Enhancing real-time decision-making with an Industrial Data Platform

An Industrial Data Platform (IDP) aggregates downstream market data, creating a platform for real-time decision-making. This can enable CPG (Consumer Packaged Goods) companies to achieve mastery over volatile trading conditions – and boost their bottom line.

Touchpoint

According to Gartner, by 2024, 75% of supply chain management companies will be using AI for demand forecasting. This rapid adoption is driven by the significant improvements AI brings to the table . For example, McKinsey claims that applying AI-driven forecasting to supply chain management can reduce errors by between 20 and 50 percent—and translate into a reduction in lost sales and product unavailability of up to 65 percent (!).

As with many areas of the CPG sector, data and analytics capabilities are foundational to many of the latest trends in the industry. Downstream market data can be used by CPG companies to enhance their marketing activities in a variety of ways: using retailer data to evaluate the effectiveness of their marketing spending, gaining customer-level insights, and developing value propositions tailored to individual customer segments.

However, in a supply chain context, that same downstream data can be used to ensure that the right goods get into the hands of the right customers at the right time and place. Deployed correctly, downstream data can revolutionize the operations of a CPG company in multiple ways.

Smarter Procurement: AI-powered analytics can analyze supplier performance data, identifying trends and potential risks. This enables CPG companies to optimize supplier relationships, negotiate better terms, and mitigate supply chain disruptions.

Optimized inventory: By finding a better balance between supply and demand, CPG companies can safely reduce stock levels by 20-30% while simultaneously improving on-shelf availability. McKinsey estimates that this will reduce warehousing costs by 5 to 10 percent, and administration costs by 25 to 40 percent (!).

Enhanced efficiency: real-time data on production rates, inventory levels, and demand forecasts can enable dynamic scheduling that maximizes resource utilization and minimizes bottlenecks.
Streamlined Logistics: AI-powered algorithms can analyze traffic conditions, distances, and delivery time windows to optimize delivery routes to reduce fuel consumption, improve delivery times, and lower transportation costs.

Wrangling the Data

However, while this Holy Grail holds the promise of using downstream data to transform demand forecasting and production, it remains tantalizingly out of reach for many CPG companies. Historically, they have struggled to generate value from their digital investments: a study from McKinsey showed that only 40 percent of CPG companies have achieved returns above the cost of capital for their digital and data initiatives.
It’s not hard to see why. CPG companies are looking to AI – which is only ever as good as the data on which it is trained –to deliver the benefits described above. However, extracting (and, in some cases, generating this data is hugely challenging.

Horizontal integration – in which a company acquires or merges with another in the same industry that is operating at the same level in the value chain – has been a significant feature of the CPG industry. Most CPG companies have grown in this way and often have dozens of sites in multiple countries with little consistency in their data architectures.  Many of these sites are using legacy OT systems that don’t currently integrate well with the IT systems on which data analytics applications reside. The data in these OT systems may be gathered manually, and there may even be ‘blind spots’ in these systems where no relevant data is produced.

On top of all that, production, packaging, and marketing departments may all be running different systems with little coordination between them. All of these siloes must be broken down so the data can be aggregated and viewed on a Single Pane Of Glass.

To do so, it is critical to understand where you are on your data journey and to address any problems you identify: this may include: the automation of manual data collection and the installation of new sensors and a very significant IT/OT integration project. You may also want to consider implementing a Unified Name Space (UNS): this not only creates a common vocabulary for the exchange of data across heterogeneous systems but can also create a single reference point for all data through the publish/subscribe (Pub/Sub) methodology. This enables systems to get real time data from the data streams to which they have subscribed, without having direct connections to every other system. This leads to improved scalability, decoupling of services, and faster response times to events across the network. Only at this point are you in a position to evolve into a data-driven forecasting mode.

Introducing Industrial Data Platforms (IDP)

In much the same way that a Customer Data Platform gathers data from multiple sources (social media, e-commerce and CRM systems, interaction with the website) to enable real-time decision-making around consumer behavior, an Industrial Data Platform (IDP) gathers information about supply chain dynamics to provide the same insights about your downstream operations. 
An IDP is a system that collects, stores, and analyzes data from industrial equipment and systems. Ultimately, the remit of an IDP would extend to upstream data (information about the different components – ranging from raw materials to energy supplies): this would help to reduce the exposure of CPG companies to supply chain interruptions (such as the attacks on ships passing through the Suez canal) or geo-political events like the war in Ukraine. However, today, most CPG companies are focusing their IDP initiatives on using downstream data to improve efficiency and optimize industrial processes. The core capabilities of an IDP are:

Consolidating data: an IDP integrates data from various sources (e.g., manufacturing, supply chain, sales, and marketing into a single location, which can help eliminate data silos and make data more accessible.  This also ensures that decision-makers have access to up-to-date information.

Enabling advanced analytics: an IDP applies advanced analytics and machine learning algorithms to interpret data in real time, identifying trends, anomalies, and opportunities more quickly. It also supports a variety of analytics tools, including SQL queries and AI/ML, which can help drive operational insights.

Improving data management: an IDP provides a centralized and scalable infrastructure for collecting, storing, processing, and analyzing vast amounts of data. It can help streamline data management by running solutions at or near the edge, where data can be organized and cleansed before being sent to the cloud.

Enhancing collaboration: An IDP can help improve collaboration between departments (e.g., R&D, marketing, sales) by providing a unified data platform where all relevant data is accessible among users. It can also improve collaboration with suppliers by sharing real-time data, leading to better alignment and faster response times.

Implementing an IDP

Implementing all the above can seem like an enormous undertaking, but it is possible to start small with a ‘lighthouse project’ that can demonstrate the ROI to be gained from a specific use case which can then be spun out company-wide. (An IDP is more likely to deliver value in factories in the global North where labor is more expensive, so it may make sense to look at projects in larger factories in these locations.) This project can be carried out ‘offline’ so there is no risk to your ongoing production capabilities and only rolled out once it is fully mature.  

Your IDP will likely be cloud-based so, consistent with the ‘start small, scale fast approach described above, you will want it to be as flexible and modular as possible with the possibility of reusing data across different cloud platforms. Deploying UNS, which creates a virtual layer that brings disparate data systems together, is one way of avoiding vendor lock-in.  You might also think about working with a partner like Orange Business that has strategic partnerships with all the Hyperscalers but is vendor-neutral.

Conclusion

As McKinsey notes, “the CPG sector faces some unique challenges. The proliferation of data and its complexity—sources are scattered across retailers, suppliers, manufacturers, and consumers—have created massive issues in terms of harnessing the data to find, track, and capture value. At its core, the reason for this low success rate is that companies fail to perform the deep organizational surgery required to affect the broad-based change that’s needed. (!)”

Implementing an IDP is therefore both a challenge and an opportunity. Breaking down these data siloes will undoubtedly require a great deal of work but the value of doing so extends far beyond the immediate IDP solution. Creating a platform for real-time data-driven decision-making has the potential to transform all aspects of their operations and drive better business outcomes.