“We are informing ourselves to death.” That was the warning cultural critic Neil Postman gave in 1990. He argued that data, once scarce, is now super-abundant — and increasingly overwhelming. Many clients echo his sentiment: “We are glutted with information, drowning in information, have no control over it, don’t know what to do with it.”
Technology giveth, and technology taketh away. While the benefits of the digital revolution are plain to see, we must not forget these challenges it has introduced. The economic value of data may be hard to quantify, but the costs are clear: inefficiencies, reduced productivity, compliance risks and reputational damage to name a few. According to the Data Management Body of Knowledge (DMBoK), organisations spend 10-30% of revenue handling data quality issues. In 2016 IBM put this at annual cost of $3.1 trillion across the US. Consequently, organisations struggle with data that is inaccessible, untrustworthy and often not relevant to their business problems.
Managing Data means managing the quality of data
The purpose of data management is to make data usable. If data is collected, stored and secured, but never used, it’s not an asset, it’s a liability! (I’ve argued in a previous post why data must be expressed in economic terms.) Data management is a multi-faceted discipline that requires leadership and commitment. Therefore, it is crucial to align with business priorities and secure stakeholder buy-in. Look no further than data quality. As the DMBoK puts it, “managing data means managing the quality of data.” If you’re responsible for data, it’s imperative to develop a business case to address quality concerns. The case must be driven by the business. Data requirements should be expressed in business terms, reflecting the needs of the organization—not shaped by the latest IT solution being pitched by a vendor. This is the approach employed by the Certified Data Management Professional (CDMP).
However, not all see it this way. While both claim to be data-driven, the ‘data hoarder’ takes a more is better approach. They gravitate toward new tech and believe AI will solve their business problems.
What does it mean to be data-driven?
Claiming to be data-driven is like claiming to eat healthy: most people believe they are, until they actually measure it. Humans are food-driven. We convert food into energy, which fuels the body. Modern organisations proudly claim to be data-driven, acting on analytics, not instinct. The ‘information value chain’ tells us that once data is converted into knowledge it becomes the fuel for decision-making.
Like a competitive eater, the data hoarder gets to work, ‘how can I ingest more data, more quickly’.
The CDMP is not so gluttonous. They recognise that like data, not all food is created equally. Fast food may be convenient, but this is at the expense of nutritional value. In the same way low-quality data may be easy to collect, it will likely lead to similar short and long-term complications further down the pipeline. A doctor would suggest a more balanced diet. Thankfully this task is made easier through regulation, which compels companies to display nutritional information. This metadata often found on packaging helps consumers make more informed decisions. Similarly, CDMPs assess the quality of data using nine dimensions: Accuracy, Currency, Consistency, Completeness, Reasonableness, Uniqueness, Timeliness, Validity, and Integrity. These dimensions help determine whether data is fit for consumption.
1-10-100 cost rule
Just as proteins, fats, carbs drive nutritional value, these data quality dimensions are the key drivers of the economic value of data. Conversely, they can drive great costs. Let’s compare our two persona’s approaches. The CDMP considers quality from capture and monitors it throughout the pipeline. The data hoarder isn’t checking the packaging; they assume all data is fit for purpose. The 1-10-100 cost rule illustrates how the later a data issue is discovered, the more expensive it becomes to fix. For example, if half the records in a dataset fail a validity test in development it may cost $1 to fix at that point. If it is only caught in testing that may be $10. If it remains undetected until production this may cost $100. However, by then it is too late. Once users lose trust is it difficult to regain it.
Tech Is the Distraction — Data Is the Trick
When faced with data challenges, many organisations instinctively reach for technology. It’s like buying a new oven and expecting to bake a better cake, without changing the recipe.
‘Big Tech’ has sold the narrative that investing in IT infrastructure will magically solve data problems. Just like a magician, they are diverting your attention from what really matters: the data. For executives, data is on their lips, but it’s IT that opens their wallets. It’s easy to see why: since 1970, processing power has increased 217 million times, and the cost of storing one terabyte of data has plummeted from billions to under $100. Humans’ evolution is a longer game; we process data with the same brains as the Ancient Egyptians.
The CDMP isn’t so easily wooed. With their DMBoK held high, they champion data, not technology as the true source of competitive advantage. They share Rashi Glazer’s view that IT is “merely the enabler of growth in the production and distribution of data”.
Does more data lead to better decisions?
The data hoarder on the other hand, is caught up in the hype. You can’t blame them, it’s natural. As Moody and Walsh (1999) observed, empirical studies show that decision makers’ perceived value of information continues to increase beyond the point of information overload. Tragically, not only does this lead to reduced performance it actually increases the decision-makers confidence and satisfaction that the right decision was made. The relationship between value and volume is not linear. This disproves the rationale of the data hoarder.
Unfortunately, this only proves that as humans we are all data hoarders at heart. The more data we collect widens the ‘data divide’ – the communication gap between business and tech. Techies talk one language, those in business another. The principle of ‘Small Data’ seeks to address this. (first learnt of this term in the I4DM Document and Content Management course) This idea recognises this paradox and calls for data managers to present data to decision-makers that is curated, contextual, and actionable. This is increasingly important as technology continues to generate ‘Big Data’.
So…does size matter?
It is not about how much data you have, but how you use it. Data management holds the key. It is the role of CDMPs to bridge this ‘data divide’ by presenting business users with data that is made accessible, with context (metadata) and is considered trustworthy (quality). By focusing on volume rather than quality, the data hoarder has gorged on too much data for too long. Just as someone may eat themselves to death, organisations may also inform themselves to death.
Declan Moore
Junior Consultant
Certified Data Management Professional
