Data Visibility & Analytics

Parker Avery Point of View

Laying the Foundation for Optimizing Information

by Sonia Hernandez

With a wealth of information available across multiple retail channels and applications, it is a great challenge to provide a consistent narrative of data and analysis across your entire enterprise. When there is not a single version of the truth, all analytics produced tend to get scrutinized by leadership and ultimately get discredited or – even worse – bad decisions are made using incorrect or flawed data.

apples & oranges

Inevitably, when business intelligence is not deemed trustworthy, the data mining and analytics functions become decentralized and silo'd, which leads to rogue reporting and limited or no data governance across business groups.

Given these challenges, a common question many retailers face is:

Where do we start in building consistency and trust with data across all areas of the business?

Balancing and optimizing inventory across the distribution network can address these difficulties and more. Part of the discussion must include where inventory is at any given moment. This requires a robust system and full integration to give an accurate snapshot based on the many different systems traditionally used within a retail enterprise.

Consider store inventory, distribution inventory and inventory in-transit (by truck, boat or air). In an ideal situation, inventory in the supply chain is fed into a supply chain visibility tool, and is then available for end users to make proper decisions, as well as to optimize fulfillment.

In this viewpoint, Parker Avery addresses these topics with detail around how integration and visibility of a retailer's inventory supply can meet the demand from all channels.



Defining a Common Ground - Referential Data

The first step in defining a common narrative and providing consistent analytics is to define a common starting point: a single version of the truth. Referential Data encompasses all data that is needed to harmonize multiple systems' key business information including customers, products, materials, suppliers and the like.

The source system (e.g., a Product Information Master or PIM) for any given set of data should be the starting point for all analytics. While the same data can reside in multiple systems, only the source should be used when producing analytics. For example, an Item or SKU may reside in many different systems (Finance, Supply Chain, Marketing, etc.), but in most cases these subscribing systems will have only a subset of the total Item/SKU population and attributes that the Item'/SKUs source system will have. When this type of data is aggregated, data issues and disparity of results only get compounded. When referential data is not in sync across systems, seemingly straightforward reporting of key metrics such as sales can become the most daunting of tasks. When Items are missing or associated incorrectly to a hierarchy, sales numbers will never match between systems. Having an agreed upon starting point is the only way to not only consistently deliver the same results, but also speak a common language.

Ensuring all analytics begin from the same starting point is also critical to creating trust throughout the organization.

An Enterprise Data Warehouse (EDW) should follow the single version of the truth that an organization has agreed upon for their referential data. The Enterprise Data Warehouse should be the central repository for harmonized source system data required for producing all of a retailer's key analytics. Having all of these data elements residing in the same place makes it possible to consistently pull repeatable and reproducible analytics. This ensures that the right data is always provided, accounting for the first component of consistent analytics.



Consistent Delivery and Timing of Enterprise Data

Having the right data is only the first piece of the puzzle. How can the right data be delivered consistently at the right time?

As the sources of all referential data are defined throughout the enterprise, the definition of how each system publishes and consumes data has to be addressed. To achieve one version of the truth, data transmission confirmation and timing are critical. This applies not only to the Enterprise Data Warehouse, but intersystem as well. It is typical to have some systems in a landscape as real time or operational in nature, and others requiring updates once per day, week, month, etc. Managing all of these interfaces as one-to-one relationships can become an increasingly daunting task over time as the systems landscape evolves and changes.

Detangling the enterprise data architecture is a must in developing a sustainable and trustworthy data landscape. Many leading companies have moved to employing Enterprise Service Bus (ESB) systems to manage the flow of data from source systems to subscribing systems. An Enterprise Service Bus essentially acts as a hub for receiving published data from source systems and transforms and queues the source system data for delivery to each subscribing system.

igh


Benefits of employing an Enterprise Service Bus system include:
> Consistent delivery of referential data
> Monitoring services for both publishing and subscribing systems
> Fast data synching and recovery when subscribing systems are down
> A streamlined and easily scalable data architecture

Having such a system in place makes it possible to get ahead of data issues before they are put in front of the business. This can be pivotal in building trust within the enterprise. Such systems also provide the ability to easily identify data transmission failures before they create issues downstream.



Final Word

Having the right data at the right time is a necessity for making informed decisions about any business, especially with the increasing complexity of the modern omnichannel retail world. The ability to have consistent analytics is not only critical to the successful measurement of a company's performance, but also to make informed decisions about where and how to drive future results.

By taking the time to get aligned on what referential data is needed, ensuring that it is available when it is needed, and instilling the inherent trust that comes from such consistency will pave the way for utilizing data and analytics for their ultimate purpose: making sound business decisions.



If you’d like to learn more about our vision or understand how you might take advantage of this strategy, contact us at Contact@parkeravery.com or call 770.882.2205.

Download PDF Version (1.5 MB)

Return to previous page