Your company needs a data strategy, or it will look at exit strategies; how and why you should use data
The amount of data generated per year has steadily increased in the past ten years, and predictions for 2023 show a generated value of 120 zettabytes. To put that into perspective, one zettabyte equals 1 billion terabytes of data.
Of course, not all of this data can be harnessed for some form of gain. However, it does display rampant data generation and why the need for more computing power has been a significant topic in recent years.
Jumping to the present day, with advancements in technology as well as the explosion of AI, data utilization is now cheaper than ever and more widely available. Harnessing the power of your data will, at one point in the future, become an essential part of any business.
Previously, only large companies got to play the data game due to how expensive it was to run such operations within a company. Today, companies can do the same with far fewer resources invested.
Studies have shown that, on average, less than half of an organization's structured data is actively used in making decisions, while less than 1% of its unstructured data is analyzed or used at all. Data Analysts' time (up to 80%) is spent discovering and preparing data. Meanwhile, 70% of employees have access to information they shouldn't.
⚖️ Balance is 🔑 key
There are plenty of ways of approaching your data strategy. In this article, we will be focusing on a business-oriented framework. This framework consists of Offense and Defense, which have tradeoffs depending on the goals and objectives your company wants to achieve. Setting these expectations is more than just a concern of the CIO and CDO. Intelligent data management is a collaborative process involving all C-level executives, starting with the CEO.
Defense🛡️ vs. Offense ⚔️
Both approaches to your data strategy are distinguished by business objectives and the activities designed to address them.
🛡️ Defense is about:
Minimizing risks through compliance with regulations (data privacy rules and financial report integrity)
Fraud detection and limitation through analysis
Data theft prevention
It also ensures the integrity of data flowing through the company's system by following a series of processes: identifying, standardizing, and governing data sources in what is called a "Single source of truth" (SSOT).
⚔️ Offense focuses more on business objectives like:
Increasing revenue
Increasing profitability
Increasing customer satisfaction
Associated activities include customer insight generation (data analysis) and customer and market data integration for data-driven managerial decision-making. These actions are relevant to customer-focused sales & marketing functions. They are more real-time and use data faster than defensive activities.
An exception would be fraud protection, which is part of a defensive strategy, but using real-time data and literal seconds can be crucial.
Every company should strive to implement both an offense and a defense, but achieving balance can be challenging. A 50/50 approach is only sometimes optimal and will vary depending on your industry.
The decision to lean more toward one side or the other is rooted in the contradiction between standardized data and flexible data.
If data is standardized, it's easier to execute a defensive strategy, such as compliance with regulations and data-access control implementation. 🛡️Defense
The more flexible data is, the easier it is to transform or interpret it for optimal integration in sales, marketing, and other customer-focused functions. ⚔️Offense
ℹ️ Data, 📊 Information, and 🏗️ Architectures
Firstly, it's crucial to understand the difference between information and data.
Raw data, such as supply costs, customer retention, sales, etc., provides limited value.
Integrating multiple instances of raw data, you can transform it into useful information that can be used for decision-making.
Let's say you run an online business.
By combining data on customer demographics, purchase history, and browsing behavior, you can create personalized recommendations for each customer, leading to increased sales and customer satisfaction.
Secondly, we need to differentiate between data architecture and information architecture.
Data architecture covers how data is collected, stored, transformed, distributed, and consumed. It includes structured formats such as file systems and databases and systems for connecting data with business processes that utilize it.
Information architecture governs the conversion of raw data into valuable and sharable information. The information architecture of a company can convert raw sales data into useful information by organizing and categorizing the data by product, region, and time period, which can then be used to identify trends and opportunities for growth.
🛡️ SSOT & ⚔️ MVOTs
As mentioned, balancing defense and offense is key to curating a successful data strategy. A flexible and realistic approach involves a single source of truth (SSOT) at a data level and multiple versions of the truth (MVOTs) for information management.
An SSOT is often a logical, cloud-based repository that contains one authoritative copy of all crucial data, such as customers, suppliers, and product details. To ensure the data is reliable, it must have robust data provenance and governance controls for both offense and defense.
The most important thing is using a common language not specific to a particular business unit or function. Creating an SSOT allows for the circulation of quality data in a practical way, which can cut costs by shutting down redundant systems.
MVOTs use data provided by the SSOT to create data into business-specific information. Thus, various groups within units or functions transform, label, and report data; they make distinct controlled versions of the truth.
Here are two examples of a good and a lousy MOVT:
✅ Good MVOT example:
A marketing team uses an SSOT to manage a company's customer data, including their contact information and purchase history. The sales team also has access to this data, which helps them make informed decisions about which products to pitch to customers. However, the sales team has the flexibility to add notes and update the data based on their conversations with customers. This creates a "good" MVOT because the sales team's input adds valuable context to the customer data, which can help improve the accuracy of the SSOT.
🚫 Bad MVOT example:
An accounting team uses an SSOT to manage a company's financial data, including expenses and revenue. However, the sales team sometimes records revenue in a separate system that is not synchronized with the accounting system. This creates a "bad" MVOT because multiple versions of the truth about the company's revenue can lead to inaccurate financial reporting and decision-making.
Business units may need MVOTs because they have different priorities and goals. Customized interpretations of data can help them gain insights tailored to their specific needs, improving decision-making and goal achievement.
🧩 Determining your strategy
Although C-level executives as a whole are responsible for discussing the matter, the CDO commonly weighs out the tradeoffs of leaning into defense or offense. The CDO must take into consideration the company's overall strategy, regulatory environment, data capabilities of competitors, maturity of its data-management practices, and the size of its data budget.
🛡️ Defensive data strategy is crucial for industries that require protecting sensitive data and heavy regulations, like healthcare, to prevent cyber attacks and legal repercussions.
⚔️Technology companies such as Google, Facebook, and Amazon commonly use an Offensive data strategy to gain a competitive edge by collecting and analyzing massive amounts of data.
🛡️/ ⚔️ The retail industry uses a balanced data strategy by collecting and analyzing data to improve customer service and optimize inventory while protecting customer data from cyber threats.
Regardless of industry, data strategies are rarely static. How a data strategy changes direction and velocity will be determined by overall strategy, competition, and market.
🗄️ Organizing data management
Typically, a stand-alone CDO accountable for the entire organization, ensuring data policies, governance, and standards are consistent throughout, is suitable for a defensive strategy.
On the contrary, an offensive strategy can be better executed through decentralized data management, with a CDO for each business unit and most corporate functions. Unit CDOs own their respective versions of the truth, while the enterprise CDO owns the SSOT.
Another important factor when choosing between centralized and decentralized data functions is funding.
Centralized budgets may seem more significant purely because they are concentrated under one CDO. They focus on minimizing risk, cutting costs, and providing better data controls and regulatory insight.
Decentralized budgets are more investment-focused, have a more tangible ROI, and may not require the same level of investment in technology and infrastructure.
In the age of big data, businesses must take a strategic approach to data management to stay competitive. The offense-defense framework presented in this article provides a practical way for companies to think about their data strategy regarding their specific business objectives. However, it's important to remember that balancing offense and defense is crucial, as a lopsided approach can leave a company vulnerable to risks or miss out on growth opportunities.
Ultimately, data management is not just a technical issue but a collaborative process that involves all C-level executives, starting with the CEO. With the right approach, companies can harness the power of their data to drive business growth and stay ahead of the competition.
