|Discipline Quick Links|
|Disciplines Master Index|
|Glossary Quick Links|
|Glossary Master Index|
This document represents an aggregated, ordered and contextualized view of the material we've been able to compile and publish that is related to the topic of "Data Management." The goal is to make this page a landing and launch point for all things related to this topic. As our content becomes more complete and more accurate, this page should become a very useful and powerful knowledge base for this topic and all parties interested in it.
You'll find that the content for this document is consistent with that of other discipline related documents. This is intentional. The consistency is based on a knowledge pattern that helps individuals learn more about different topics, quicker and more efficiently. We hope you find the material useful and easy to learn.
It's important to realize that content in this document and any related sub-documents are constantly evolving. Therefore, we recommend you check for updates, regularly, to keep up with the latest material.
The Foundation always welcomes your feedback and suggestions for improvement, as we're always looking for ways to improve our solutions and offerings to the general community.
All solutions published by the Foundation are subject to the terms and conditions of the Foundation's Master Agreement.
This document or artifact, along with everything in it, is intended to act as a "Framework" that addresses various aspects of Data Management.
The readers will notice that most sections in the Table of Contents (TOC) use a format where the TOC entry is prefixed with a topic name, followed by a short descriptive title (i.e. "TOPIC_NAME: TOPIC_RELATED_SECTION_TITLE"). This is intentional and represents a format by which the Foundation may achieve things like the identification of appropriate topic areas, the segregation of distinct topic areas from each other, the appropriate ordering of topic areas, and achieve the maintenance of consistency, both, within and across different IT Disciplines.
To elaborate, this artifact is intended to:
From the Foundation's perspective, if done correctly, all of the above will allow the Foundation to properly decompose, document and publish content related to each sub-area or sub-topic for each IT Discipline, including this specific discipline (i.e. "Data Management").
From the reader's perspective, if done correctly, all of the above will allow him or her to easily find and learn about specific areas of interest associated with this and all other IT Disciplines in a manner where the reader may effectively consume and digest material in small atomic segments that act as repeatable and more effective learning units.
As this artifact evolves and progresses, the reader will see it address key areas of the professional IT Discipline "Data Management" that range from its detailed definition through closely related terms, phrases and their definitions, to its detailed specification of Data Management Capabilities, and all the way through to defining, delivering, operating and supporting Data Management Services.
As mentioned previously, this document will continue to evolve and the Foundation recommends the reader check back, regularly, to stay abreast of modifications and new developments. It is also important to understand that the structure of this artifact may change to meet the needs of such evolution.
Before moving on to learn more about the rest of the Data Management framework, we suggest that you take some time to familiarlize yourself with the following very basic term(s)...
"1. A documented, stored and loosely or highly structured set of one or more values or symbols that are arranged in a manner that often has specific meaning or context and that sometimes conforms to some form of specific constraint or set of constraints.
NOTE: Data can be stored in multiple constructs, such as a Record in a File or Database, or as text in raw Content."
"1. The professional discipline that involves working with, in or on any aspect of planning, delivering, operating or supporting for one or more Data Items or any and all solutions put in place to deal with such Items.
2. The solution set that a person or organization puts in place to manage one or more Data Items.
3. The process or processes put in place by a person or organization to assist in the management, coordination, control, delivery, or support of one or more Data Items.
4. The Enterprise Capability that represents the general ability or functional capacity for a Resource or Organization to deal with or handle one or more Data Items. Such a term is often used by Information Technology (IT) Architects when performing or engaging in the activities associated with general Capability Modeling."
In addition to the above basic term(s), you can also learn a great deal about Data Management by familiarizing yourself with the broader spectrum of terms that make up the Data Management Glossary...
Language between IT professionals and the businesses we serve is often a significant barrier to success, as we often spend countless hours trying to interpret each other's meanings. This is often also true between IT professionals who are taught to use certain terms and definitions as part of the organizations and industries they serve. It's when you start to jump from organization to organization, from enterprise to enterprise, and from industry to industry that you realize how much time and effort is wasted on just getting language and meanings correct. For these reasons, the Foundation puts a great deal of focus on terms and phrases, as well as their corresponding definitions. We highly recommend you spend time learning and understanding all of the related terms and phrases, along with their meanings, for all areas of "Data Management."
