Customer renewal rates are dropping, and your CEO is on the warpath. You need to find out why and fast.

At most large companies, that is a pretty tall task. Information about customers is likely scattered across an assortment of applications and devices ranging from your customer relationship management system to logs from customer-facing applications, order entry forms, notes on sales rep laptops, and even social media posts.

Each of these sources can yield some insight into the cause of customer defections. None can give you a definitive answer. For that, you need a modern observability solution which doesn’t silo data, but brings it all together in one place. By embracing this, companies can examine a broad range of customers, determine patterns that are common to non-renewers, and test changes that might fix the problem.

The Danger of Data “Nuggets”

Data fragmentation is one of the biggest factors that frustrates decision-making at large organizations. With operational data spread across hundreds of thousands of devices, leaders don’t know what information they have, its condition, or whether it can be believed. Multiple versions of the same data exist in different locations, and duplication is rampant. Data owners hoard information in the belief that it enhances job security.

Isolated bits of information exist without context or integration. While each is useful individually, they fail to provide a comprehensive view of the bigger picture. Without that broad perspective, interpreting these “data nuggets” can lead to dangerously misleading conclusions.

The situation is an unfortunate byproduct of the changes in data ownership expectations that kicked off more than 40 years ago when PCs entered the workplace. Prior to that, institutional knowledge was locked up in employees’ brains or corporate mainframes accessible only to a privileged few. Many efforts were made over the years to harmonize data silos, ranging from file shares to Sharepoint to Slack. They have helped, but integration challenges between these disparate platforms have only reduced data islands, not eliminated them.

Aggregate and Conquer

To harness data’s full potential, you need data management strategies emphasizing aggregation and insight extraction. Value comes from analyzing comprehensive datasets, identifying patterns, and drawing meaningful conclusions.

Technology now exists to support large, centralized data repositories where data can be stored, normalized and accessed without a computer science degree. Data lakes and lakehouses are reliable, scalable, and relatively inexpensive platforms for aggregating information. Integration tools make data accessible without the need for tedious manual cleansing. Breakthrough technologies like object storage and Apache Iceberg allow multiple people to operate on the same data without compromising integrity or consistency and obviate the need for a rigid IT structure.

Organizations that take advantage of these technologies will leap ahead of their competition. By combining disparate datasets, they can uncover correlations and trends that would be invisible in isolated data nuggets. Metadata provides context and background information that enhances the interpretability of the data. Advanced analytics tools, both stand-alone and in combination with other functions like observability and visualization, extract insights from aggregated data.

AI Multiplier

Artificial intelligence takes these capabilities to a new level. Machine learning and data mining can identify patterns and predict trends. Users can ask questions in natural language instead of SQL. Machines can discover trends in volumes of data too large for humans to process. We are still in the early days of understanding how generative AI will change how we interact with and understand data.

Taking advantage of consolidated data is more than a matter of adopting tools. Strong data governance policies are required to ensure data quality, consistency, and security. Many governance frameworks are available to help maintain data integrity and prevent redundancy and inconsistency issues.

While data nuggets are a convenient way to collect and store information, they fall short of unlocking data’s true potential. Organizations can transform fragmented data into a powerful asset by embracing data management strategies emphasizing aggregation and insight extraction. The shift from piecemeal data to holistic data management is crucial for gaining meaningful insights, enhancing decision-making, and driving innovation in the AI age.