Current techniques for data analytics have changed the way we view data – spawning a new industry/role (a business analyst who understands machine learning and big data). However, with the very nature of the term ‘analytics’ implies a post event investigation. In the new disrupted value chain, this is no longer good enough. Putting interventions long after an event has taken place only mitigates future issues, while potentially impacting current business.

Incorporating data streaming aids with flagging issues and enabling almost real-time interventions. Removing lag time from event and action improves responsiveness and creates an agile value chain that can truly cater for JIT.

But what is data streaming and how does it impact the supply chain? This article will delve deeper into the topic and highlight specific use cases.

What is Data Streaming

Large amounts of unstructured and semi data is continuously generated by different sources at high speed within the entire value chain. This type of big data is forcing many organizations to focus on how they process, interact and store the data. Data streaming can perform real-time analysis on streaming data, and it differs from data lakes in speed and continuous nature of analysis, without having to store the data first.

Why Data Streaming

Big data analytics, trend analysis, bench marking is generally based on rolling averages, numbers rolled up and reported weekly and or monthly in arrears, all very good but nothing that makes an operational difference in enabling competence on the ground. When an economy is distressed the average is not good enough to influence the bottom line, getting granular in the reporting and ensuring that the granular change or deviation is noted in time enough to be fixed or adjusted to make sure a sale can take place.

Data Streaming in Action

Understanding the principle of data streaming is better done through an illustrative use case currently in running in South Africa.

Let’s take a personal care and cleaning manufacturer:

They have 5 products (with different pack configurations) that competes in the sanitizer category in one of the major retail chains. This chain has committed to providing an ePOS stream. The daily till sales and basket data feed for the sanitizer is processed.

Combining the tills of the product by SKU, as well as stock in store and stock in the Distribution Centre, MACmobile enables the manufacturer to identify Key Performance Areas against which performance benchmarks are established and any and all deviations are reported against. The deviation reporting is live and communication is made to merchandisers and or retail management on the ground – speeding up the response time.

The impact on the Supply Chain

This real-time access to streaming sales data removes the potential losses due to reacting to issues after the fact. We move into a proactive state where remediation can be done in the moment. Having merchandisers empowered with image recognition technology provides real-time proof of repair, as well execution and planogram compliance in each of the retail channels, pricing compliance via the reading of Perpetual Inventory (PI) labels and comparing to a pricing data base per retail outlet ensures maximum return on investment.

Over a period, we are able to map retail sales per sku, per store, per region, across the country allowing for intelligent stock allocation, reducing returns and wastes and allowing for Just-In-Time manufacturing to satisfy customer demand as it varies.

Ensuring that products are visible on shelf at all times is the crux of successful manufacturing and retail. With Data Streaming we are able to have real-time access to a wealth of data to allow for smart resource allocation based on proactive and actionable remediation, instead of reactive event tracking. We can empower manufacturers to leverage a single version of the truth through the entire supply chain, so that they can move to optimal, JIT manufacturing across the country.