|Data Management Glossary|
|Centralized Data Management||Data Management Project|
|Data||Data Management Reference Architecture|
|Data Automation||Data Management Release|
|Data Capacity Management||Data Management Report|
|Data Catalog||Data Management Reporting|
|Data Catalogue||Data Management Roadmap|
|Data Configuration||Data Management Role|
|Data Configuration Item||Data Management Rule|
|Data Configuration Management||Data Management Schedule|
|Data Cost||Data Management Security|
|Data Database||Data Management Service|
|Data Decommission||Data Management Service Assurance|
|Data Delivery||Data Management Service Contract|
|Data Dependency||Data Management Service Level Agreement (SLA)|
|Data Deployment||Data Management Service Level Objective (SLO)|
|Data Document||Data Management Service Level Requirement (SLR)|
|Data Document Management||Data Management Service Level Target (SLT)|
|Data File Plan||Data Management Service Provider|
|Data Framework||Data Management Service Request|
|Data Governance||Data Management Software|
|Data History||Data Management Solution|
|Data Identifier||Data Management Stakeholder|
|Data Inventory||Data Management Standard|
|Data Item||Data Management Strategy|
|Data Lifecycle||Data Management Supply|
|Data Lifecycle Management||Data Management Support|
|Data Management||Data Management System|
|Data Management Application||Data Management Theory|
|Data Management Best Practice||Data Management Training|
|Data Management Blog||Data Management Vision|
|Data Management Capability||Data Management Wiki|
|Data Management Center of Excellence||Data Management Workflow|
|Data Management Certification||Data Metadata|
|Data Management Class||Data Migration|
|Data Management Community of Practice (CoP)||Data Plan|
|Data Management Course||Data Portfolio|
|Data Management Data||Data Portfolio Management|
|Data Management Data Dictionary||Data Processing|
|Data Management Database||Data Record|
|Data Management Demand||Data Records Management|
|Data Management Dependency||Data Repository|
|Data Management Discussion Forum||Data Reuse|
|Data Management Document||Data Review|
|Data Management Documentation||Data Schedule|
|Data Management File Plan||Data Schematic (Schema)|
|Data Management Form||Data Security|
|Data Management Framework||Data Software|
|Data Management Governance||Data Strategy|
|Data Management Knowledge||Data Support|
|Data Management Lessons Learned||Data Taxonomy|
|Data Management Metric||Data Termination|
|Data Management Operating Model||Data Tracking|
|Data Management Organization||Data Tracking Software|
|Data Management Plan||Data Transaction|
|Data Management Platform||Data Unique Identifier|
|Data Management Policy||Data Verification|
|Data Management Portfolio||Data Version|
|Data Management Principle||Data Workflow|
|Data Management Procedure||Decentralized Data Management|
|Data Management Process||Enterprise Data Management|
|Data Management Professional||Federated Data Management|
|Data Management Program||Regional Data Management|
|Data Management Project|
Please refer to the IT Glossary for other terms and phrases that may be relevant to this professional discipline.
Readers may also refer to the Taxonomy of Glossaries for terms and phrases that are semantically grouped according to IT Disciplines or enterprise domains.
This Data Management Glossary is a contextual subset of the master IF4IT Glossary of Terms and Phrases. The master glossary can be used by you and your enterprise as a foundation for broader understanding of Information Technology and can be used as a teaching and learning tool for those you work with, helping to ensure a common and more standard language.
A Capability, as it pertains to Information Technology (IT) or to an enterprise that an IT Organization serves, is defined to be "A manageable feature, faculty, function, process, service or discipline that represents an ability to perform something which yields an expected set of results and is capable of further advancement or development. In other words, a Capability is nothing more than "the ability to do something" or, quite simply, a Feature or Function. Therefore, when applied to an enterprise, a Capability represents a critical Enterprise Feature or Enterprise Function.
When it comes to Capabilities, there are multiple types that an enterprise needs to be aware of. Examples include but are not limited to:
As can be seen above, there are Capabilities that are associated with Resources, Organizations, and Assets such as Systems. All are important to an enterprise.
In the case of this IT Discipline (i.e. Data Management), we use the word Capability in the context of an Enterprise Capability or an IT Capability, which are both equivalent to Enterprise Disciplines or IT Disciplines, respectively. In short, the Capability of Data Management represents the ability to deal with any and all Data Items and anything relevant that is related to or associated with any Data Items.
If you think about it, a capability is really nothing more than a "verb" or "action that represents "the ability to do something." Understanding this allows us to derive a consistent and highly repeatable set of sub-capabilities for any Noun we're dealing with. For example:
In summary, the implication is that the Enterprise Capability or Enterprise Discipline known as Data Management is the superset of all the above Sub-Capabilities, as they pertain to or are applied to the discipline-specific Noun: "Data." This now translates more specifically to:
For a more complete list of very specific Capabilities/Disciplines, refer to the Foundation's Master Inventory of IT Disciplines. It is important to note that this inventory is in a flat or non-hierarchical form, specifically because "hierarchy" is almost always a matter of personal preference or context (what hierarchy is important to one Resource or Organization may be unimportant to another's needs or requirements). Therefore, the Foundation has published its inventory of Capabilities in a non-hierarchical, flat form.
This now brings us to a very obvious problem that surrounds Capabilities, which is the fact that there are simply too many "granular" or "specific" Capabilities to document and publish in any single Capability Model. The end result is that a Capability Model may become unwieldy because of trying to incorporate so many different specific Capabilities. Also, Capability Modeling "Purists," who all have their own (and very differing) opinions about how Capability Models should or should not be represented, almost always refuse to get into the details. To address this, we recommend using a generic set of Capabilities that map to and are driven by the Systems Development Life Cycle. For example:
As you can see from the above, we now have a very limited, controlled and manageable set of Discipline-specific Capabilities for the Discipline Data Management.
