Bad data leads to bad decisions, yet data quality problems run rampant in today’s organizations. Surveys show that as many as 60 percent of companies consider their data to be unreliable, with as much as 25 percent of the information in the average database containing inaccuracies.
It’s these numbers that impact the effectiveness and trustworthiness of the analytics dashboards you’re creating.
A common mistake we often see is organizations trying to “work around” the data quality problem, and the problem with this is two-fold.
- This adds unnecessary cycles to workflows, and time costs money. A recent survey cited 31 percent of analysts and data scientists claim to devote up to three hours a day cleaning data.
- You will run the risk of introducing a trust issue. Once your users think that the data they are making decisions on isn’t right, they’ll lose trust in the environment and look for other solutions.
No matter how sophisticated your visualizations and analytics are, you’ll experience major problems if the underlying data is questionable.
A successful BI and analytics strategy shouldn’t focus on the pretty picture. It should start with the data. First, connect your analytics to the data you need either in place or by moving it. Consider the latency demands on that data. Do you need real-time access or is near real-time good enough? Then, consider incorporating data quality management and master data management into your overall BI and analytics process. This will help you identify and correct bad information before it reaches your end users. Capabilities like profiling, cleansing, matching and merging, and a single view of data across all sources can prepare and optimize data for analysis by ensuring its accuracy, completeness, timeliness, and consistency.
Confidence in your data is vital to promoting widespread adoption among your users, which will drive greater value. No matter how sophisticated your visualizations and analytics are, you’ll experience major problems if the underlying data is questionable.
Our customer, St. Luke’s Healthcare Network, successfully cultivated 31 diverse data sources to create an enterprise health data warehouse with a single view of key business entities. By mastering all patients, facilities and providers, they eliminated errors and duplications and created consistent reference sets for all of their BI and analytical applications. Stakeholders now have access to a single source of trusted information that is actionable and drives constant improvement. Simply put, trusted data has made St. Luke’s a better organization.