Tips for Developing a Data Strategy: Managing Your Data Appropriately - dummies

Tips for Developing a Data Strategy: Managing Your Data Appropriately

By Ulrika Jägare

After your company’s objectives have become clearer, your CDO, as part of an overall data science strategy, needs to create a business-driven data strategy fleshed out with a significant level of detail. In addition, that person needs to define the scope of the desired data-driven culture and mindset for your company and move to drive that culture forward. Here, you discover what a CDO needs to keep in mind in order to accomplish these tasks, as well as an example of a data strategy scope.

Data science: Caring for your data

One key aspect in any data strategy involves caring for your data as if it were your lifeblood — because it is. You need to address data quality and integration issues as key factors of your data strategy, and you need to align your data governance programs with your organizational goals, making sure you define all strategies, policies, processes, and standards in support of those goals.

Organizations should assess their current state and develop plans to achieve an appropriate level of maturity in terms of data governance over a specific period. It’s important to recognize that data governance is never complete; by necessity, it evolves, just as corporate needs and goals, technology, and legal and regulatory aspects do.

Governance programs can range from establishing company-level, business-driven data and information programs for data integrators, to establishing customized, segment-based programs for the business optimizers and market disruptors/innovators. However, even the best strategy can falter if the business culture isn’t willing to change. Data integrators flourish in an evidence-based operational environment where data and research is used to establish a data-driven culture, whereas business optimizers and market disruptors/innovators need to adopt a “fail-fast” agile software development culture in order to increase speed-to-market and innovation.

Data Science: Democratizing the data

As important as it is to understand the value of the data your company has access to, it’s equally important to make sure that the data is easily available to those who need to work with it. That’s what democratizing your data really means. Given its importance, you should strive to make sure that this democratization occurs throughout your organization. The fact of the matter is, everyone in your company makes business decisions every single day, and those decisions need to be grounded in a thorough understanding of all available data. It has become obvious that data-driven decisions are better decisions, so why wouldn’t you choose to provide people with access to the data they need in order to make better decisions?

Although most people can understand the need for data democratization, it isn’t at all uncommon for a company´s data strategy to instead focus on locking up the data — just to be on the safe side. Nothing, however, could be more devastating for the value realization of the data for your business than adopting a bunker mentality about data. The way to start generating internal and external value on your data is to use it, not lock it up. Even adopting a radical approach of a totally open data environment internally is better than being too restrictive in terms of how data is made available and shared in the company.

Data science: Driving data standardization

A third key component in any data strategy is to standardize to scale quickly and efficiently. Data standardization is an important component for success — one that should not be underestimated. A company cannot hope to achieve goals that assume a 360-degree view of all customers underpinned by the correct data without a common set of data definitions and structures across the company and the customers.

TM Forum, a nonprofit industry association for service providers and their suppliers in the telecommunications industry, developed something they call the Information Framework (SID) in concert with professionals from the communications and information industries working collaboratively to provide a universal information and data model. (The SID part of the name comes from Shared Information Data model.) The benefits of this common model come from its ability to significantly support increased standardization around data in the telecommunications space and include aspects such as;

  • Faster time to market for new products and services
  • Cheaper data and systems integration
  • Less data management time
  • Reduced cost and support when implementing multiple technologies

Organizations have long recognized the need to seek standardization in their transactional data structures, but they need to realize the importance of seeking standardization in their analytical data structures as well. Traditional analytics and business intelligence setups continue to use data warehouses and data marts as their primary data repositories, and yes, they are still highly valuable to data-driven organizations, but enabling dynamic big data analytics and machine learning/artificial intelligence solutions requires a different structure in order to be effective.

Data science: Structuring the data strategy

The act of creating a data strategy is a chance to generate data conversations, educate executives, and identify exciting new data-enabled opportunities for the organization. In fact, the process of creating a data strategy may generate political support, changes in culture and mindset, and new business objectives and priorities that are even more valuable than the data strategy itself. But what should the data strategy actually include? The list below gives you an idea.

  • Data-centric vision and business objectives including user scenarios
  • Strategic data principles, including treating data as an asset
  • Guidelines for data security, data rights, and ethical considerations
  • Data management principles, including data governance and data quality
  • Data infrastructure principles regarding data architecture, data acquisition, data storage, and data processing
  • Data scope, including priorities over time

Don´t mix-up the data strategy with the data science strategy. The main difference is that the data strategy is focused on the strategic direction and principles for the data and is a subset of the data science strategy. The data science strategy includes the data strategy, but also aspects such as organization, people, culture and mindset, data science competence and roles, managing change, measurements, and business commercial implications on the company portfolio.