As a reminder, the above Capability representations are "suggestions" for baselining or initializing your own Enterprise Capability Model (ECM). It's recommended that you take the time to work with your enterprise stakeholders to improve upon and/or customize your own ECM so that you can help meet their needs. However, with that being said, it's always a better idea to go in with a baseline that you can modify rather than building your own solution from scratch, especially if your goals are to standardize, not reinvent the wheel, and not deviate too far from what other enterprises are doing to model their own environments. This is especially true if you've never had any experience building ECMs that have gained and maintained full adoption.
Why do enterprises perform Capability Modeling? Enterprises most often build Capability Models that are associated with Data Management for the following reasons...
Capability Modeling Recommendations: Some things to consider and keep in mind when working on or creating your Data Management and Enterprise Capability Models...
Learn More About Capability Models: Taking the time to learn about and understand Capability Models, what they're for, and how they're used may help you learn how Data Management better fits into the broader enterprise. Therefore, we suggest you spend some time reviewing and understanding the IF4IT Enterprise Capability Model...
Here's a very simple fact... If an enterprise does not establish and enforce clearly defined Ownership (i.e. a Resources and his or her Organization are assigned as accountable ownership) for Data Management, the enterprise has automatically set itself up for failure in its implementation of that discipline. Therefore, if you and your enterprise want to implement and maintain a successful solution for Data Management, there must be a clearly defined Owner that can and will be held accountable for getting work done, providing transparency, helping with strategy setting, and coordinating implementation of Data Management as a fully functional and mature enterprise Service.
Having clearly defined Ownership should not be confused with having fully dedicated Resources that spend one hundred percent of their time working on Data Management. In fact, smaller enterprises can rarely afford to dedicate full time Resources, like larger enterprises can, to all enterprise IT Disciplines. This being the case, all IT Disciplines, including Data Management, should "always" have clearly defined Owners so that there is always a clear point of accountability and contact for any issues or work that need to be addressed.
In addition to the common best practice of having clearly assigned Ownership for Data Management, it is also considered a best practice to clearly publish and socialize Data Management Ownership details to a centralized location (often referred to as a "Service Catalog" or an "Enterprise Service Catalog"), along with Ownership details for all other IT Disciplines, so that the entire enterprise has constant access to it.
Figure: How Ownership of the Capability Data Management fits into the Canonical Model for IT
The above figure helps us understand how Capability or Discipline Ownership fits into the Canonical Model for Information Technology (IT) (i.e. "Think," "Deliver," and "Operate"). Owners are assigned to individual Disciplines or Capabilities, such as Data Management, and are instantly made accountable to the enterprise for the results of all Data Management Thinking activities (i.e. Strategy, Research, Planning and Design), all Data Management Delivery activities (i.e. Construction, Deployment and Quality Assurance), and all Data Management Operations activities (i.e. Use, Maintenance and Support). Done correctly, Data Management Ownership is constant and ongoing. It's important to understand that such assigned Ownership should "never" end so that there is clear and constant accountability and transparency for all aspects of the Canonical Model to the enterprise.
Not having clear Ownership for Data Management means that there is no clear understanding of who is accountable for it, who can provide understanding of what's going on within it, who can help the enterprise provide short term and long term descriptions of work being performed within the Discipline area to improve it over time for its customers, and who can help with getting work done that's associated with it. It means your or your enterprise's implementation for Data Management will be highly incomplete and erratic because no one is constantly (or even partially) watching over the Discipline and its needs for maintenance and evolution. Not having clear Data Management Ownership is a recipe for confusion and, sometimes, even chaos.
In summary, if you and your enterprise truly want to be successful with your implementation of Data Management, ensure that a clear and highly accountable owner is identified and assigned to the Discipline. Publish those ownership details, preferably in an enterprise's Service Catalog, and socialize it so everyone knows whom to go to for answers and for help with Data Management related work. In other words, if you want to implement Data Management as an enterprise Service, then you absolutely must start with clearly defined, published and socialized Ownership.
Throughout the Foundation's documentation, you will continuously run into the references of "Nouns and Verbs." These concepts are key to consistency and standardization, throughout the IT Industry, down to each and every IT Discipline. Given that we've discussed the impact of "Nouns" on the discipline of "Data Management," this section will start to discuss the importance of "Verbs" or "Actions" that can be performed with or against the key Noun or Nouns associated with this Discipline. To reiterate, Verbs or Actions allow us to clearly understand what can be performed on or with the Noun in question. As will be discussed in the next section, Verbs or Actions will also help us clearly identify whom it is (i.e. the "who" or more specifically the Roles) that performs or executes such Verbs or Actions against a Discipline and its associated Noun or Nouns. As will be discussed later, Verbs or Actions will also help identify key Attributes (i.e. Field Names) that are necessary for the very data definition of the Noun or Nouns for this Discipline and will even help identify which Verbs or Actions can be automated for this Discipline.
