Advertisment

Less ROI, More Lawsuits – Why Big Data goes awry?

Look beyond IT perspective; smell for business perspective, ethics, and governance too

author-image
Pratima Harigunani
New Update
Image courtesy of TAW at freedigitalphotos

Arvind Purushothaman

Advertisment

While large businesses and IT organizations are convinced of the potential of Analytics within their industry, often they are uncertain on how to embark successfully on a Big Data Analytics journey for the enterprise. Big Data isn`t all about technology but it’s about how one strategizes the use of data.

Big Data solutions are most successful when approached from a business perspective and not solely from an IT perspective. The use of Big Data and Advanced Analytics techniques typically requires combining multiple data sets from diverse sources. If not thought through, the output garnered from these techniques can have unintended consequences for both individuals and organizations. These include lack of return on investments, incorrect business decisions, loss in customer confidence, and can often lead to potential lawsuits for organizations.

Here are a few things organizations must consider while planning Big Data & Advanced Analytics initiatives -

Advertisment

1. Well-defined outcomes - The expected outcomes must be well thought through. The desired ROI must be well defined, and should be constantly monitored for results.Organizations must create a feedback loop to ensure that it’s not an academic exercise for them. It should be kept in mind that for achieving the results, all potential barriers including process and cultural barriers must be eliminated.

Arvind P, VirtusaPolaris Arvind P, VirtusaPolaris

2. Over-engineering the solution–Organizations must be cautious about over-engineering the analytics initiatives. Essentially, it means that companies can use advanced analytics techniques only when they value over and above other forms of rule-based analytics. If the rule-based analytics are not yielding the required results, it may be beneficial to try other advanced analytics techniques and compare them with more traditional results.

Advertisment

3.Quality of data - The success of advanced analytics is dependent on the availability of good quality data. Before embarking on any analytics initiatives, it is essential to profile the quality of data especially, for the key variables that will be used in the model. It is critical to put together diverse data sets by understanding the relationships which can help in building a conceptual data model.

The quality of output is highly dependent on the completeness of the data, and its character. For example, if health-care decisions for individuals are based on advanced algorithms and empirical data, it is important to ensure that the algorithm has access to clean and complete data sets. In case of Machine Learning techniques, it is highly crucial to train the ‘machine’ on a realistic set of data that is available in the organization.

4.Ethical use of data - The ethical use of data is critical in today’s time. It ensures that organizations are not viewed in a negative manner by customers or subject to lawsuits. Data privacy laws are still evolving; hence it is important to assure that organizations are self-regulating. This can happen at multiple levels starting with collecting the data from individuals.

Advertisment

Full disclosure on how data can be used is essential. For example, using data to predict potential defaulters of loans, to reduce risk exposure is a valid business case. However, the use of similar data to profile and potentially deny services based on age, ethnicity and other discriminatory parameters would be an unethical use of Big Data Analytics which could lead to lawsuits.

What is more important, is that organizations should remain sensitive to personal preferences of individuals and not expose it to a wider audience through targeted marketing. While organizations have access to individuals’ personal data, it is imperative that they ensure it is not hacked or leaked.

5.Control over the use of data - Governance over Big Data, advanced analytics and its usage is becoming critical day by day. Organizations must have a Data Governance function that governs the use of data as a whole including Big Data &Advanced Analytics. One of the main questions which needs to be answered is who owns the data. Are data sets used in algorithms complete and clean? What is the intended use of advanced analytics? Any use case that involves using sensitive information must be well thought through from all angles, and steps must be taken to ensure its accuracy.

The five points highlighted above are critical in securing success in this happening field of Big Data Analytics. If followed correctly, organizations can reap tremendous benefits and this can be a source of competitive advantage.

(Arvind Purushothaman is VP –Data & Analytics, VirtusaPolaris. Views expressed here are of the author and CyberMedia does not necessarily endorse them.)

big-data analytics