Start with the Users
As with any system implementation initiative, the user community defines the acceptance criteria for data migration efforts. Valiance recommends that the users be brought on-board the migration effort early in the process to set these acceptance criteria; either directly or through the appropriate representation. Additionally, Valiance understands the need to efficiently leverage the user’s time—We will use available information from the build team and focus our efforts where the users have the most impact, supplementing user defined information with information that may be derived from the source application via Valiance’s proprietary technology (refer to the next section, Derive Information). Valiance recommends that the following information be garnered directly from the user community:
- Identify source data stores
- Categorization—logical grouping of source data for migration
- Data exclusion—list of source system data that will not be migrated
- Enumerate the business processes that are associated with the source data stores
- Ascertain the following for each data store: initial object classifications (records, documents, folders or other relevant object types), status (work-in-process, effective or archive), version requirements, rendition requirements, authoring tool(s), user types supported,
- Scheduling concerns for the migration including permissible downtime (a weekend, or a week night) and critical business issues (a submission deadline).
- List any required external system interfaces
- Review via a demonstration, the source system application
Valiance will organize and document this information in a Consolidation Catalogue. This Catalogue will be augmented in the next few steps with information derived directly from each of the source applications.
Data Discovery: Derive Information from Source and Target Applications
The Consolidation Catalogue is a top down view, a guide to the appropriate data sources and target applications. From here, Valiance’s Data Discovery may be used to get a detailed perspective on the source and destination application and provide sufficient information for the initial mapping between the two. Data Discovery is bottom up detailed view that is derived using software tools (TRUcompare) and a bit of analysis to provide the following:
- Categorization— logical grouping of target data
- Define queries—technical definitions that define the scope of the source data sets
- High-level mappings—If the target types are defined, define the source to target high-level mappings for category of data or content
- Identify gaps—the high-level mappings will likely have gaps for supported source data categories in the destination application, for example, insufficient object type definitions (typically requires sub-type definitions) and modifications to the cabinet folder hierarchy
- Identify data requirements—The target application has data requirements such as mandatory fields, valid value lists and other field-level validation checks. Each of these should be enumerated and used as the basis for testing the sources for data mismatches.
- Define likely sources for data cleansing—detect documents/objects that cannot be categorized and require data cleansing (manual or automated).
- Generate list of records that will be migrated based upon criteria for review by the business. This list may be detailed, or if appropriate, summarized for larger volumes.
Define the System-Level Migration Complexities
Migration efforts are difficult to precisely scope until data discovery is executed (should yield a large percentage of issues) and an initial migration is tested with TRUcompare (gets the rest of the data-level issues identified). If done well, the initial migration will be tested about half way through the migration process (if not done well, it will be done later). Valiance understands that completing 50% of the migration effort is often too late for budgeting purposes.
Valiance recommends that each system-level migration effort be reviewed and an expected level of complexity estimated. The estimate will be based upon the volumes, percentage “not categorized,” number of related business processes, the number of document types, data quality, the level of understanding of the source data and the expected throughput . This level of complexity will be used, along with the other data previously catalogued, to estimate the labor required to migrate and validate each system.
Master Data Management
In many corporate environments, it is not uncommon for key application data (non-transactional data) to vary across departments. For example, the definitive list of product names often vary across large corporations which provide barriers to perform corporate-level reporting. Consolidations offer an opportunity to review and if necessary, define corporate standards for these key data sets.
Valiance recommends identifying likely candidates for these key data sets and defining a process to define and manage these data sets centrally (Master Data Management). To be effective, this process needs to start early in the overall consolidation effort and involve each of the relevant user communities. The alternative to ignoring Master Data Management is to revisit the definition of key data sets multiple times as part of the system-level migration efforts.
Create High-Level Plan and Estimate Effort
Valiance’s approach for scoping is user-centric and leverages our existing toolset to obtain detailed information directly from the source and target applications. This approach goes beyond the typical “survey” results, with significant system-level detail including: initial mapping information, high-level data cleansing requirements, external system interfaces, volumes, throughputs and takes into account each of the key user groups scheduling requirements.
Valiance compiles this information in a Migration Plan for a number of key constituents including: the user communities to review/approve their system’s information, for senior management’s review of the later stages of the migration effort, for QA or validation review, the migration tool selection process and finally, to set the appropriate expectations between the migration team and others involved with the deployment activities.
Valiance will use the Migration Plan to create a detailed task level of effort estimate to complete the migration process. For each migration, we will list each task, each of the relevant user roles and percent allocation for each of the user roles. These system level estimates will be input into the overall deployment plan to define the high-level resource requirements and the resource requirements by role at any given point. It is expected that this information will be used for planning purposes and to play out “what if” scenario to resource the remaining steps required for a successful migration (or consolidation) process.
|