As a reminder, the base Noun for the discipline known as Data Management is: "Data," which is sometimes referred to as a the Noun: "Data Item."
By now, it should be becoming apparent that verbs represent a baseline for defining solid functional requirements and sub-capabilities for what would be a part of any good Data Management System or Service. What this means is that if you and/or your Organization is looking for a solution in this space (e.g. the purchasing or building of a software solution or the implementation of a Service to address the needs of Data Management), you could use discipline-related verbs to drive the foundation of what the solution should or shouldn't do, as mapped to specific stakeholders that will use or provide the solution.
Examples of the types of Verbs or Actions that are important to this Discipline include but are not limited to:
The above list represents a very small subset of all Verbs or Actions that are relevant for this Discipline. The more complete set can be found in the Roles section of this document, where readers can see the direct correlation of Verb to Noun and to, both, Generic Role and Discipline Specific Role.
An "action" or a "verb" is something that can be performed on or with a specific "noun." The reason it is important to itemize all relevant verbs is because we can now start to determine what we can or cannot do with the noun in question, where in this case the noun is "Data."
|Actions/Verbs||Example as Applied to "Data"||Generic Roles||Discipline-Specific Roles|
|Administrate||Administrate Data||Administrator||Data Administrator|
|Approve||Approve Data||Approver||Data Approver|
|Architect||Architect Data||Architector||Data Architector|
|Archive||Archive Data||Archiver||Data Archiver|
|Audit||Audit Data||Auditor||Data Auditor|
|Bundle||Bundle Data||Bundler||Data Bundler|
|Clone||Clone Data||Cloner||Data Cloner|
|Code||Code Data||Coder||Data Coder|
|Configure||Configure Data||Configurer||Data Configurer|
|Copy||Copy Data||Copier||Data Copier|
|Create||Create Data||Creator||Data Creator|
|Decommission||Decommission Data||Decommissioner||Data Decommissioner|
|Delete||Delete Data||Deletor||Data Deletor|
|Deploy||Deploy Data||Deployer||Data Deployer|
|Deprecate||Deprecate Data||Deprecator||Data Deprecator|
|Design||Design Data||Designer||Data Designer|
|Destroy||Destroy Data||Destroyer||Data Destroyer|
|Develop||Develop Data||Developer||Data Developer|
|Distribute||Distribute Data||Distributor||Data Distributor|
|Download||Download Data||Downloader||Data Downloader|
|Edit||Edit Data||Editor||Data Editor|
|Educate||Educate Data||Educator||Data Educator|
|Export||Export Data||Exporter||Data Exporter|
|Govern||Govern Data||Governor||Data Governor|
|Import||Import Data||Importer||Data Importer|
|Initialize||Initialize Data||Initializer||Data Initializer|
|Install||Install Data||Installer||Data Installer|
|Instantiate||Instantiate Data||Instantiator||Data Instantiator|
|Integrate||Integrate Data||Integrator||Data Integrator|
|Manage||Manage Data||Manager||Data Manager|
|Merge||Merge Data||Merger||Data Merger|
|Modify||Modify Data||Modifier||Data Modifier|
|Move||Move Data||Mover||Data Mover|
|Own||Own Data||Owner||Data Owner|
|Package||Package Data||Packager||Data Packager|
|Persist||Persist Data||Persister||Data Persister|
|Plan||Plan Data||Planner||Data Planner|
|Purge||Purge Data||Purger||Data Purger|
|Receive||Receive Data||Receiver||Data Receiver|
|Record||Record Data||Recorder||Data Recorder|
|Recover||Recover Data||Recoverer||Data Recoverer|
|Register||Register Data||Registrar||Data Registrar|
|Relocate||Relocate Data||Relocator||Data Relocator|
|Reject||Reject Data||Rejecter||Data Rejecter|
|Remove||Remove Data||Remover||Data Remover|
|Replicate||Replicate Data||Replicator||Data Replicator|
|Report||Report Data||Reporter||Data Reporter|
|Request||Request Data||Requestor||Data Requestor|
|Restore||Restore Data||Restorer||Data Restorer|
|Review||Review Data||Reviewer||Data Reviewer|
|Save||Save Data||Saver||Data Saver|
|Search||Search Data||Searcher||Data Searcher|
|Split||Split Data||Splitter||Data Splitter|
|Sponsor||Sponsor Data||Sponsor||Data Sponsor|
|Store||Store Data||Storer||Data Storer|
|Strategize||Strategize Data (or Set Data Strategy)||Strategizer (or Strategy Setter)||Data Strategizer (or Data Strategy Setter)|
|Support||Support Data||Supporter||Data Supporter|
|Test||Test Data||Tester||Data Tester|
|Train||Train Data||Trainer||Data Trainer|
|Upgrade||Upgrade Data||Upgrader||Data Upgrader|
|Upload||Upload Data||Uploader||Data Uploader|
|Verify||Verify Data||Verifier||Data Verifier|
|Version||Version Data||Versioner||Data Versioner|
|View||View Data||Viewer||Data Viewer|
At a minimum, the above list of Verbs can be used to help identify, track, and manage the basic "Features" required by and associated with Data Management, even if your enterprise doesn't maintain a Capability Model that lists specific Data Management Capabilities. Application designers, developers, and architects often find such Verb Lists or Feature Inventories to be invaluable.
