4 Criteria of Actionable Insight in Big Data

This is the final instalment of a three-part series on data-driven productivity, and why so many organizations fail at getting true value from their analytics efforts. We began by clearing up some buzz words and defining an important hierarchy between “data”, “information”, and “insight”. We explored the friction between going from passive data to actionable data, including shortcomings of traditional tools, errors of old thinking, and the need for a data-driven culture shift.

Today, we explore the four key criteria of actionable insight. It’s important for organizations to constantly monitor how their data is parsed and communicated in order to clearly differentiate noise from information-driven value. How data is captured, contextualized, organized, and delivered are all critical variables of a successful big data strategy.


To ensure that your organization’s data initiatives are on the right track, ask whether your current framework considers the following characteristics. If your analytics is falling short of data-driven decision making, it may be time to rethink your team’s approach, tools, or process. 

When it comes to analytics of any kind, placing it within the proper context is key. You may feel that you’ve come across a new insight but if it lacks the appropriate background to put it into context, it can raise more questions than it can drive actions—or worse, it can drive the wrong actions. 

For example, say a manufacturing organization has been meticulously tracking its sourcing in order to cut spending and optimize savings. In doing so, the company realizes that it can pay $1 for a component which they have historically purchased for $2. That is a 50% price cut which, from an in-house perspective, is huge. 

However, let’s say the same organization analyzed and benchmarked their data against the rest of the world. The reality could turn out that other manufacturers are paying an average of $0.50 for the same component, making your new $1 pricing twice the competition’s cost. Working with your in-house data in an isolated silo will only provide limited intelligence. Deriving the right actionable insight requires accompanying context and a wider sampling of data that offers a more complete understanding of the information.


While data relies on logic and reasoning, decisions are still ultimately made by humans. This means that a level of bias and subjectivity comes into play when determining whether a data-driven discovery is notable or just noise.

Relevance requires the insight to be delivered to the right person at the right time in the right framework. If the finding is not routed to the appropriate decision-maker in time, chances are it will be too late to act or even investigate. The value of data is as good as its ability to be acted on.

If an insight does not adequately help to explain why something has occurred, it is not being specific enough. It’s important to aim to produce very detailed and complete insights, rather than going by high-level metrics that lack the context or precision to answer questions.

Descriptive analytics has a way of fooling professionals that they are sufficiently “working with data” because it gives them a good grasp of what has happened. If the data does not explain why it has happened, not enough is being understood about its triggers and conditions. For manufacturing and sourcing especially, moving from descriptive to predictive analytics will be a huge jump in both data capability and competitive advantage. Gaining the ability to know what may happen first requires a full-picture understanding of why. 


How the insight is communicated plays a huge role in whether or not it is actionable. By design, data needs to be filtered, processed, and organized for it to become human-friendly information. If this step is done poorly, it causes a major break in the path towards actionable business insight.

Organizations that are struggling to connect data to action should pay close attention to whether or not their data is being presented in a way that supports communicating its values. This doesn’t just stop at pretty graphs and messaging: it needs to live in a user-friendly tool that encourages staff of all levels to draw from and collaborate with it. 

Which of these has your organization achieved? Which is the most challenging? Let us know in the comments section below. 

0 comments on “4 Criteria of Actionable Insight in Big Data

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.