Downstream Demand Data
One of the most compelling issues facing many consumer goods (CG) companies today is how to leverage the mountain of downstream demand data to which they have access. Demand data comes in many forms; while point-of-sale (POS) data is the most prevalent, demand insights can also be gleaned from warehouse withdrawal data, transactional data from collaborative relationships such as VMI or CPFR, ultimately from RFID data, loyalty card data, etc. CG companies have historically received POS data from syndicated sources, and more recently, directly from the retailers themselves. Oftentimes, however, the volume of data can be overwhelming and is therefore not leveraged to it's potential.
Direct Correlations
Research shows a direct connection between leveraging downstream demand data and performance. In its DDSN reports, AMR Research proves a direct correlation between leveraging demand and performance: Specifically, companies that are 30 percent better at demand forecasting average 15 percent lower inventories, 17 percent stronger perfect order fulfillment and 35 percent shorter cash-to-cash cycle times. In the perfect order metric, the difference shows up strongly in the stock-outs component. Companies with stronger demand forecasting capability average 0.5 percent stock-outs, while the companies with weaker capability average 5 percent stock-outs -- a full 10-times higher. Clearly, the ROI is there, so why aren't more companies taking advantage of the available data?
The answer lies partly in organizational alignment -- the processes that could truly leverage the insights aren't necessarily working with the same data, systems and procedures. Today, demand data is primarily used in three distinct areas, which are usually independent silos that don't share findings. Those areas are:
Sales - for account management
Demand Planning - to generate a forecast
Supply Chain - to support specific VMI or CPFR relationships
This leaves out other processes and departments that could benefit from the demand insights, particularly Research and Development (R&D) and operations. Sales and Operations Planning (S&OP) is another important function that rarely sees this data, yet could be the process to bring the most disciplines together to enable profitable demand response.
The Problem Lies in the Data
The most frequently cited reason for not leveraging demand data is the data itself. It is often dirty, unreliable, non-standard and each retailer expects something different in return for sharing the data. In addition, each downstream application that could leverage the data needs access to downstream data, but in different contexts, formats, at different frequencies, and often require complex customization or re-implementation to use the downstream data. In the past, companies that have exploited this data, particularly POS data, have had to build solutions from scratch that addresses the data and technology requirements. Perhaps the best known example is Anheuser-Busch's Budnet application, which is used to leverage direct store delivery (DSD) sales information, as well as track promotional and competitive insights.
As the need for technology to help manage the data escalates, solutions have emerged that provide Demand Signal Repository (DSR) capabilities: these applications gather, cleanse, normalize, aggregate and analyze all types of downstream demand data. Figure 1 illustrates the data flow and the potential uses for the data:
Account management
Category management
Promotion management
Vendor-Managed Inventory (VMI)
Corporate forecasting
Demand insights
Often each group or department will have a different business sponsor, answer a different business question and require a different data definition. This adds to the data complexity, yet emphasizes the need for each business process to use the same source of data, the elusive one version of the truth. This can only happen if there is a central repository to manage data cleansing and normalization, and then feed aggregated slices of the data to the appropriate application or department who can then act on the data.
Many CG companies are starting to invest in this area and understand that leveraging demand data is a pre-requisite to becoming demand-driven. However, there is often what AMR calls "organizational tension" on how to start. Many IT departments prefer to implement a comprehensive solution to maximize the use of downstream data in all business applications. However, AMR also reports that line-of-business (LOB) owners are often more narrowly focused, seeking a quick return from optimizing their local processes and systems. In these cases, LOB owners, usually sales, and IT should strategize on how to best meet the needs of the enterprise without compromising the immediate needs of any particular department.
Rules of the Road
AMR published several rules of the road, involving technology that can guide companies undertaking this initiative. They include:
Don't wait around for a data standard. There won't be one.
It takes time and money. Typically two years and $1 million to $5 million.
Companies that have invested in data synchronization and MDM efforts have a leg up in data harmonization.