A Taxonomy, in its noun form, is defined as:
...a documented and orderly set of types, classifications, categorizations and/or principles that are often achieved through mechanisms including but not limited to naming, defining and/or the grouping of attributes, and which ultimately help to describe, differentiate, identify, arrange and provide contextual relationships between the entities for which the Taxonomy exists.
From this general definition, we can derive that the definition for a Data Management Taxonomy is:
...a documented and orderly set of types, classifications, categorizations and/or principles that are often achieved through mechanisms including but not limited to naming, defining and/or the grouping of attributes, and which ultimately help to describe, differentiate, identify, arrange and provide contextual relationships between Data Items, Entities or Types.
In short, what this means all means is that a Taxonomy is nothing more than a classification or typing mechanism and that a Data Taxonomy is nothing more than a classification or typing mechanism that helps people and systems distinguish between different Data Items, Entities, Types, Records or any other Data Management element you can think of.
It's important to understand that Taxonomies can be as simple as a list of relevant terms or phrases with respective meanings or definitions or they can take on more complex forms, such as hierarchical and graphical model structures that can be homogeneous and heterogeneous in nature. More complex Taxonomies include examples such as "Visual Taxonomies" and "Audible Taxonomies" but, expect in the case of very special technologies, are typically out of scope for general Information Technology (IT) Operations.
The Foundation directs readers to its ever-evolving Inventory of Taxonomies for Standard Taxonomy suggestions. Specifically, readers may want to start with the Taxonomy of Taxonomies, which helps make it clear that the IT Industry is composed of many hundreds if not thousands of Taxonomies, Classifications, Categorizations or Types.
While Taxonomies represent organized classifications or types, you can think of Ontologies as the design and representation of entire lanaguages, with the specific intent to control things like structure, behavior, representation, and meaning. Without getting into a theoretical conversations about Ontologies, you can view this entire article as a foundation for the ontology of Data Management. Or, in other words, a Data Management Ontology.
Throughout this artifact/framework, you will find things like Data Management related terms, phrases, definitions, roles, responsibilities, nouns, verbs, classifications, and so on, all as a means of definining a standard representation for and interpretation of the language of Data Management.
It is only through the definition, communication, and establishment of such Ontologies that we can standardize language and communication associated with Data Management, whether it be between humans and/or systems.
When we talk about Life Cycle (or lifecycle) for Data Management, it's important to keep in mind that there are two different types of Life Cycles that apply. The first is a Data Life Cycle, which addresses Data Management data or entities, and the second is associated with delivering Data Management Assets like Systems or Software solutions.
Data Management Data Life Cycle Phases:
Data Lifecycle (or Life Cycle) for any and all data is the period from the "inception" of data through to its ultimately being "purged" from existence. This is no different for Data Management related data.
Like the data associated with any other professional IT Discipline, Data Management related data adheres to the following common Data Lifecycle Phases:
Figure: Data Management Lifecycle Phases
The above Life Cycle Phases represent the high level transitions that occur from the inception of Data Items or Entities all the way through to their complete elimination from existence. A more detailed breakdown of these transitions or phases represents what are referred to as "Data Management States."
Data Management Systems Development Life Cycle (SDLC) Phases or Data Management Software Development Life Cycle (SDLC) Phases:
The SDLC is a means for facilitating and controlling how IT Professionals deliver Assets, such as Data Management Systems and Software. In this case, you should default to the master SDLC, which is used to deliver any Asset of any type, including those associated with the Data Management discipline.
There are probably no greater or more important tools for providing Data Management transparency and direction than the collection, ordering, categorizing, grouping, and maintenance of all related Data Items. In other words, Data Management Inventories.
In short, an Inventory represents a list of individual things or instances of things that are typically all of the same Noun Type or Data Type, where these instances are described and detailed by their Attributes, along with the Data and Information that act as values for such Attributes.
At a minimum, Data Management Inventories are used for the establishment of solid Data Configuration Management practices, as the Data Instances tracked within such Data Inventories act as Configuration Items (in Target and/or Dependency form) for key Configurations (Data Management Configurations or otherwise).
Inventories are also used for solid decision making. Good decisions, either strategic or tactical, are made based on having good Data and Information. And, good Data and Information only come from taking the time to follow best practices associated with Inventory Management. It's only through building such Inventories that an enterprise can achieve solid Data Management Business Intelligence and Reporting.
Also, it's these very same Inventories that act as the foundation for understanding and managing Total Cost of Ownership (a.k.a. "TCO") for Data Management. Without such Inventories, trying to understand your costs can be nothing more than uneducated guessing.
The obvious place to start is with Data Inventories and then move on to surrounding Inventories that are directly and indirectly related to Data Management.
