How to mitigate customer pain points through implementation of digital transformation (DX) technology
What will not change in the retail industry? Even 10 years from today, customers will still want more product selections at a lower price through more convenient services. Recognizing these constant customer demands, Amazon has continued to grow sustainably by focusing on three factors: selection, price, and convenience (or SPC). In other words, Amazon has been enhancing all aspects of the customer experience—including easy product searches, one-click payments, low prices, and faster last mile delivery—to boost customer satisfaction, traffic, and sales transactions. This effort has led to an increase in registered brands and a more diverse number of products within Amazon. As a result, this effective business model has continuously driven customer satisfaction upward. Let’s have a look at the Amazon digital transformation (DX) use cases at different stages of the retail value chain.
Prerequisites: Demand forecasting and big data–based customer analysis
Amazon collects all data from its sales channels (including webpages, voice devices, and stores) and saves them in a data lake. This vast amount of data is then analyzed through demand forecasting and personalization analysis tools such as Amazon Forecast to improve understanding of customers. As these tools continue to speed up decision-making, retailers will be able to supply the products that consumers need in a more timely manner.
Amazon first applies DX technology to its supply chain management. Supply chain management generally begins with demand forecasting. By accurately predicting demand, rather than under- or overpredicting it (which can result in product unavailability or a rise in unnecessary costs incurred from excessive inventory and disposal), retailers are able to enjoy higher customer satisfaction, stock adequate inventory levels, and save costs.
Until 2007, Amazon adopted a traditional way of conducting demand forecasts. But over time, ML was introduced and deep learning was applied starting in 2015. Demand predictions became 15 times more accurate in response. To achieve such improvements in performance, Amazon employed data lake–based data analysis. Data analysis that uses AI- and ML-based Amazon Forecast systems requires three main types of data: past time series data (sales and inventory data), externally related data (weather and competitor data), and metadata of products (origin and category).
More Retail Ltd. (MRL) is one of India’s top four grocery retailers, with a revenue of several billion dollars. It has a store network of 22 hypermarkets and 624 supermarkets across India, which is supported by a supply chain of 13 distribution centers, seven fruit and vegetable collection centers, and six staples processing centers. With such a large network, it is critical for MRL to deliver the right product quality at the right economic value while meeting customer demand and keeping operational costs to a minimum. MRL used Amazon Forecast to increase their forecasting accuracy from 24 percent to 76 percent, which led to a reduction in waste of up to 30 percent in the fresh produce category, an improvement of in-stock rates from 80 percent to 90 percent, and a rise in gross profit of 25 percent.
When an order is placed with accurate demand forecasting, the product is received at an Amazon Fulfillment Center. Amazon is maximizing work and cost efficiency in its fulfillment centers by utilizing robot solutions such as Kiva, which automatically moves shelves loaded with products. Since the introduction of Kiva, the required warehouse area has decreased by 22 percent, and the operating cost of the fulfillment center has dropped by 20 percent.
DX technologies such as computer vision, AI, and ML can also identify the location and quantity of all products in Amazon Fulfillment Centers in real time. Amazon Logistics Centers apply the “random stow method,” which randomly stows products from a diverse number of categories and brands throughout the fulfillment center. This way, pickers don’t have to worry about loading products in a designated location. When picking is initiated, Kiva moves the shelf that the product is stored on so that the picker can select the appropriate products. This system, powered by DX technologies, increases picking speed and reduces picking errors while optimizing the fulfillment center space. Thanks to integrated data, there is also no need to conduct separate inventory checks.
Personalized marketing: Enhanced purchase conversion rate
Sales activities and operations provide another use case for Amazon DX technology. To enhance the customer experience, Amazon has to mitigate obstacles throughout the customer’s entire purchase journey. Amazon is using its latest devices and services—such as Alexa with Echo, Amazon One, and Amazon Go—to do just that. And in doing so, Amazon is improving its customer experience journey and reducing consumer inconvenience.
For instance, Amazon is expanding its Just Walk Out technology on Amazon Go to reduce waiting time at checkout. In 2021, Just Walk Out was applied to Amazon Fresh, a 3,300 m2 large supermarket, and to Whole Foods Market in Washington, DC. Amazon plans to continue expanding its presence in the smart store segment with Just Walk Out technologies, including Amazon One and the Smart Shopping Dash Cart. These tools help customers to pay without having to carry credit cards or mobile phones at all.
DX technology can also be used in the personalized marketing aspect of the value chain. Amazon has a purchase conversion rate of above 30 percent for Prime members. This is significantly higher than the average for general ecommerce platforms, which is approximately 3.3 percent. For ecommerce platforms to secure such high purchase conversion rates and customer loyalties, firms must be equipped with personalized recommendation systems and services. As described above, Amazon collects all available data and enhances understanding of customer needs through tools that are based on analytics—all, of course, while strictly adhering to the Amazon Web Services (AWS) Data Protection Policy.
The major Korean retailer Lotte Mart, for example, used Amazon Personalize to establish a personalized product recommendations system and develop a marketing campaign to distribute personalized coupons. This resulted in a five-fold increase in customer response rate to recommended products and a 40 percent rise in new product purchase rates.
Required for the future: A next-gen commerce platform strategy
Ecommerce interfaces first appeared in the 1990s on webpages and progressed into the mobile era in the first decade of the 21st century. Chances are high that the next generation of commerce is going to use voice-based interfaces, which will ultimately evolve into zero-effort commerce.
Amazon currently offers a variety of shopping experiences with Alexa, its voice-based Echo speaker. In the future, the market for voice commerce is expected to expand as advancements in AI technology provide enhanced personalized services.
The metaverse platform is also predicted to create additional purchasing channels. It is expected to go beyond its original function of providing a virtual space for avatars and gaming to further identify consumer purchase patterns and shop for consumers based on their purchase data. AWS can help architect metaverse platforms through real-time spatial/interactive 3D technology, image/text analysis, chatbot generation services, and more.
Responsibility for the future: ESG initiatives
Finally, Amazon has been developing environment, social, and governance (ESG) initiatives based on data and DX technology. Amazon has been particularly focused on the environmental aspect of ESG and is in the midst of developing efficient product packaging to achieve carbon neutrality.
Amazon has adopted envelope-style packaging and zero packaging to reduce the use of cardboard boxes. To achieve this, Amazon uses big data to calculate the optimal packaging for products and determine whether to ship products in envelope-size packaging or with special packaging due to fragility.
For instance, Amazon has reduced the overall packaging for toys to maximize efficiency. In 2019 alone, 400,000 packages were altered and improved. Amazon plans to decrease the box packaging ratio to 15 percent by 2030 and increase the proportion of products that need zero packaging to 50 percent. Learn more about this effort in this AWS re:Invent 2019 session.
Retailers who want to develop their own ESG businesses should first focus on driving ESG-friendly strategies in their main business areas. It is equally important to identify businesses that have enough data that can be secured to start data analysis and effectively build a smart retail business.
Learn more about digital transformation solutions for retailers on AWS.