Data Management

What are Key Data Management Principles?

Ethereal blue light streaks across a black void, symbolizing the guiding principles of data management.

Data Management Principles – The Key to Better Data Management

The impact of the global pandemic has highlighted that your primary asset is your people, and ensuring their health and wellbeing is a key management principle. The second most precious asset is data, which requires a similar management perspective to maintain sustainability and competitiveness.

Globally renowned futurist Bernard Marr stated,

“As the world becomes smarter and smarter, data becomes the key to competitive advantage, meaning a company’s ability to compete will increasingly be driven by how well it can leverage data, apply analytics and implement new technologies. In fact, according to the International Institute for Analytics, by 2020, businesses using data will see $430 billion in productivity benefits over competitors who are not using data.”

Senior management leads the creation of straightforward principles for the acquisition and management of data. These principles are then linked to SMART Outcome Key Results (OKRs) to ensure that your data controls and maintenance allow agile problem solving, decision making, customer analysis and business protection.

The most fundamental Data Management Principles and Data Governance principles are:

  • Design a strategy and vision on what data is required to keep you in business secure and competitive
  • Create data accountability by having every piece of information owned by a business domain leader or product owner
  • Make data a responsibility of everyone
  • Understand and monitor the lifecycle of data from acquisition to destruction
  • Benefit from the DevOps and Agile principles of “Right the First Time” to ensure quality, timely, accurate, secure, and relevant data
  • Create metadata to automate the use, storage and maintenance of data

Design a data management strategy 

You generate data every time you perform a task. You acquire data from the Internet, IoT devices, mobile devices, newsfeeds, vendors, social media and more. How do you know what data is needed and that it is of the quality and type required? What will you do with the data once available in your business? Senior management can address these concerns using a powerful lean technique called Value Stream Management (VSM). Using nothing more than a few post-it notes or tools such as Tasktop, leaders can better understand:

  • How and where data is used
  • What data is used
  • What happens to data after use
  • Where is data redundant or repetitive
  • Where are the data gaps that block the flow of work

The outcome of a value stream mapping event is a list of iterative steps that will create the vision for mature data management as defined within your master data management framework. Your data management strategy should articulate:

  • Guardrail governance controls and policies
  • Archival, storage, and recovery of data policies against global regulatory requirements
  • Domain or business area rules for data applications, services and products
  • Role definitions which could include significant vendor partners
  • Data protection and security, with actions to take if hacked or data infringement occurs
  • Pilot schemes to ensure that your strategy is viable, flexible and will not negate your agility in the market
  • Data incident and defect monitoring, alerting and actions
  • Financial rules for the cost of data acquisition, control and storage
  • Training on data use and management is required

Your critical Data Management Principles should enable an agile and flexible organization ready for the digital economy by giving each principle an OKR or KPI. You should track these metrics on real-time dashboards for all critical data and as needed for the rest.

Roles in the data management system

Senior management must make it clear that data management, security and safety is the obligation of every employee and vendor partner. Poor data mismanagement should be finable, costing an employee a job offence or a vendor its partnership with you.

Data management is a team effort. The VSM exercises mentioned earlier mapped the flow of data across tasks concerning customer services and products. This visualization acts as an internal marketing tool for managers at all levels to explain why data management is key to the success of an employee’s role and the business.

Data management roles should have the following functions:

  • Data Owners: every piece of data and information must have accountable staff that will govern and oversee the acquisition, usage, control, financial management, security and storage of all data under their value streams. Data Owners are business domain managers or product o
  • Data Stewards: these are the individuals responsible for the quality of data flowing across a value stream(s). Quality means that data should be timely, relevant, and complete in all attributes or fields. Examples of data stewards are data process or governance owners, developers, database administrators, team leads, SaaS suppliers or external data providers.
  • Data Custodians: are those that create and manage your data repositories and applications. Database administrators, developers, SaaS, and data providers are examples of Data Custodians.
  • Data users: the rest of your organization, each with a role to protect and benefit from the use and management of data.

Risk and compliance teams working with data owners should at least annually review these roles to ensure their relevance.

Control of data throughout its lifecycle

Good data quality management enhances problem solving, decision making and performance of tasks. Good data management practices will be rewarded by your customers with their trust in your organization via the services they purchase. The following comments and questions will guide the maturity of data management activities based on your key Data Management Principles.

