It is crucial for a company’s success to manage humungous amounts of data, but the sad part is that most companies fail to do so, even after the emergence of Chief Data Officers (CDOs) and data-management functions. According to studies conducted cross-industry, less than 1% of unstructured data is used or analyzed at all and about 80% of analysts spend their time just exploring and preparing data. All this can lead to more frequent data breaches.
You thus need to have a coherent strategy in place to organize, govern, analyze, and deploy an organization’s information assets. This blog helps you learn how to build a robust data strategy applicable across industries and various levels of data maturity. As you build this strategy and practice it across your organization, your organization develops superior data management and analytics essential capabilities, those that support managerial decision making and increase financial performance.
Types of Data Strategies
Before we start building a robust strategy that your business needs, let’s understand the difference between data and information as well as between data architecture and information architecture. In the words of Peter Drucker, information may be defined as “data endowed with purpose and relevance”. It is the data architecture of a company that defines how data is gathered, stored, transformed, distributed, and consumed. Information architecture controls the rules and processes that transform data into useful information.
Let us understand this better with the help of an example. Data architecture might feed raw daily sales and advertising data into information architecture systems, like marketing dashboards. In these dashboards, this data is integrated and analyzed to show relationships between ad spend and sales by region and channel.
A more pragmatic and flexible approach to data and information architecture consists of a single source of truth (SSOT) and multiple versions of the truth (MVOTs). The SSOT operates at the data level, while the MVOT supports information management. An SSOT is a source from which multiple versions of the truth (MVOTs) are developed. In most organizations, the concept of a single version of the truth is fully comprehended and accepted by IT and across the business.
As multiple groups within functions or units change, label, and report data, they create different, controlled versions of the truth that, at the time of being queried, produce consistent, customized responses as per the predetermined needs of the groups. MVOTs inferred from a common SSOT support high-level decision making.
Remember not having an SSOT can lead to confusion. But the fortunate part is that artificial intelligence (AI) tools that can sort through such chaotic data to assemble an SSOT are easily becoming available. Leveraging AI enables companies to shut down their redundant systems and thus save on large IT costs.
The SSOT-MVOTs model is conceptually straightforward, but it needs robust data controls, technology, governance, and standards. Ideally, senior executives actively participate in data governance committees and boards. What’s critical to understand is that SSOTs remain unique and valid and that MVOTs deviate from the source only in a carefully controlled manner.
Pro tip: It needs flexible data and information architectures that allow single as well as multiple versions of the truth to support an offensive-defensive approach to data strategy.
Offense vs. Defense
Data offense and defense are distinguished by different business objectives and the activities designed to address them. Data offense emphasizes supporting business objectives, like increasing profitability, revenue, and customer satisfaction. Typically, it consists of activities that generate customer insights or integrate different market and customer data to support managerial decision making through interactive dashboards.
Offensive activities are the most relevant for customer-focused business functions, like sales and marketing. Often, they are more real-time than defensive work is.
Data defense, on the other hand, is about mitigating downside risk. It includes activities like ensuring regulatory compliance, building systems to prevent theft, and identifying and limiting fraud through analytics. Besides, defensive efforts guarantee the integrity of data flowing through the internal systems of a company by detecting, standardizing, and governing authoritative data sources in an SSOT.
Every organization requires offense as well as defense to succeed and striking the right balance is tricky. The challenge for CDOs and the rest of the C-suite is to set up the appropriate trade-offs between offense and defense, assuring the best balance that supports the overall strategy of the company.
The Elements of Data Strategy
Striking a Balance Between Offense and Defense
Offense and defense often need varying approaches from IT and the data management organization. The offense is about partnering with business leaders on strategic initiatives. Leaders are always happy to collaborate on optimizing marketing and trade promotion spending. Defense, on the other hand, is operational and day-to-day.
CDOs find that their ideal data strategy focuses on offense and flexibility or defense and control. It’s not prudent to default to a 50/50 split, instead of making well-thought-out trade-offs. To ascertain a company’s existing and desired positions on the offense-defense spectrum, the CDO must keep in mind, the company’s overall strategy, the regulatory environment, the maturity of its data-management practices, the data abilities of its competitors, and most importantly, the size of its data budget.
Irrespective of what industry a company belongs to, its position on the offense-defense spectrum is hardly static. As competitive pressure intensifies, an insurer may decide to focus more on offensive activities. For instance, a hedge fund may find itself in a difficult regulatory environment that demands rebalancing its data strategy towards defense. How the data strategy of an organization changes in velocity and direction is a function of its overall strategy, competition, culture, and market.
Are you Ready with Your Data Strategy?
Emerging technologies may enable the next gen of data management capabilities, possibly simplifying the execution of offensive and defensive data strategies. Machine learning (ML) is easing the creation of an SSOT in most companies. The promise here is more dynamic, less-expensive SSOTs and MVOTs. The data strategy framework will emerge more relevant as blockchain technology grows more prominent. Companies that haven’t yet built a data strategy and a robust data-management function will be required to catch up faster or begin to plan for their exit.