Close this search box.

Creating an effective data strategy

data leaders

Data strategy is essential for any organisation serious about maximising the value of its data for commercial and operational advantage.

Data strategy aims to make data available, maintain data consistency and accuracy and support data security, according to a new report from Alation. Properly implemented, a data strategy can unlock immense value in areas such as: Operational efficiency; Process optimisation; Rapid decision-making; New and enhanced revenue streams; Optimised customer satisfaction and Accelerated onboarding for data users.

These areas apply to virtually all businesses and sectors, further underscoring the importance of best practices around data management for companies operating in dynamic and competitive marketplaces.

Offensive strategies are focused on extracting value from data to understand customer needs and behaviours, and then delivering enhanced products and services. These strategies may use data to create customer insights or integrate data from customers and markets to inform high-level decision-making. This allows data to deliver a positive overall impact on profitability, revenue, and customer satisfaction.

Defensive strategies focus on mitigating risk and instilling security to ensure compliance with local, national and international regulation. However, offensive and defensive data strategies are not mutually exclusive. All businesses will have some requirements to protect data for ethical and legal reasons. And all businesses can, to some degree, benefit from greater insight to help understand their customers and markets, and then shape their offering accordingly.

Any organisation seeking to implement an effective data strategy will need to: Assign responsibility for implementing policies and processes; Define policies for sharing and processing data; Create processes for naming and storing data and Establish measurements for keeping data clean and usable.

Identify: A company needs to know what data it already has. All critical data elements (CDEs) should be collated and inventoried with relevant metadata, then classified into relevant categories and curated — organised and managed with active metadata management.

Store: In businesses where individual departments have their own databases for metadata management, data will be siloed, meaning it cannot be shared and used across the business and its full power is therefore not harnessed. A platform that centralises metadata is, therefore, critical. This will ensure collection across multiple tools and sources, supporting the productive reuse of metadata, visibility into data history, and effective governance and stewardship. A platform also ensures the scalability and flexibility needed for analytics, while assisting different departments in understanding the value of data lineage.

Curate: Data needs to be packaged in a way that is easy for users to access and understand. All data needs to be cleansed, validated, and then transformed into a format that enables it to be understood and used across the organisation. Once the most critical data is curated, templates should then be created for a business glossary, data dictionary, and business metadata. This will help organise data vocabulary and track how many data assets or terms are uploaded within the network.

Process: It’s vital to establish clear processes for creating, sharing, and governing data. The presence of data technology does not itself guarantee that individuals know how to effectively access and use data. However, for data to be democratised effectively, it must be made available to non-technical users in a way that aids understanding. Training on policies, guidelines, and technology itself is key.

Also key is ensuring the right analysts are in place to support business functions. Some business processes may need to be reviewed to include data analysis – even going as far as requiring specific data to make a business decision. Creating a clear process with documented steps will help.

Govern: Modern data governance models need to be flexible and dynamic while taking an active approach to risk management and compliance with local and global regulation. They should respond to multiple styles and be sensitive to context while encouraging innovation.

Two main governance models exist: a centralised model, where rules around data use and processes are set by one group; and a decentralised or federated model, where several groups will have authority over data. This model tends to be adapted in organisations with differing departmental needs. The model chosen will ultimately depend on the needs of the organisation.

Whichever route is chosen, the data governance model must be able to evolve to meet changing organisational needs, ensuring consistently high data quality, effective usage, and continued regulatory compliance.  New platforms can effectively democratise data to more people, using AI and ML to mask sensitive or personal information contained therein. Such solutions conceal sensitive data from those without the authority to access it (meeting compliance demands) while still allowing them to analyse it and glean new insights.

CTS The industrialisation of IT
CTS - Industrialisation of IT
Related Posts
CTS The industrialisation of IT
Others have also viewed

UK businesses see boosting connectivity as integral to growth

Study reveals a great opportunity for alternative network providers (AltNets) to meet growing demand for ...

Germany Energy Efficiency Act demonstrates importance of data centre supply chain collaboration

Following the signing into law of Germany’s Energy Efficiency Act (EnEfG), energy solutions specialist Aggreko ...

Systemair look to Infor’s cloud solution to deliver more sustainable products

Systemair is moving its core business system to Infor CloudSuite Manufacturing, aiming at smoother integration ...
Data Centre

Vertiv collaborates with Intel on liquid cooled solution

Vertiv is collaborating with Intel to provide a liquid cooling solution that will support the ...