Data Acquisition:

  • Who can acquire data?
  • Where can data be obtained?
  • What tests validate that data can be accepted into your data repositories or applications?

Data Creation:

  • What processes or tasks generate data?
  • How many of these tasks are manual?
  • If manual, should they be automated?
  • Is the data being created for use elsewhere?
  • Is the data being created already available?
  • If data is of no use, can we stop making this data?
  • Is the data ready for consumption, or does it have to be supplemented?

Data Sharing:

  • Is data easily accessible, timely, relevant, complete, accurate and of quality?
  • What actions are required if a task(s) cannot access data?
  • How is data validated as it continues its journey downstream in a process?
  • If data changes, what happens to the original data?
  • How certain are you that approved applications or individuals access your data and that there is no unauthorized use?
  • Is your data being offloaded (to tools such as Excel) and, therefore, no longer under your management?
  • Do applications or manual processes generate duplicate data, and if so, how can this be mitigated?
  • How is data cleansed before use? (A common practice for externally-sourced data)
  • Is the training on data management appropriate to each role, and does it occur?

Data Storage:

  • Where are we storing data? Our infrastructure or less expensive cloud storage?
  • Do we make the best use of data warehouses or data lakes to store data for product and service use?
  • How do you know if the data is stale or no longer required?
  • Data storage options are hot if needed immediately or cold if not. Both of these have cost implications that should be under the supervision of business domain managers or product owners.

Data archiving and recovery:

  • Where is the best place for this data to reside, be archived, and backed up?
  • How certain are we that we can recover data when and as needed?
  • Do we take advantage of cloud storage options and business continuity options?
  • What length of time or regulatory rules must be adhered to when storing data?
  • Is our data stored someplace that could be illegal?
  • When upgrading technology, do we ensure that archived data is still retrievable and usable by our products and services?
  • How often do we test business continuity or legal compliance?

Data deletion and destruction:

  • Data may be deleted or destroyed based upon several variables such as time or regulatory mandate
  • The act of data deletion must be checked and verified, both manually and via automation
  • How do you know that data has not been inadvertently destroyed and what actions are necessary to recover lost data?

Metadata strategy in data management

Each aspect of data movement, change, storage, security, and deletion should be logged using an approved data control toolset and verified by the data owner. The creation of a metadata catalog will assist in this endeavor.

All data has a set of attributes such as:

  • Time created
  • Application used in
  • Time to keep
  • Where acquired or created
  • Quality metrics (accessibility, timeliness, relevance, etc.)

Metadata captures these attributes and stores them in a series of catalogs. An analogy would be the card catalog of a library showing information about all the books it has.  Metadata simplifies the control and governance of data by ensuring that all data characteristics are logged, monitored, and reported as dictated by data management policies. Metadata also confirms that the key Data Management Principles are being followed and can act as the input to leadership data dashboards.

Properly defined metadata will show:

  • Where data was acquired or created
  • How data was accessed and by what or whom
  • Date last used or modified
  • Improvement of data management activities
  • Application enhancements regarding data
  • Where to receive or store data
  • How data benefits data analysis and decision making
  • Automation possibilities for data deletion or archival tools

Summary of Key Data Management Principles 

Data is your second most valuable asset, after staff. Senior leadership must signal their belief in the principles and importance of data management. Senior leadership should regularly review designing an agile strategy, assigning data roles, assuring policy adherence, controlling the cost of data management, and assessing if data is serving its intended purpose.

Senior leadership should create a Data Management Manifesto that highlights its vision and principles. Hand each domain leader or product owner four post-it notes, two green and two red. Ask the leaders to write one comment they expect about data on each green note and one they do not desire on the red. Once categorized, the company can see their shared views and derive their principles and actions accordingly (lean technique). Placing a signed data manifesto onto the corporate website will help sell the importance of data management, control and security.

Examples of Key Data Management Principles derived from a Manifesto exercise are:

  • We value data and expect it to be suitable and relevant every time
  • All data will be owned by an accountable party
  • Our rules and policies must be flexible but also adhered to, and we will monitor this on a real-time basis for critical data and as needed for the rest
  • Data security is of paramount importance to our company
  • It is the role of everyone to manage data

Your key principles should help you define, design, create and manage your data repositories, customer data platforms (such as customer relationship management), financial records, competitive analysis and response, and product and service delivery improvements.