Linked In Ten steps to effectively use Big Data - Sedgwick

Ten steps to effectively use Big Data


Big Data is a new catch phrase, not a new concept. It's about making sure you're utilizing all of the information you have available in order to make meaningful decisions about your business, your program…whatever you're driving toward, and to link analysis with action.

Leveraging the techniques found in data mining, the framework for Big Data takes two paths for risk managers and claims administrators:

  • The first is data depth. Data depth describes the volume of data available within a content vertical, such as workers' compensation or general liability-bodily injury claims.
  • The second path is data breadth. Data breadth provides data across content verticals, such as integrating workers' compensation with human resources, company exposure, disability cases or group health data to name a few examples.

The path taken depends on the analytical question or business need driving the analysis. Depth or breadth alone may be enough or may provide such a partial picture that the endeavor is worthless.

The key is to not to let the data be the starting point. Think in terms of the following ten steps as you consider the use of Big Data:

Step 1 – Definition: First, define the business question.

Step 2 – Scope: Determine the scope of the question – what would be actionable? Remember that the answer does not have to explain everything, but meaningfully move the bar forward.

Step 3 – Identification: Identify the necessary data and its location.

Step 4 – Data preparation: The biggest hurdle to Big Data, data mining, or any form of analysis, data preparation is a topic unto itself but will likely include: data integration, derivation of variables, cleansing and formatting.

Step 5 – Analysis: Yes, you have to wait until step 5.

Step 6 – Validation: Accuracy and precision of your results…what are the strengths and weaknesses of the analysis?

Step 7 – Interpretation: What is the analysis telling you in terms of opportunity? How is the data transformed into information that is actionable?

Step 8 – Deployment: Outside of data preparation, deployment is the single biggest hurdle. How do you get the information from point A to point B?

Step 9 – Action: How can you leverage the information to change results?

Step 10 – Evaluation: Did the action taken actually change the result? Assess with an eye toward the next opportunity.

To learn more about how Sedgwick is using these principles to better manage clients' risks, as well as some ideas about future trends in Big Data, listen to our podcast. Then tell us your thoughts – how are you using Big Data in your own organization?

Keith Higdon, SVP, Decision Support Services

Back to Blog
Back to top