Who wants a Real-Time Supply Chain?

Blog Data Science

Everyone, it seems. Take Procter & Gamble. In a recent talk, Procter & Gamble’s SVP of Product Supply, mentioned they have created a “real-time instrumented supply chain,” which they believe could achieve an upside of 1-2% sales increase, 2-5% margin improvement, and 5-10% improvement in asset utilization.

Only several years ago companies updated their supply chain plans approximately once a month, whereas today forecasts and plans are adjusted twice a day for some product categories. Such frequent updates enable responding much faster to changing demand and allow implementing a more accurate resupply of products to stores.

Nevertheless, achieving a real-time supply chain is not trivial and typically involves revisiting the deployed technology stack.


The hurdles on the way to real-time supply chain

During the past decade various events such as the rise of Amazon and the likes, the birth of Black Friday, global sourcing and the omni-channel approach, have revolutionized the way customers are served. Consequently supply chains have been pressured to lower cost, achieve faster delivery times, improve service levels and be more agile.

Access to real-time data is key in this process as it enables instant-decision making and real-time logistics solutions. Supply chain managers can seize time-sensitive opportunities or avert the consequences of unplanned events.

What, then, is preventing companies from implementing a real-time supply chain?

Data silos. Many companies and business units still operate in data silos, using numerous legacy systems. Such systems are typically disconnected, each with its unique deployment. While each system optimizes a specific part of the chain, it cannot assist in managing the impact on the rest of the chain. Manual workarounds must be implemented to obtain the consolidated indicators/insights required to run the day-to-day business and make operational decisions.

Less than optimal software. It’s surprising how many companies entrust a mission-critical task to desktop software like Excel® or Access®, which to be fair, were never designed to process volumes of fast-moving data on a continuous basis. As a result, many processes are error prone and time consuming. It may take hours, if not days, to assess the impact of transport delays, order changes, shortages on lead times, and order fulfillment.

Many alerts – but not the right ones. Commonly, there is never a lack of alerts in supply chain systems. Legacy systems will flood users with alerts, red cells and emails. However, these will rarely be intraday events, which impact the entire end-to-end supply chain. The fact is – users only want to be alerted when the impact of an event requires them to make an immediate decision; for example, when an inbound transport delay will affect a customer order planned for the next x days.

The inability to access data in real time leads to sub-optimized decision making, especially with forecasts that reach 80% accuracy at best. Companies are unable to swiftly synchronize their plans with the reality of execution or to respond fast enough to unplanned events.

In-memory technology as an enabler of real-time supply chains

Real-Time Supply ChainIn recent years In-memory computing has made a major impact on business analytics, allowing companies to change the way they conduct business. It provides business users with immediate real-time access to the right information for making informed decisions and answering the most complex business questions.

In-memory computing is all about speed, scale and real-time analytics. These elements are essentially the prerequisites for a real-time supply chain.

Data aggregation. In-memory computing can support the instant aggregation of data arriving from numerous operational systems (e.g. WMS, TMS, ERP, POS, Planning tools etc). It can do this incrementally, without any data latency and ensure that users always have the freshest and most detailed transactional data.

User-defined alerts. With in-memory technology, users can create their own user-defined alerts to identify outliers immediately when they occur, and focus on events that require immediate attention.

On-the-fly simulations. The speed and calculation power of in-memory technology lets users create on the fly simulations to review impact on order fulfillment rates, lead-time, performance, capacity, costs, revenue etc.

ActivePivot and Real-time Supply Chain

ActivePivot’s in-memory aggregation technology provide a powerful and flexible approach for solving data analytics challenges, allowing teams to make more informed supply chain decisions.

Exception reporting, simulations and real-time analytics support day-to-day operational decision-making. When an event occurs, e.g. change in product availability in the WMS, that affects your overall service level, your team is alerted and can immediately see the impact of the event across the entire supply chain in terms of costs, lead times or service levels.

By simulating different scenarios optimal decisions can be made regarding the next steps to be taken, to avoid any service levels issues. For example, supply planners can simulate an increase in orders of a certain product, evaluate the impact on raw material inventory levels (current and projected), and instantly adjust/recalculate the optimal supply plan accordingly.

Learn how ActivePivot’s in-memory analytics can help you Discover, Sense and Respond to unplanned demand and supply events.