Dears,
Despite multi-million dollar investments in data warehouses, ERP’s and CRM systems, many companies are still struggling to create a consolidated view of the customer and deliver the right data, to the right person at the right time. As a matter of fact, many sales and marketing executives still identify incomplete and inaccurate customer information” as their biggest obstacle when attempting to optimize customer interactions across all channels.
The goal of Customer Data Integration
“To provide a combination of technologies, processes and services to develop and maintain an accurate, timely and complete view of the customer across multiple channels, business lines and, potentially, companies, in which there are multiple sources of customer data in various applications and databases.”
Will Customer Hub be the final enabling technology which allows companies to reach the Holy Grail in customer management and create a truly consolidated view of the customer?
Before answering this question, it is necessary to answer two prerequisite questions?
• Why have ERP and CRM systems been unable to allow companies to create a consolidated view of the customer?
• What is different about CUSTOMER HUB that will allow it to succeed where data warehousing has failed?
WHAT ABOUT ERP AND CRM SYSTEMS?
In most organizations, ERP and CRM systems have been unable to deliver the complete view of the customer because 100 percent of all customer information is rarely stored within the walls of a single application. It is common practice for companies to use ERP applications from one or more vendors for Back-office functions, a combination of custom-developed applications and packaged CRM applications for sales automation and customer service and custom-built applications for other customer touch points such as the Web. While technically possible to standardize on a single vendor for all Customer-related applications today, it remains highly impractical for most organizations.
As a result, many organizations build a data warehouse to achieve a complete view of the customer. Unfortunately, not many data warehouses are real-time, which is a critical requirement when creating a complete view of the customer to support the enterprise. In addition, data warehouses require a substantial development effort and do not include a business services layer to encapsulate and protect the customer data from direct access by external systems.
The unique feature of the new breed of Customer Hub applications is that they offer a layer of business services, which act as a broker between external systems requesting customer data and the actual customer data. In reality, it is more appropriate to think of CUSTOMER HUB as a data governance platform than as an application. Typical business services within a CUSTOMER HUB application suite include the following:
Administrative services which allow administrators to do the following:
• Define ‘trust levels’ for each source system that request to insert/update individual customer records and/or attributes
• Define security levels for each source system and user when accessing customer records
• Approve the addition of new customer records
• Review and correct customer data that is incorrect
• Data manipulation services which receive requests from source systems to insert/update customer data and apply the previously discussed ‘trust levels’ when doing inserts/updates
• Data access services which receive requests from source systems to read customer data
• “Publish and subscribe” services which publish new/updated customer externally so that customer data in external application which are subscribers can be updated
• Data quality services which perform such tasks as name and address validation and standardization, duplicate identification and house holding
In reality, some companies have been able to custom-build a real-time data warehouse which includes all or some of the above functionality to deliver a complete customer view. However, due to the high cost of this approach and because numerous vendors, such as Oracle’ Siebel, SAP and Siperian to name a few, are delivering pre-built Customer Hub applications with some or all of the above functionality, it is our opinion that the custom route for creation of a customer repository to deliver the complete customer view to the enterprise should be the path of last resort.
We at Sierra Atlantic believe that Customer Hub represents a natural next step beyond data warehousing, and will allow many companies to achieve the goal of a complete customer view more quickly and at a more affordable cost than before.
The most common customer hub architecture is referred to as the “Coexistence Style”
Within the “Coexistence Style,” external applications with customer data communicate with the “customer master” via the business layer, which provides all the services previously discussed. Ideally, whenever an external application needs customer data not stored locally, it queries the “customer master” instead of interfacing with another external system that may have the data. This avoids creation of a spider web of interfaces between source systems, which are difficult and expensive to maintain. As a result, the “customer master” becomes the hub in a hub-and-spoke architecture. Once this approach has been successfully implemented, a new philosophy towards the storage of customer data can be realized. Going forward, whenever a new customer attribute is needed, the new data element can be
stored exclusively in the “customer master.” Over time, this will allow an organization to not only create the complete customer view, but also provide architectural separation between their application layer and data layer. As a result, they ultimately achieve a CUSTOMER HUB architecture style referred to as the “Transaction Hub Style”
Within the “Transaction Hub Style,” all customer information is stored centrally within the “customer master.” None of the individual applications store customer data locally - they simply access the customer information stored centrally. Although it is impractical for most organizations to focus on this CUSTOMER HUB architecture style initially (because it requires substantial enhancements to most source systems), it is still noteworthy to be aware of this style, since this should represent a desired end state for most organizations. In addition, having this style as a desired end state will impact ongoing packaged application selection and development.
