Two years after the pandemic turmoil, retail is finally ready for a long-awaited reset that may help retailers take a more stable stance. Getting there, however, will require rethinking the old ways, especially when it comes to the sector’s relationship with digital systems.
Back in the day, retailers used operational data to assess employee performance and dig into customer behavior trends to then take corrective measures. But in 2022, this approach is rather reactive. The exponentially growing volumes of enterprise data — half of which still goes unused for analytics — are calling for retailers to adopt new ways of managing and analyzing their data.
Below, we explore how AI-driven data analysis solutions are helping forward-thinking retail companies combat the fallout of the pandemic, keep operating costs under control and win back customers.
AI-Driven, Company-Wide Data Analytics: Retail’s Next Frontier
Currently, there are three types of data solutions on the market:
- Software solutions that solve a specific business task, like visualizing enterprise data, locating SKUs, ensuring data interoperability between enterprise systems, preventing fraud, and more. Examples of such software include Power BI, Tableau, RabbitMQ, and others.
- SaaS platforms that allow managing all supply chain, sales, and customer data via a single platform. Examples of such platforms include SAP, Salesforce, Snowflake, and more.
- Enterprise-wide data ecosystems comprising enterprise IT infrastructure, business applications, and AI-driven analytics tools that aggregate, process, store, analyze, and visualize business data, as well as trigger automatic actions relying on the gained insights.
Due to the further development of artificial intelligence for retail and the democratization of data, enterprise-wide data ecosystems are winning over traditional business intelligence and analytics tools:
- By running structured and unstructured data from internal and external sources through artificial intelligence algorithms, retailers can identify recurring patterns and proactively foresee scenarios that may affect critical operations.
- By granting access to AI-driven insights to every employee regardless of their technical background, retailers remove operational bottlenecks and support employee decision-making with machine-generated insights.
The Promise AI-Driven Data Ecosystems Hold for the Future of Retail
During the pandemic, retail has faced its share of challenges. With a 10.5% decline in sales and 12,200 brick-and-mortar stores shutting down in the US alone, retailers turned to technology to keep their businesses afloat. AI-driven data ecosystems have become a basis for company-wide IT system deployments.
Researchers find that the global market of retail AI software will reach $27 billion by 2027, with a year-over-year growth rate of 29.7%.
And though retail tech investments have decreased by 43% from the previous peak in Q4 2021 (mainly due to the rebound of brick-and-mortar stores and retailers foregoing putting much money into ecommerce solutions), totaling $13.8 billion in Q3 2022, the sector continues to invest in in-store technology, systems connecting physical and digital shopping experiences, as well as logistics and supply chain management software.
The more often retailers adopt digital systems, the more valuable information they get access to. They can aggregate, analyze, and capitalize on this data, using it to reduce operational costs, attract new customers, and identify opportunities for growth.
Top 3 Goals Retailers Achieve with AI-Driven Data Ecosystems
Balance supply and demand
Retailers globally have leveraged AI-driven systems to combat the fallout of the COVID-19 crisis.
Zara, a global fast-fashion retailer, made $254 million in net sales in Q2 2020 and reopened 98% of its brick-and-mortar stores by adopting an enterprise-wide digital transformation strategy with customer intelligence, personalization, and data-driven supply chain management at its core. The company managed to avoid overproduction and overstocking by using demand data to manufacture the next batch, dynamically tracking each SKU, and redistributing items across stores.
Another fast-fashion retailer, H&M relies on artificial intelligence to track purchases across stores, predict trends, and restock according to demand.
Europe’s largest online retailer, Zalando, uses AI to find a higher price point that preserves sales in the long run. The software comes up with the optimal price for the company to sell the same number of items but at a higher margin. With dynamic, informed price updates and regular deals and discounts, the retailer wins more, while keeping an overall basket cost balanced.
Similarly, Albertsons, one of the US’s largest food and drug retailers, relies on artificial intelligence to personalize pricing. The company uses historical sales data to predict how customers would respond to price changes. The retailer uses the insights to adjust the prices dynamically, if a predicted reaction is negative, to retain customers.
Pinpoint opportunities for cost reduction
Tesco, a British supermarket chain, powered in-store refrigerators with sensors to track temperature data over the year. They found out that the refrigerators had been running at a temperature lower than necessary. The insights helped the company achieve an energy cost reduction of 20%.
Another example comes from the US. New York’s Westside Market pioneered the scan-and-go technology paired with a mobile app. The solution automatically identifies items in a shopping cart and prompts shoppers to pay for the goods by scanning barcodes that pop up on the screen. This allowed the retailer to cut staffing costs and retain in-store visitors amid the pandemic by minimizing human interaction.
What does it take to build an AI-driven, enterprise-wide data analytics solution?
Although artificial intelligence has been on the radar for a long time, most businesses haven’t jumped on the opportunities it offers just yet. So, venturing into retail software development to adopt an intelligent data ecosystem, it is better to start small and opt for a trial-and-error approach until a full-featured solution is developed.
To successfully implement an AI-based data solution, retailers need to:
- Identify business problems they are trying to solve by adopting technology
- Design a vendor-neutral, technology-agnostic architecture that would allow any stakeholder to access and manage the future system’s data
- Craft a data governance strategy highlighting quality, security, and management of the enterprise data, as well as covering user roles and permissions
- Identify and prioritize use cases
- Build data components for high-priority, low-risk use cases that will make up an MVP. Make sure it allows for scaling and building on more advanced functionality and expanding to more use cases
- Get feedback from stakeholders and tweak the solution accordingly
- Repeat the steps above until a comprehensive data ecosystem is developed.
Opting for iterative development allows retailers to start generating value from day one without high upfront investments. This approach is a good fit for both enterprises and smaller businesses that want to test out AI waters. Still, regardless of the company’s size or profitability, tapping into company-wide data interoperability and perfecting data analytics will allow businesses to balance near-term challenges and long-term objectives.