No application technology provider offers an out-of-the-box product.
A DSR is a must have. However, do not attempt to build a DSR yourself. While many companies have attempted to build a DSR themselves, the use of pre-built data models and services from companies, like Teradata, VeriSign and Vision Chain, have proven most successful.
Direct Correlations
Research shows a direct connection between leveraging downstream demand data and performance. In its DDSN reports, AMR Research proves a direct correlation between leveraging demand and performance: Specifically, companies that are 30 percent better at demand forecasting average 15 percent lower inventories, 17 percent stronger perfect order fulfillment and 35 percent shorter cash-to-cash cycle times. In the perfect order metric, the difference shows up strongly in the stock-outs component. Companies with stronger demand forecasting capability average 0.5 percent stock-outs, while the companies with weaker capability average 5 percent stock-outs -- a full 10-times higher. Clearly, the ROI is there, so why aren't more companies taking advantage of the available data?
The answer lies partly in organizational alignment -- the processes that could truly leverage the insights aren't necessarily working with the same data, systems and procedures. Today, demand data is primarily used in three distinct areas, which are usually independent silos that don't share findings. Those areas are:
Sales - for account management
Demand Planning - to generate a forecast
Supply Chain - to support specific VMI or CPFR relationships
This leaves out other processes and departments that could benefit from the demand insights, particularly Research and Development (R&D) and operations. Sales and Operations Planning (S&OP) is another important function that rarely sees this data, yet could be the process to bring the most disciplines together to enable profitable demand response.
The Problem Lies in the Data
The most frequently cited reason for not leveraging demand data is the data itself. It is often dirty, unreliable, non-standard and each retailer expects something different in return for sharing the data. In addition, each downstream application that could leverage the data needs access to downstream data, but in different contexts, formats, at different frequencies, and often require complex customization or re-implementation to use the downstream data. In the past, companies that have exploited this data, particularly POS data, have had to build solutions from scratch that addresses the data and technology requirements. Perhaps the best known example is Anheuser-Busch's Budnet application, which is used to leverage direct store delivery (DSD) sales information, as well as track promotional and competitive insights.
As the need for technology to help manage the data escalates, solutions have emerged that provide Demand Signal Repository (DSR) capabilities: these applications gather, cleanse, normalize, aggregate and analyze all types of downstream demand data. Figure 1 illustrates the data flow and the potential uses for the data:
Account management
Category management
Promotion management
Vendor-Managed Inventory (VMI)
Corporate forecasting
Demand insights
Often each group or department will have a different business sponsor, answer a different business question and require a different data definition. This adds to the data complexity, yet emphasizes the need for each business process to use the same source of data, the elusive one version of the truth. This can only happen if there is a central repository to manage data cleansing and normalization, and then feed aggregated slices of the data to the appropriate application or department who can then act on the data.
Many CG companies are starting to invest in this area and understand that leveraging demand data is a pre-requisite to becoming demand-driven. However, there is often what AMR calls "organizational tension" on how to start. Many IT departments prefer to implement a comprehensive solution to maximize the use of downstream data in all business applications. However, AMR also reports that line-of-business (LOB) owners are often more narrowly focused, seeking a quick return from optimizing their local processes and systems. In these cases, LOB owners, usually sales, and IT should strategize on how to best meet the needs of the enterprise without compromising the immediate needs of any particular department.
Rules of the Road
AMR published several rules of the road, involving technology that can guide companies undertaking this initiative. They include:
Don't wait around for a data standard. There won't be one.
It takes time and money. Typically two years and $1 million to $5 million.
Companies that have invested in data synchronization and MDM efforts have a leg up in data harmonization.
No application technology provider offers an out-of-the-box product.
A DSR is a must have. However, do not attempt to build a DSR yourself. While many companies have attempted to build a DSR themselves, the use of pre-built data models and services from companies, like Teradata, VeriSign and Vision Chain, have proven most successful.