How to formulate a Data Strategy

Dpulse believes that organizations need to take the following steps in order to be able to maximize the value of their data;

• The formulation of a data strategy.

• Acquiring the appropriate technology for both hardware and software > and the integration of data sources.

• Transformation of the organization.

• Making sure the organization is supported by appropriate data > governance and data security.

The key factor in building a data-driven organization is a clear and thought-through data strategy derived from the business strategy, led by a data champion with insight into the overarching goals and operation processes of the business.

The ultimate goal of many data projects is to build a culture of customer- and process analytics, where data teams are responsible for building a centralized repository for the organization’s data, and equipping the rest of the business with the knowledge and skills to use and interpret it effectively.

We will now take a closer look at the different steps that need to be taken to come to a meaningful data strategy;

6 areas that will be looked at:

• Business Strategy & Dashboard

• Data availability & integration

• Analytics – Data competenties

• Data Governance

• Security & GDPR

• Organizational readiness

Business Strategy & Dashboard

The data strategy should be derived from and aligned with the company strategy, mission and vision. The data strategy should be focusing on the customer and/or on the internal processes. Does the organization support the direction the management wants to go into? Make sure you understand the pain points and the business needs.

The dashboard should support the measurement of the KPI’s that also should be derived from the overall goals. The KPI’s should be SMART. Are your employees satisfied with the dashboard and do they use it for the day-to-day operations? Can the visualization be improved (interaction) so the data can be better understood and used by a broader part of the operation? Is the dashboard also supporting autonomous teams that enables delegation of decision making?

Data availability & integration

In this stage we want to understand which data is available and how the relationship is between the different sources of data; the data model or data architecture. It makes clear which data is generated and where it is stored. As an example; data from social media (Twitter, Facebook, Instagram) might be available, but not stored as it might not be clear how this data can be used. Data lakes/date warehouse can do the job, but what will be the cost of storing and analyzing this data. For example through sentiment analysis, and what are the expected advantages of doing this in terms of returns.

These considerations should result in a high level Hardware and Software architecture with an overview of the data that need to be integrated.

Analytics – Data competencies

The data strategy should give an overview of the different big data techniques and tools that can be used and what they can yield. Of course with business cases, taking into account the data competencies the organization can call upon.

Big Data analytical techniques are, amongst other, machine learning, deep learning, predictive analytics and prescriptive analytics. With regards to software tools decisions need to be taken if business intelligence tools like Power BI and Tableau (low code tools) are used for analysis or tools like R and Python, highly reproducible business intelligence tools. We want to emphasis that not only the return on investment needs to be looked at, but also the organizational readiness. Are the new ways of working accepted and is the workforce ready to use the new techniques. A long term training and/or hiring plan should be part of the Data Strategy.

Data Governance

Data governance is about quality and consistency of data. Data analysis should lead to reliable results and it should be able to be executed when needed.

The organization needs to have a written policy in all stages of the data lifecycle, including acquiring, maintaining, using, and archiving or destroying data. The policy needs to be reviewed and audited regularly. A description of the required data management roles along with their responsibilities and their decision rights should be available.

Security & GDPR

A comprehensive security and privacy procedure that is regularly reviewed should be in place. Is there a continuity plan for data services and infrastructure in an event of a data breach, loss, or other disaster?

Organizational readiness

The organizational impact from going to a traditional organization to a data driven organization shouldn’t be underestimated. It is critical to make sure from the start of the process that the employees in all levels of the organization are involved and part of the decisions that are taken and aware of the direction the company is going in.

Therefor Dpulse advises to:

• Communicate the vision and strategy. Do it frequently and carefully.

• Create a sense of urgency for your new plan, using as much data as possible. Visualize it as effectively as possible. Tell stories.

• Project teams should be assigned to work on different parts of the project.

• Identify people that need training in the new analytical techniques and tools.

• In general people should be trained to use data to make measurable business performance improvements and informed decisions. The ultimate goal is to ‘democratize’ access to data.

• Is the organizational structure still up to speed with the new requirements of the data-driven organization - is it time for a CDO – a Chief Data Officer?

Once the assessment has been finalized DPulse advises to summarize the Data in a document that proposes what to do now and in the future with your data (and what not), time-phased, with SMART goals and with business cases for investment and support in order to explain the ‘Why’. Along with a draft implementation roadmap with priorities, projects and action items with milestones, usually looking 3 years ahead. Important is to review and update the Data Strategy at least yearly.

Dpulse’ experience is that depending on the complexity and the ‘data maturity’ of the organization it will take 6 – 16 weeks to formulate the Data Strategy.

If you find it hard to execute the assessment yourself or it can’t be prioritized, DPulse is able to help you to take this important step in making your data a critical and valuable asset.