Additionally, there are many other types of Inventories that are common and important to Data Management, which include but are not limited to examples such as:
If you and/or your enterprise are not collecting and maintaining such Inventories, you're probably considered to be very low on the efficiency and effectiveness maturity scale.
It's important to keep in mind that collecting and managing Data Management Inventories is something that should be performed across all phases of Data Management Lifecycle and across all Environments (i.e. Data Management Environments). Both are considered to be very important Best Practices. For example, you and/or your enterprise cannot get a complete understanding of Data Management costs or impacts without knowing all related Inventory Items in all environments. And, tracking across all lifecycle phases gives a temporal perspective that is important for things like problem analysis, historical reporting, and the reconstruction of state (i.e. Configuration Management).
NOTE: Data Management Inventories are also important for other enterprise functions, such as Architecture and Design. Such Inventories represent the foundation for understanding an enterprise's Current State and are critical for planning Future State and any related strategies, roadmaps, and transition plans for facilititating change.
Building environments that are specific to and for the discipline known as Data Management is no different than doing so for any other discipline area. The reader should, therefore, refer to the IT Environment Framework to understand such environments.
As with any professional Discipline, the place to start with when dealing with Data Management specific metrics is with standard metrics categorizations. Standard Metrics Categorizations, or what are commonly referred to as "SMCs," include but are not limited to...
Data Management Quantitative Metrics: Quantitative metrics for Data Management often revolve around the "counting" of key constructs that are associated with the Discipline. For example, the number of Data Items or Entities that have been Created, Edited or Modified, Copied or Cloned, Destroyed, Archived, Restored, etc. (Note the correlations to key Data Management Verbs!). Also, the counts for things like the number of Data Management Stakeholders, such as but not limited to Paying Customers, End Users, Employees, Consultants, etc. are also very useful.
Data Management Qualitative Metrics: Qualitative metrics for Data Management often revolve around concepts such as Data Management Defects, Failures, Problems, Incidents, and/or Issues. So, for example, if we were to capture the number of Data Management Defects (i.e. their counts) over time, we could do things like see if Defect quantities are going up or down, over time, allowing us to explore that area for things like correlating Causes and Effects.
Data Management Time Metrics: When dealing with Data Management Time Metrics, there are usually two forms. The first was introduced in the previous paragraph, which has to do with capturing and measuring things like Quantitative or Qualitative Metrics, over time. In this case, we capture other metric categories, over time, with the intent to see how they change and perform, based on modifications to the Data Management Operating Environment. The second form of Time related metrics has to do with system or operational performance, such as in the case of how long it takes to process a Data Management Request, from the time it is created to the time the Requester gets a satisfactory deliverable that allows him or her to move on with his or her work.
Data Management Utilization Metrics: Utilization Metrics specifically have to do with the consumption of Data Management specific solutions or deliverables. For example, tracking the number of Data Management Service Requests, over periods of time, along with their corresponding Data Management Deliverables, allows one to measure how active Data Management Services are against other Services that may exist within the Enterprise.
Data Management Financial Metrics: As is always the case for any single Discipline, Financial Metrics for Data Management always revolve around things like revenue, expenses, and profits, both, for operators of the Service or Services and for consumers of the Service or Services. For example, if a Data Management Request is invoked by a Data Management Customer (acting as the "Requester"), it becomes important to be able to identify and understand what the cost is to that Customer who is invoking the Request, and it also becomes important to understand why that cost is what it is. In the case of Services that do not yield revenue or profits, measuring costs is a strong way to, at very least, help understand the costs associated with each Service being performed by, within, external to, and for the Enterprise and its Customers.
Note: It's important to understand that, when it comes to metrics, enterprises should take a "Crawl," "Walk," "Run" approach to collecting, working with, and understanding them. That is, you cannot get to complex metrics collection, dissection, analysis, and understanding until you start with basic metrics and slowly work your way to more complex metrics representations.
One of the most important concepts you will learn about Data Management (or any Discipline, for that matter) is the notion of implementing the Discipline as an accountable, planned, controlled, transparent, and managed "Service."
In short, Services represent a logically "bounded" and repeatable sets of work types, activities or tasks that are performed by humans and/or machines, with the specific intent to provide outputs or deliverables, in the form of solutions for the requesting Stakeholders who are commonly considered the customers of such Services. In other words, we perform and/or provide a Service to deliver very specific solutions to very specific Stakeholders who are looking for a means to solve a certain problem they have.
A Data Management Service is defined as:
"1. A set of solutions, either transactional (i.e. Transactional Data Management Services) or dial-tone (i.e. Dial-Tone Data Management Services), that are being or have been put in place to yield an intended, controlled, expected, repeatable and measurable set of results or deliverables for Data Management specific Customers, Consumers or Clients.
NOTE: Data Management Service Consumers or Clients can be either Human Resources or Systems."
All Services, including Data Management Services, can be performed manually (i.e. by people), automatically (i.e. by machines such as Computers), or by a combination of the two (i.e. a hybrid that is both manually and automated).
