Every day, in every organization, senior leadership teams rely on data to make key decisions about hiring, project assignment, vendor selection, employee promotions and more. But research continues to show that the data being presented to leadership is inaccurate, leading to tactical errors, gross inefficiencies, and other negative consequences.
Yet now, more than ever, data accuracy is vital. As John Hall, a partner with KPMG, put it, “Complex analytics underpins most of today’s most important business issues, so it’s vital that business leaders can trust in the data, in the analytics, the algorithms and the decisions that they make by that data.” The more your company relies on data to make decisions, the more imperative it is that you have good data.
How big a problem is bad data?
You may already know, based on anecdotal evidence, that the data you’re receiving is flawed—or perhaps, like many leaders, you’re unaware of how bad data is affecting your organization. Either way, good data proves that the problem is worse than you think:
- Around 56 percent of CEOs surveyed in a KPMG study expressed concern about the integrity of the data they were using for decision-making.
- 36 percent of CEOs feel that they are unable to make data-driven decisions until they have invested significantly in data quality.
- S. companies responding to a Nielsen survey reported that 32 percent of their data is inaccurate, a seven percent jump from the same survey a year ago.
- Of 75 executives taking part in a study published by Harvard Business Review, only three percent reported that their departments fell within an acceptable range of accuracy for data records.
- On average, 47 percent of data records that are newly created has at least one critical error.
The Costs of Inaccurate Data
The costs associated with these critical errors are enormous. Data Driven recommends calculating the cost using the “rule of ten”: “It costs ten times as much to complete a unit of work when the data is flawed in any way, as it does when it is perfect.” And those costs only continue to rise with our growing dependence on data. A study by IBM indicates that bad data costs U.S. companies $3.1 trillion per year.
The intangible effects can be far more pernicious. A full 95 percent of companies want to spin their data into insight. If that data is faulty, it may prevent them from generating useful insight, and even guide them in the wrong direction. Bad data also weakens decision making, as the quality of decisions is directly correlated with the quality of the data. And, obviously, incorrect data also makes it difficult to successfully execute any data strategy.
The Causes of Bad Data
The single primary factor behind most bad data is human error. In fact, 61 percent of companies report human error as the driver behind their inaccurate data. The reason is obvious: The majority of data is collected manually, which makes it vulnerable to error or tampering.
There are myriad secondary sources of inaccurate data, including:
- Data migration and conversion projects, in which data is lost or mishandled by employees
- The prevalence of siloed departmental data strategies
- System and program errors
- Lack of a data management strategy developed by company leadership
The Solution to Bad Data
With such complex, diverse causes behind bad data, there are no shortcuts to solving the problem. The first step is for leadership to become aware of the outsized role human error contributes. Then, leadership should develop a sophisticated, optimized approach to improve data quality, ensuring that all levels of the organization can provide them with real, accurate data. They can accomplish this by making sure that:
- The data is automatically captured and cannot be corrupted by human error.
- The data captured provides a comprehensive view, along with the capability to view either part or all of the data.
- The data captured can be integrated across multiple enterprise applications, to provide a complete overview as well as checks and balances.
There are tools available to help leadership teams efficiently automate data collection while fulfilling all of these requirements above, and all but eliminating the possibility of human error. Leaders who take advantage of these tools can improve data accuracy, reduce costs, and make decisions more confidently. Finally, after implementing a data management solution, leaders should move beyond trust and verify that all data provided is of sufficient quality to make well-informed decisions and successfully guide the organization’s strategy.