Nick Calver, Marketing Director
Each grocery retailer presents its trading data in a different way, but you want to operate consistently across them all. You can accomplish so much more with one approach to reporting, analysis, and sharing insights. However, that is easier said than done. Isn’t it?
Grocery trading data represents a valuable picture of shopper demand. However, creating a clear image from the different parts presents a complex problem. Grocery datasets that seem similar in principle are very different in practice: various formats, structures, terminology, measures, definitions, product codes and descriptions, collection methods, and granularity, to name just a few.
They differ so much that it’s hard to establish where the similarities lie once you explore them in detail.
Data lakes are meant to help CPGs process and harness those insights
In response, many organisations have adopted “Big Data” technology to address this complexity. CPGs worldwide have implemented data lakes to address the challenge. Effectively serving as a demand signal database, a data lake harmonises incoming demand indicators to enable users to detect trends across a wide range of aggregated data. That’s the intention, at least; the reality is often different. While many organisations have begun to invest in data lakes, limitations at both technical and operational levels have shown that the value delivered can often fall short of expectations.
The reality of the data lake is usually a little different from the expectation
For CPGs, we tend to see six specific points of failure when it comes to encapsulating demand signals inside of data lakes:
- Limited input: Many CPGs struggle to get a broad enough spectrum of information into their data lake, resulting in overly simplistic outputs and a lack of nuance and depth.
- Design shortfalls: Data lakes aren’t specific to the CPG industry. As a result, generalist data lakes can struggle to cope with data in the kind of quantities that CPGs deal with.
- Standardisation difficulties: With data arriving in multiple formats from a wide variety of sources, transforming it into a standardised format for further analysis can be time-consuming and expensive.
- The pace of change: The sheer speed with which the consumer goods industry moves means that keeping a data lake up to date with market changes can be tough.
- Duplicated effort: With bespoke practices and domain-specific knowledge required for every retailer that a CPG serves, effort is often wasted on replicating the same process across each.
- Fragmented data: A data lake is meant to enable aggregated analysis across the data set. But with little commonality between data from different retailers, this often proves impossible.
The result is that, rather than getting the insights they’re expecting, many CPGs end up with an inefficient, inconsistent, and ultimately ineffective platform. With fragmentation rife and system complexity, meaning that any modifications or additions to the data lake can have a substantial impact elsewhere, usage tends to be restricted to key accounts only.
Solving the common conundrums of data lakes
For CPGs looking to leverage their demand signals fully, considering demand intelligence platforms can be a viable alternative. These platforms serve as a demand signal repository (DSR), providing the essential capabilities needed to manage the challenges of a data lake. They offer the functionality and flexibility required to simplify and streamline the organisation of demand signals. This ensures both accessibility and reliability. Key capabilities include:
- Translating: The ability to navigate between different retail customer terminologies, timings, and technologies. This reduces the need to understand the idiosyncrasies of various retailer systems and enables consistent reporting, analytics, and insight.
- Linking: They connect different supply chain signals — such as stock movements, inventory positions, and till sales — to move away from isolated data points to rich, related flows. This allows activities at any point in time to be understood in the context of their origin and subsequent impact.
- Representing: They handle all forms of product information, including consumer units, traded units, and shipping units. This capability facilitates switching between case and unit perspectives of trading performance.
This approach helps reduce, rather than create, more complexity. It enables:
- The standardisation of diverse demand signals into a consistent and easily manageable format.
- The easy addition of new datasets or modifications to existing data for optimised and efficient analysis.
- Enhancing data accessibility and democratisation through dashboards and sharing, making it easier to find and view the data.
- The application of machine learning and AI techniques universally across the data set, rather than on individual pools.
Demand intelligence platforms can also help simplify product matching complexity
Best-in-class demand intelligence platforms also offer tools to transform the existing product-matching experience. Matching products across different retailers can be challenging since they often classify or name identical products differently. This discrepancy can lead to inconsistencies in data feeds and skewed numbers, as a single product might be perceived as two or three separate SKUs by a retailer. Consistent reporting across multiple retailers is crucial for accurate category performance evaluation. However, achieving this consistency can be difficult unless products are uniformly classified. It is a major pain point.
With data management tools, you can benefit from automated product matching between your product definitions (codes, descriptions, units of measure, etc.) and those in your customers’ trading data. Data owners and administrators simply need to approve or amend match suggestions. Afterwards, they can rely on the platform to provide trading performance information. This information can be presented from your product perspective across customers, or on a customer-by-customer basis using their perspectives.
In summary
If you are considering how to address the challenge of disparate trading data from various supermarkets, then demand intelligence platforms can offer a comprehensive solution to achieve consistency across all retailers. By simplifying the integration and analysis of diverse datasets, they can facilitate coherent reporting and insight sharing, transforming a complex task into an efficient and manageable operation.
Want to see SKUtrak in action? Contact us today for a demonstration of how SKUtrak can fuel consistent, cohesive analysis – without the complexity.