Although CUSTOMER HUB initiatives ultimately involve implementation of new software and hardware platforms, technology is definitely not the starting point for a CUSTOMER HUB initiative. Alternatively we at Sierra Atlantic advocate adherence to our ‘Data-to-Intelligence “ assessment methodology which initially focuses on the needs of the business community.
The typical sequence of events within our ‘Data-to-Intelligence’ methodology is as follows:
• The first step is to conduct interviews with relevant end-users to understand key business issues and determine Customer Hub-specific capabilities that a “customer master” must support. Capabilities can be either transactional or analytical. Transactional capabilities are supported by creation of process flows while analytical capabilities are supported by defining metrics. In addition, a parallel activity to develop a high-level understanding of the existing systems is also conducted.
• After finalizing the Customer Hub-specific capabilities that must be supported, the next step is to create a conceptual data model (entity-level) of customer data to be stored within the “customer master” In addition, this deliverable should estimate data volumes, insert/update rules, etc. This deliverable will also be of value later when selecting the CUSTOMER HUB package, since the data modeling flexibility of CUSTOMER HUB tools vary.
• The software/hardware components should be finalized as part of the ‘Solution Architecture’ step, making the build vs. buy decision where appropriate. If the decision to buy the software/hard, the next step would need to include a tool selection RFP
• Create an implementation plan which includes design, development, testing and rollout of the CUSTOMER HUB application.
To avoid project failure, Customer Hub implementation initiatives must begin with an assessment such as our ‘Data-to-Intelligence’ assessment described above. Since the Customer Hub initiatives require substantial commitment of resources, it is also impractical to pursue a ‘big bang’ project approach. Alternatively, we at Sierra Atlantic advocate a phased approach in which Customer Hub functionality (components of the “customer master”) is delivered incrementally in smaller phases. Ideally, each phase should support one or more of the following:
• Improved and/or new business processes which are now possible due to the availability of accurate and timely customer data
• Improved and/or new analytical capabilities which are now possible due to the availability of accurate and timely customer data
• Elimination of IT infrastructure previously focused on integration of customer data across systems (i.e. the spider web)
Good Luck
Your Partner and Companion
DC*
Despite multi-million dollar investments in data warehouses, ERP’s and CRM systems, many companies are still struggling to create a consolidated view of the customer and deliver the right data, to the right person at the right time. As a matter of fact, many sales and marketing executives still identify incomplete and inaccurate customer information” as their biggest obstacle when attempting to optimize customer interactions across all channels.
The goal of Customer Data Integration
“To provide a combination of technologies, processes and services to develop and maintain an accurate, timely and complete view of the customer across multiple channels, business lines and, potentially, companies, in which there are multiple sources of customer data in various applications and databases.”
Will Customer Hub be the final enabling technology which allows companies to reach the Holy Grail in customer management and create a truly consolidated view of the customer?
Before answering this question, it is necessary to answer two prerequisite questions?
• Why have ERP and CRM systems been unable to allow companies to create a consolidated view of the customer?
• What is different about CUSTOMER HUB that will allow it to succeed where data warehousing has failed?
WHAT ABOUT ERP AND CRM SYSTEMS?
In most organizations, ERP and CRM systems have been unable to deliver the complete view of the customer because 100 percent of all customer information is rarely stored within the walls of a single application. It is common practice for companies to use ERP applications from one or more vendors for Back-office functions, a combination of custom-developed applications and packaged CRM applications for sales automation and customer service and custom-built applications for other customer touch points such as the Web. While technically possible to standardize on a single vendor for all Customer-related applications today, it remains highly impractical for most organizations.
As a result, many organizations build a data warehouse to achieve a complete view of the customer. Unfortunately, not many data warehouses are real-time, which is a critical requirement when creating a complete view of the customer to support the enterprise. In addition, data warehouses require a substantial development effort and do not include a business services layer to encapsulate and protect the customer data from direct access by external systems.
