopping or loyalty programmes. With personalisation and recommendations, retailers can significantly enhance sales and the customer experience. In a recent implementation in India, our customer increased sales volumes by 6.3% based on personalisation.”
Argility says the Google Cloud, BigQuery enterprise data warehouse and the Vertex AI machine learning (ML) platform underpin the company’s ability to innovate and achieve faster time to value for its customers.
Salic says: “The entire platform has been developed as an extension of Google Cloud, using their data analytics tools and services. At the platform’s core is BigQuery, which hosts all our input and output data. The BigQuery TCO was lower than its competitors, and because it is a completely managed service, it allows us to focus on business requirements instead of managing infrastructure.”
“BigQuery is incredibly scalable. This is important for us because we work with massive volumes of data – tens of billions of rows per month.”
Another key advantage of BigQuery is that it uses SQL syntax, which meant that Argility’s team could lean on their decades of experience and get straight to work. “They didn’t need to retrain and reskill, they could just hit the ground running,” Salic says.
Argility uses Vertex AI to manage the end-to-end machine learning life cycle. Salic says: “We wanted one platform, a managed service we could deploy, and start implementing the business logic. So by combining these two platforms, we could do the job of 20 people with just 10, and we no longer needed large infrastructure teams. We shut down our two data centres and moved everything to the Google Cloud, resulting in significant savings. This has enabled us to become more competitive, develop and deploy solutions faster, and achieve faster time to value.”
Argility aims to continue harnessing Google innovation to improve its portfolio. For example, Salic says Google’s image recognition models could prove useful for the group’s Smollan business, which focuses on point-of-purchase retail solutions. “For big brands and retailers, merchandising is crucial. To ensure the shelf displays and pricing are correct, merchandisers and shelf packers take photos and feed them back to the servers; however, ensuring that the photos match the planograms can be complex and costly. Integrating this process with a Google model could simplify our processes significantly,” he says.
Share