Also, all Services, including Data Management Services, can be either transactional or dial tone, in nature.
In the case of Transactional Services for Data Management, a Service Request is submitted and that Request is fulfilled as part of a process that is either manual, automated, or a hybrid of both (e.g. a Service to perform maintainance on your Data Management System).
In the case of Dial Tone Services for Data Management, a Service is expected to be up, running, available, and accessible to an End User so that he/she/it may perform some controlled and highly repeatable function (e.g. a "Data Management System" that is up and running all the time).
Data Management Service Components: The successful implementation of Data Management as a set of Services for your enterprise usually implies that a number of key components have been established to support it. These components are:
Data Management Ownership: The most important thing to understand about a Data Management Service is that, in order for such a Service to be successful, there must be a clear and accountable Owner for it. That is, there needs to be a very clear and accountable named person or organization that owns and is fully responsible for the Service, all of its sub-Services and, most importantly, all of the Service's "Outcomes." Without clear ownership, Services are almost never successful. And, for those few occasions where Services are successful without clear ownership, you can assume that they're successful because the people working in those Service areas are acting as heroes, or... the those Services are just plain lucky (that kind of luck doesn't last for long).
Data Management Service Inputs: There are typically two types of inputs to any Data Management Service. The first is what is known as a "Data Management Service Request" and the second really represents any and all supporting artifacts that are necessary to support such requests, including but not limited to Data and Information in the form of Documents, either electronic or paper in form. Many would argue that the "money" to pay for the Service execution of the Request would be the third but, for now, we will assume that payment is controlled through the Data and Information provided to the Service Operators, in support of the Request.
Data Management Service Outputs: The outputs of any Service are often referred to as the Service's Deliverables. Therefore, the readers should be aware that the terms "Data Management Outputs" and "Data Management Deliverables" are synonymous and interchangeable. All work performed in any enterprise is, by default, a Service that is being performed for someone else and, therefore, all work or Services yield results. These results are the Service's Outputs or Deliverables and a good Service ensures that such Outputs are appropriately documented to the consumers of said Service. This means that for any given Data Management Service Request Type or Category there will be one or more clearly defined and documented Outputs or Deliverables, making it clear to the consumer what he, she, or they will get in response to their Request. This can be as simple as an answer to a question or as complex as the Merger of two enterprises.
Data Management Service Levels: Service Levels represent "performance agreements," contractual or otherwise, that dictate how well a Data Management Service should perform, most often keeping the Customers, Consumers, Clients or End Users of the Service in mind. Data Management Service Levels can come in many forms and are often worked out by the Customers paying for the Services and the Service Providers who sell or provide the Services. In many cases, Service Levels are also self-imposed by the Service Providers performing the Services as a means to set expectations for Service Customers. In short, Data Management Service Levels are constraints, limitations, and/or expectations that are tied directly to Data Management Service Deliverables. They represent measures for things like quality, efficiency, and cost against said Deliverables or Outputs that allow the consumer of such Services to measure what they actually get against what they expected to get.
Assuming an enterprise pursues the establishment of Data Management as a set of controlled Services, there are three common paradigms for doing so. These include:
Centralized Data Management is defined as:
"1. The term or phrase that implies establishing and/or practicing the Discipline known as Data Management as a concentric and singular set of organizations and services, usually in order to serve an entire enterprise, regardless of geographic location, further implying full centralization and no federation of any and all Data Management associated Work, Activities, Actions, Tasks, Capabilities and/or Services."
Federated Data Management, which is also referred to as Decentralized Data Management, is defined as:
"1. The term or phrase that implies establishing and/or practicing the Discipline known as Data Management in multiple pockets, communities, or organizations, further implying no centralization in the implementation and execution of Data Management associated Work, Activities, Actions, Tasks, Capabilities and/or Services."
There are clear tradeoffs to each of the two models. For example, in a Centralized paradigm, it's normally easier to coordinate work and provide broad coverage, across many areas of the enterprise and relevant stakeholders. However, it becomes far more difficult for a centralized organization to properly fund and staff resources and services in order to perform all required work across all stakeholders, in a much larger enterprise.
It's also important to note that a third paradigm also exists as an option. This is known as a Hybrid Data Management paradigm or model. In this case, there is a centralized Data Management organization that is often responsible for things like centralized governance, command, control, and communications, while federated staff and services deal with localized forms of Data Management. In this type of paradigm, federated staff and services usually report direclty into their local management but may have matrix reporting or responsibilities into the Centralized Data Management organization.
A "Principle" is defined as being: "A professed assumption, basis, tenet, doctrine, plan of action or code of conduct for activities, work or behavior." Therefore, we can deduce the definition of "a Data Management Principle" to be:
Data Management Principle: "1. A professed assumption, basis, tenet, doctrine, plan of action or code of conduct for any activities, work or behavior associated with the Discipline known as Data Management."
