Data Detox: leveraging your greatest asset

Data is among the industry’s most powerful assets with a perceived value of 37% to 40%. Understanding its significant value, is it not time to leverage this asset effectively?

In the world of insurance, quality data is vital. In addition, accurate and reliable data is required to make quality business decisions, however in contrast; unreliable data can have disastrous consequences such as missed opportunities and damage to one’s reputation.

Quality data is now a necessity, one that direct underwriters understand very well. By leveraging this asset, they have successfully penetrated an industry that was wholly broker based.

In order to remain competitive, brokers and agents will need to change their mindset, their processes, and their culture when it comes to data.

Why detox?

The insurance industry is being increasingly driven by data. However, before data needs to be used, it needs to be accurate, complete and consistent. Data not meeting these quality standards will need to be cleaned.

Data cleaning, cleansing or scrubbing is a process used to identify inaccurate, incomplete or invalid data to improve quality through correction.

Data quality problems exist for a number of reasons, but the two most common reasons are data availability and incorrect capturing of information by a user. Completing and correcting data can be a time consuming and tedious process, but it cannot be ignored.

Other benefits of clean data include an increase in retention ratios through enhanced customer relationships, growth through targeted campaigns, single view of the customer and compliance with industry regulations.

The detox process

An iterative process that has been widely adopted for data cleansing consists of the following steps:

- Data analysis or data auditing;
- Workflow specification and execution; and
- Post processing and controlling.

Data analysis or auditing entails the use of database methods, statistical methods, or both, to identify anomalies and their location within data. This may be done manually or completed through a commercial software package such as Microsoft Access.

After the data has been analysed, the workflow - steps that one will perform to achieve high quality data - needs to be specified. Again, these steps can be performed manually, if the data set is small or through a software package for larger data sets.

However, brokers will need to consider the time, effort and cost involved when specifying the workflow. Thereafter, the workflow must be executed until all data sets have been reviewed and corrected.

Upon completion of the workflow, the results have to be reviewed for accuracy and completeness. If the cleansing process was completed automatically, a manual intervention will be required to correct data that could not be rectified.

This process should yield noticeable results in comparison to the quality standards of accuracy, completeness, consistency and uniformity. The process will need to start over for the data that has not met the required quality standards.

Making it a habit

Prevention is better than cure and the same rings true for quality data. After completing the data cleansing process, the broker will need to ensure that the data quality is maintained.

This can be achieved through implementation of system controls such as data validation i.e. rejecting incorrect data before it gets processed by the system. In addition, there are tools available that will be able to cleanse the data as it is captured thus avoiding the need for a full data cleansing exercise.

Preventing incorrect data and maintaining quality data goes beyond the tools and processes in place. It requires a major cultural shift in the thinking and working habits of all those involved in the data value chain.

Ultimately, with good quality data comes opportunities which, if leveraged correctly, can add significant value to both the business and the customer.