Palekar believes that among the multiple BI solutions available in the market, some are by established players in the country and some are by relatively newer ones. The skill sets available for both these kinds of solutions vary according to the solution itself.
"Few solutions, though technologically superior, pose challenge from a resource availability point of view. Hence, this becomes a critical factor when it comes to selecting a BI solution," he says.
Sangoi mentions that before internal users, be it business or IT, are put on the job, they need to understand that a BI project is very different from a traditional IT project. In an IT project, the organization first defines the outcome, then does requirement study, as-is and to-be process definitions, and implements the same with technology-based solutions after training the stakeholders.
"In analytics projects, the hypotheses are built based on the theories developed, tested through experiments, refinements are based on the findings, and the whole process is repeated for continuous improvements. Hence, a CIO needs to prepare his team for a change of mindset in managing the projects," he explains.
Another challenge in choosing the skill sets required for BI is ensuring complete synchronization between the teams handling the technical aspect and the business analyst. Also, it works a lot better if both the technical expert and the business analyst have some knowledge about both aspects of BI i.e. the technical and business acumen.
"Analytics by its very definition needs someone who understands the business and is a good business analyst. Apart from that, he/she also needs to understand analytics tools and know how to leverage them for successful implementation," says Sangoi.
What are CIOs Looking for?
Agrawal says that there should be awareness of what is important to the business, and what the various perspectives that should be considered in the design are. Such factors would be important in the design and success of the solution.
"On the analytics projects, more than pure IT capabilities, a CIO should look for people who are at ease working with large set of numbers and a good statistical background," Sangoi explains.
Sangoi says that Statistical knowledge here could range from knowledge of concepts such as measures of central tendency, estimation, probability, outliers, normal distribution and others to confidence intervals, hypothesis testing, regression, correlation, data visualization, time series, segmentation, forecasting and expert system, fuzzy logic, and artificial neural networks among others.
Sangoi suggests that analytics-based teams should typically be cross-functional. "People with robust understanding of business, and computer science; statistical graduates; and data scientists should typically form part of such a team," he explains.
Since the ultimate goal of such projects is to take decisions based on data and evidence for overall business and consumer benefit, it can be very useful if the team also consists of members who have necessary understanding of cognitive and behavioral science.
Sign up for Computerworld eNewsletters.