'Data Mining Aspect Of BI Facilitates Predictive Analysis'

Business Intelligence has become an integral decision making tool for many organisations in the current context. BI has moved on from traditional reporting tools to far more advanced predictive analytical tools. Today, BI is closely associated with enterprise wide performance management activity. Bhavish Sood, principal analyst, Gartner, in an interaction with Biztech2.0, discusses how BI is widely used for predictive analysis and performance management.

What are some of the major driving factors for BI?

The primary driver for a BI solution is that people at each level of the organisation need information to perform their job on a day-to-day basis. Without information or contextual data, one may not be able to efficiently perform his/her job, be it that of a supply chain manager, CIO or CFO. The decision making process has moved away from ‘gut and feel’ to fact-based analysis. To put it in other words, you can say that it has moved away from historical analysis to predictive analysis. Hence, information is extremely significant to make critical decisions, and this information is provided by BI solutions, which makes a strong case for their adoption.

More from Biztech

Future Group - Reliance Retail Deal approved by CCI RBI ban on cryptocurrencies takes effect; prohibition could force investors to tap the black market

Which verticals will lead the adoption of BI tools?

BI as a practice or as a management technique can cater to almost all verticals. However, which verticals will spend more on BI solutions vis-à-vis which others depends upon the amount of competition in a particular industry and the ability to drive customer efficiency and satisfaction based on ‘intelligent information’. For example, in India the BFSI vertical would be more likely to use BI solutions due to its extremely competitive nature. Similarly, in the retail industry, BI is a huge differentiator as it can help to measure customer loyalty, footfalls, shelf space return, store profitability etc. Post the announcement that mobile number portability will be allowed in India, telcos are also expecting a huge churn of customers and will use BI to cope with the new challenges.

BI is becoming increasingly significant for performance management. Could you elaborate on this?

There is a stepping stone to start any activity. On day one itself, you can’t say that I am implementing BI as well as handling performance management. Typically, I have observed that in India BI generally starts as a departmental initiative. So let’s say in case of the retail industry, one can apply BI for a customer loyalty programme to start with. Here BI helps to understand how many customers are members of the loyalty programme, how many members are frequent buyers, what is the average bill size, how much time they devote to each brand etc.

Let me give you one more example, the CFO may require some financial information to understand several aspects like the store return, shelf life of each product or the level of inventory etc. Thus, BI may be required not merely for historical and operational purposes, but also to measure financial and non-financial metrics and that is what performance management is all about.

How does BI lead to predictive analysis?

Reporting is one large component or subset of BI. BI can involve multiple technologies. The historical measurement is very easy to figure out, though ability to predict what is going to happen is of essence. Referring to my earlier example, telcos are worried about the oncoming number portability issue. They would like to know about the propensity or the percentage by which their customers are expected to churn. Thus, there is a historical aspect of BI, which is reporting centric and there is a data mining aspect, which allows predictive analysis and provides insight into what could potentially happen in the future.

How can an enterprise move towards the BI and PM curve?

An organisation should start with planning and preparation of a BICC (Business Intelligence Competency Community proposal). Then you look at how internal customers are currently accessing information, what is the frequency of information they require, how they get that information and what are the various ways by which you can enable them easier access to that information. BI, in conjunction with data warehousing, allows you to build a common set of metrics and also helps to bring consistency in enterprise wide data. Thus, BI is now moving away from being a departmental initiative to having an enterprise wide view and is helping organisations make fact-based decisions.

What are some of the critical things that CIOs need to keep in mind while developing a BI strategy?

BI is not merely a technology platform. While considering a new BI initiative, CIOs generally focus on what kind of reporting tool they need to buy. They need to move away from this practice and focus on key metrics at the business strategy level and also the prime objectives of having BI.

There are several things, which they need to keep in mind while deploying BI solutions, but primarily they need to link it with business strategy. The best way to do this is to study their annual report and look at the prime challenges of the company and the sector as a whole. They can also talk to CXOs and organisation leaders to understand their information needs.

The other important thing is that you cannot have one set of tools and user access designed for every employee. CIO and CXO level people will want to look at scorecards and KPIs to measure the performance of the business. People, who take fact-based decisions, will need ad-hoc analytical or predictive model kind of capabilities. People, who want to migrate, can be provided with BI in interactive, visual formats or familiar formats like MS office, where they can easily use BI for consumption.