A "Best Practice" is defined as being: "One or more Activities, Actions, Tasks or Functions that often do not conform with strict Standards and that have evolved, over time, to be considered as conventional wisdom for consistently and repeated achieving Outcomes or Results that can be measured as being equal to or above acceptable norms." Therefore, we can deduce the definition of "a Data Management Best Practice" to be:
Data Management Best Practice: "1. One or more Data Management related Activities, Actions, Tasks or Functions that often do not conform with strict standards and that have evolved, over time, to be considered as conventional wisdom for consistently and repeatedly achieving Outcomes or Results that can be measured as being equal to or above acceptable norms."
The plural form of this term would be "Data Management Best Practices."
Common Data Management related principles and best practices exist to help achieve higher than average expectations of quality and to ease in the implementation, support, operations, and future change associated with the solutions industry professionals put in place to address the needs of this Discipline and all its related stakeholders.
While this entire document is meant to represent and serve as a set of common principles and best practices for Data Management, the following list represents a summary of some very basic examples of what implementers, supporters, and operators of Data Management should constantly be working toward:
|Principle or Best Practice||Description|
|Establish and always have very clear Ownership for Data Management.||Establishing, publishing and socializing clear Ownership for Data Management allows an enterprise and all its Resources, regardless of their geographic location, to assign accountability for all aspects of the Discipline. It also ensures that there's always at least one person that everyone can go to for transparency into the Discipline as well as for handling work that is associated with the Discipline.|
|Define, Collect, and Manage Relevant Data Management Inventories.||As an IT professional, there are probably few things that are as important as knowing what is or is not in your portfolio, as well as understanding key traits about your portfolio. You cannot achieve this without the transparency provided by your inventories. Therefore, it is critical that you clearly define, collect, manage, and govern any and all relevant Data Management inventories. Lack of Data Management Inventories means no transparency, a chaotic and immature environment, and (even worse) the implication that you don't know how to do your job.|
|Always use standard terminology for Data Management, in order to standardize communications between stakeholders.||It is often argued that the biggest mistake you can make is to create your own words and/or your own definitions, when communicating with others. There is no place where this is more accurate than in the field of Information Technology. IT Stakeholders make up their own words and definitions far too often, or let their business constituents do so. When you make up words or definitions, or you let others do so, you're creating a grave injustice for your organization. Self invented terminology and grammar often leads to poor communications, which in turn leads to redundancy of solutions, higher complexity of environments, slower delivery times, and much higher costs. Therefore, the IF4IT always recommends that you leverage standard terminology for Data Management, whenever possible.|
|Centralization of Data related data.||While often impossible to centralize and collocate all Data related data and information, especially in a geographically dispersed environment, Data Management related stakeholders should always strive to centralize all data and information. The goals are to eliminate data fragmentation, improve source of truth for data, reduce the number of systems needed to support stakeholders, reduce the complexity of solutions, improve usability, and to ultimately reduce the costs associated with Data Management.|
|Clearly define, implement, track, and analyze Data Management Metrics.||In order to successfully set up the discipline of Data Management and its related Services, it is critical to clearly define, track, and constantly analyze Data Management metrics. Such metrics include but are not limited to Supply and Demand Metrics (i.e. Operational Metrics), Performance Metrics, Quality Metrics, and Financial Metrics.|
|Transparency of Data related data.||Stakeholders should always strive to make any and all Data Management data transparent to all other appropriate stakeholders, at a minimum, and often to the entire enterprises. The exception when private user data must be protected. Many stakeholders often make the mistake of treating internal operational data as private or protected. This often creates a data silo and will often lead to internally silo-ed organizations that revolve around such data silos.|
|Do not let "perfection" of Data Management solutions stand in the way of "good enough solutions".||Often, Data Management stakeholders "overthink" solutions, leading to the impression that best-of-breed or perfect solutions are more effective than "good enough" solutions. Experience tells us that "good enough" is, almost always, the better path to follow. We live in an age where technologies grow old in the blink of an eye. Even the implementation of something that looks perfect, today, will look antiquated, tomorrow. This is especially true if your enterprise doesn't have a long term funding plan and commitment to improvements and upgrades of the solution(s) put in place.|
|Follow industry Standards, Best Practices, and Guiding Principles for Data Management, whenever possible".||One of the most common errors many enterprises make is to create solutions from scratch or without the guidance, assistance and/or experience of others who have created such solutions, before them. Whenever possible, the IF4IT recommends that you research existing Standards, Best Practices, and Guiding Principles to avoid the mistakes of others, while also gaining from their successes. Remember, we live in a vast world. Chances are very high that someone else has already experienced the pain you're about to create for yourself. Wise people will always look to learn from such people's experiences before they go down the road of implementing their own solutions.|
|Work toward and maintain a Single Source of Truth (SSoT), whenever possible.||While it may be impossible to truly maintain a Single Source of Truth (SSoT) for all data items at all times, especially in the case where the same data entity or instance enters an enterprise through unique data channels, it is an accepted, industry-wide best practice to always work toward such a goal.|
The Information Technology (IT) Learning Framework. A tutorial that helps understand Information Technology and how disciplines, such as this one, fits into the bigger picture of IT Operations.