Q&A: Taking a dive into data analytics

With many organisations making data generation a higher priority, data analytics solutions have become increasingly popular as well, as organisations attempt to streamline their IT processes and turn their growing volumes of data into keys for cost-effectiveness and efficiency.

We spoke to Sridhar Iyengar, VP of product management at real-time IT management company ManageEngine, about the data landscape and the role of self-service IT analytics.

  1. What is driving companies to acquire so much data? Which sectors have been at the forefront of this?

There are three factors driving companies to acquire more data. One, companies are using more applications, which are easier and cheaper to access, so they are generating more data. This trend is fed by a) the rise of mobile computing and the lower costs associated with mobile devices, b) the rise of cloud computing, and c) the creation of more applications for all facets of business, reinforcing IT and software as the underlying fabric of all types of business.

Two, storage is cheap. It costs less to store data today compared to a few years or a decade ago. And three, people always knew there were insights in their business data, but they didn't have an easy or inexpensive way to analyse that data for insights. Now, cloud and evolving technologies such as distributed processing and analytics software have made it easy to process this data in an inexpensive way.

The banking, finance, and insurance sectors have been leading the acquisition of more data due to the higher impact of data on revenues and profitability. In healthcare, the analysis of electronic health records for predictive modelling can move healthcare treatments from reactive to preventive mode, thereby resulting in better medical outcomes and lower costs.

The pharmaceutical sector generates vast amounts of data through drug discovery and clinical trials, which have to be analysed efficiently. Improving customer experience is very important in the retail and e-commerce sector. Large amounts of customer data need to be analysed to better understand buying patterns; demographic, regional, and seasonal trends; and pricing decisions.

  1. What sort of barriers are presented by storing this data in silos?

When data sits in silos, teams can't share their data with other teams. In turn, they can't fully understand their business problems because they don't have holistic visibility of those problems, which could include:

  1. How to correlate marketing spend with new revenues earned?
  2. Has the support responsiveness improved or not?
  3. Are customers happier with the product now compared to a year ago?

When businesses ask such questions, the only way to find answers is to look at data holistically across departments such as sales, marketing, and support. Data residing in silos prevents both effective data analysis and optimal business answers.

Another barrier presented by storing data in silos is the difficulty of accessing information and reports when you need them, which slows down decision making. Storing data in silos also prevents you from correlating and cross-referencing data that can be derived as relevant and therefore can provide accurate insights.

  1. How can enterprises, particularly those already heavily invested in gathering data, move away from working in silos in a cost-effective manner?

Enterprises can move away from working in silos in a number of ways. The more time consuming and expensive option is to normalise data using tools that correlate or merge data scattered in silos into a common unified format or data set. The most cost-effective approach is to standardise on one analytics tool that can ingest, correlate, and visually present a unified view of all the data, irrespective of where it resides.

  1. What is the most effective practice in regards to storing, tagging, and categorising data?

Such data classification, identification, and storage techniques depend on 1) the type and confidentiality of the data, 2) the organisation's regulatory and security policies, and 3) the specific situation. However, in general, data classification should be done based on the analysis and reporting requirements of the organisation.

Once those are determined, the metadata for classification should be derived in consultation and agreement with all teams that are involved in producing and consuming these data sets as well as security and governance teams. This work should not be done in isolation by the data architecture team alone as they are dependent on both the data source as well as internal or external analysts that would create visual analytics and they would drive some of these data classification requirements as well.

  1. Why does so much of the discussion surrounding data analytics focus on visualisation?

Graphical visualisation has been the best way for humans to grasp things quickly. Applying the same concept to data analytics, graphical representation of insight from analysis is the easiest form by which to understand the data analysis. Unless the information and insight from the data analysis can be presented in an easy to understand way, it will be very difficult for organisations to glean useful insights and arrive at the right decisions.

  1. In what ways have the companies you’ve worked with become more cost-effective and efficient through data analytics? Which data sources tend to contribute the most valuable data?

We have customers who use our business intelligence and analytics solutions, Zoho Reports and ManageEngine Analytics Plus, to analyse sales, marketing, finance, and IT data to help them derive accurate insights for better decision making.

Using our solutions, our customers are able to spot patterns within their data and make decisions quickly. This has enabled them to become more agile and responsive to situations. Additionally, understanding their data better has also made them more aware and be able to predict seasonal patterns. That is, they understand their business better and can serve customers better and eventually be more profitable.

  1. More companies have begun acquiring data. How have data analytics platforms changed to address the practice becoming more mainstream? Are they being designed with newcomers and smaller companies in mind?

Data analytics is a need for organisations of all size and types. Historically, such tools were expensive and complex to use, putting them out of reach of smaller or mid-sized organisations. In recent years, the transformation of analytics from a consultative-driven solution to a self-service solution has made it possible for even a smaller organisation with no data experts to analyse data by their non-expert data analyst.

That and the rise of cloud-based analytics solutions that offer large data storage and processing required for scalable needs will be major factors driving data analytics to become mainstream.

  1. How important is mobile optimisation for IT professionals using data analytics?

Mobile optimisation is key to presenting data analytics insights on a mobile device. Unless the analytics tool is optimised for mobile, the presentation will be difficult to understand, and users won't be able to understand the insights and make the right decisions.

Image source: Shutterstock/tonefotografia