The unique feature of the new breed of Customer Hub applications is that they offer a layer of business services, which act as a broker between external systems requesting customer data and the actual customer data. In reality, it is more appropriate to think of CUSTOMER HUB as a data governance platform than as an application. Typical business services within a CUSTOMER HUB application suite include the following:
Administrative services which allow administrators to do the following:
• Define ‘trust levels’ for each source system that request to insert/update individual customer records and/or attributes
• Define security levels for each source system and user when accessing customer records
• Approve the addition of new customer records
• Review and correct customer data that is incorrect
• Data manipulation services which receive requests from source systems to insert/update customer data and apply the previously discussed ‘trust levels’ when doing inserts/updates
• Data access services which receive requests from source systems to read customer data
• “Publish and subscribe” services which publish new/updated customer externally so that customer data in external application which are subscribers can be updated
• Data quality services which perform such tasks as name and address validation and standardization, duplicate identification and house holding
In reality, some companies have been able to custom-build a real-time data warehouse which includes all or some of the above functionality to deliver a complete customer view. However, due to the high cost of this approach and because numerous vendors, such as Oracle’ Siebel, SAP and Siperian to name a few, are delivering pre-built Customer Hub applications with some or all of the above functionality, it is our opinion that the custom route for creation of a customer repository to deliver the complete customer view to the enterprise should be the path of last resort.
We at Sierra Atlantic believe that Customer Hub represents a natural next step beyond data warehousing, and will allow many companies to achieve the goal of a complete customer view more quickly and at a more affordable cost than before.
The most common customer hub architecture is referred to as the “Coexistence Style”
Within the “Coexistence Style,” external applications with customer data communicate with the “customer master” via the business layer, which provides all the services previously discussed. Ideally, whenever an external application needs customer data not stored locally, it queries the “customer master” instead of interfacing with another external system that may have the data. This avoids creation of a spider web of interfaces between source systems, which are difficult and expensive to maintain. As a result, the “customer master” becomes the hub in a hub-and-spoke architecture. Once this approach has been successfully implemented, a new philosophy towards the storage of customer data can be realized. Going forward, whenever a new customer attribute is needed, the new data element can be
stored exclusively in the “customer master.” Over time, this will allow an organization to not only create the complete customer view, but also provide architectural separation between their application layer and data layer. As a result, they ultimately achieve a CUSTOMER HUB architecture style referred to as the “Transaction Hub Style”
Within the “Transaction Hub Style,” all customer information is stored centrally within the “customer master.” None of the individual applications store customer data locally - they simply access the customer information stored centrally. Although it is impractical for most organizations to focus on this CUSTOMER HUB architecture style initially (because it requires substantial enhancements to most source systems), it is still noteworthy to be aware of this style, since this should represent a desired end state for most organizations. In addition, having this style as a desired end state will impact ongoing packaged application selection and development.
Although CUSTOMER HUB initiatives ultimately involve implementation of new software and hardware platforms, technology is definitely not the starting point for a CUSTOMER HUB initiative. Alternatively we at Sierra Atlantic advocate adherence to our ‘Data-to-Intelligence “ assessment methodology which initially focuses on the needs of the business community.
The typical sequence of events within our ‘Data-to-Intelligence’ methodology is as follows:
• The first step is to conduct interviews with relevant end-users to understand key business issues and determine Customer Hub-specific capabilities that a “customer master” must support. Capabilities can be either transactional or analytical. Transactional capabilities are supported by creation of process flows while analytical capabilities are supported by defining metrics. In addition, a parallel activity to develop a high-level understanding of the existing systems is also conducted.
• After finalizing the Customer Hub-specific capabilities that must be supported, the next step is to create a conceptual data model (entity-level) of customer data to be stored within the “customer master” In addition, this deliverable should estimate data volumes, insert/update rules, etc. This deliverable will also be of value later when selecting the CUSTOMER HUB package, since the data modeling flexibility of CUSTOMER HUB tools vary.
• The software/hardware components should be finalized as part of the ‘Solution Architecture’ step, making the build vs. buy decision where appropriate. If the decision to buy the software/hard, the next step would need to include a tool selection RFP
• Create an implementation plan which includes design, development, testing and rollout of the CUSTOMER HUB application.
To avoid project failure, Customer Hub implementation initiatives must begin with an assessment such as our ‘Data-to-Intelligence’ assessment described above. Since the Customer Hub initiatives require substantial commitment of resources, it is also impractical to pursue a ‘big bang’ project approach. Alternatively, we at Sierra Atlantic advocate a phased approach in which Customer Hub functionality (components of the “customer master”) is delivered incrementally in smaller phases. Ideally, each phase should support one or more of the following:
• Improved and/or new business processes which are now possible due to the availability of accurate and timely customer data
• Improved and/or new analytical capabilities which are now possible due to the availability of accurate and timely customer data
• Elimination of IT infrastructure previously focused on integration of customer data across systems (i.e. the spider web)
Good Luck
Your Partner and Companion
DC*
No comments:
Post a Comment