Re-defining the retail value chain
Let’s look at a fast-growing global fashion retailer. Like many of their peers, they were experiencing excessive store-level inventory at the end of each season.
Traditional methods of clearing inventory, such as promotions and mark-downs, were leading to revenue loss and missed opportunities. Pre-season and centrally managed sales forecasts proved to be time-consuming, and inaccurate. Forecasting was largely based on historical data, experience, and intuition of individual sales-planning professionals. Furthermore, long planning horizons and siloed manual planning with delayed market responses, contributed to suboptimal results.
The retailer was looking to increase sales by significantly improving their forecasting accuracy, to leverage intelligent product segmentation in order to tailor supply chain fulfillment strategies, and to optimize their supplier network.
With advanced analytics and machine learning, they were able to drop forecasting error rates significantly—from >90 percent to around 30 percent. This had a significant positive margin and brand image impact by eliminating excessive season-end promotions and mark-downs. And 20 percent of their production capacity was reallocated closer to the consumer markets using network and flow path optimization. In total, the retailer was able to achieve >$200m in annual benefits.
Enabling supply chain transformation with intelligent technologies
By putting data at the core of their operating model, supply chain organizations can develop powerful new capabilities, processes, and metrics. Leaders who make the transformation increase their forecasting accuracy, identify and resolve issues in real time, and create new segmentations that enable them to deliver on consumer requirements with speed, specificity, and scale.
This transformation is enabled in the following ways:
Establishing end-to-end visibility and centralizing control
Intelligent technologies make it possible to manage supply chain complexity. They capture, process, and utilize vast sets of structured and unstructured data to provide real-time visibility. Supply chain organizations will leverage this power by creating “control towers” to centralize data and decision-making. New network-based planners will drive internal, vendor, and customer collaboration to align expectations, develop plans, and manage exceptions.
Creating new performance engines
AI and machine learning enable powerful resolution engines. These are based on real-time, root-cause analyses, to automate the execution of supply chain functions, as well as optimize transactions to meet strategic objectives. AI can process data in enormous quantities to perform real-time, what-if analyses and supply chain management (SCM) modeling to optimize the supply chain along more variables than ever before. When a significant change occurs, the engine will determine the impact on key performance indicators (KPIs) and make immediate supply chain decisions that help achieve business outcomes and optimize financial results.
Managing through agile decision-making
Currently, supply chain activities are driven by sales and operations planning cycles, and executed manually. Leveraging the power of intelligent technologies, future management models can be fully collaborative, data driven, and platform based. On an ongoing basis, participants can share qualitative information and real-time data from the supply chain systems, review reports, and discuss implications.
As exceptions are identified or opportunities arise, planners can create resolution options, share with stakeholders, discuss on the collaborative platform, and take immediate action. Additionally, all qualitative information that is currently being exchanged—and lost—in phone calls and emails is saved along with the changes. As such, sales and operational planning will transform from manual and linear to continuous and collaborative.
Developing a personalized and flexible approach
AI’s computer power makes it possible to create more and more granular segmentation strategies that address consumers’ personalized needs by channel, service level and locality. In addition, real-time visibility into market data will produce greater insight, variation, and urgency of understanding and meeting demand requirements. This empowers supply chain organizations to rapidly develop multiple supply chain models for a single, integrated network.
Unifying and amplifying performance management
Advanced analytics enable a new way to measure and manage performance. In addition to tracking traditional functional metrics, supply chain organizations can determine impact on ultimate business outcomes. They can develop new end-to-end metrics to measure performance and cost across multiple functions, such as procure-to-pay and order-to-cash. Additionally, using predictive scenario modeling, organizations can evaluate the business impact of various options, like corrective plans for missed supplier shipments or new promotional campaigns.
Ongoing optimization to achieve outcomes
Machine learning enables the attainment of goals over time by self-learning, predicting, prescribing, and optimizing supply chain performance automatically across the functions. Automation can flag and resolve exceptions in real time. Machine learning-based algorithms can predict these exceptions and supply chain outcomes. As the nature of the exception or resolution process changes over time, cognitive computing learns and adapts to it. This enables supply chains to handle more complexity, making them more dynamic, flexible, adaptive, and efficient.
Intelligent technologies—fueled by end-to-end data—can add essential value to a supply chain organization. Not only do they unify the supply chain, creating new efficiencies and operational capabilities, they unlock capital to reinvest in new business models that enrich customer experiences, build competitive advantage, and support profitable growth.