Dear Vandana
The post from RNair2000 is an excellent one and gives you a good overview.
Here are a few pointers on the topic from my end:
1. Analytics is not for everyone! Analytics is a process of correlating and evaluating data... And therefore, if you are in a company that doesn't have a reasonable quality of data, then it might give you a mucked-up scenario. For instance, if you have partial data... and decide to refine it later, your analytics tool will have a nightmare!!! We were evaluating training needs of drivers and the accidents said that the buses had 'hit' another vehicle... Now, there was no record of whether it hit the front or back or side! These details affect the training requirements and technological upgrades considerably.
2. If your decisions aren't data driven, you are definitely not for analytics. I recall in one of my pharma clients... The client wanted to bring in a 'standard process', but failed to actually follow that... In fact, he went ad hoc, despite spending a fortune on his performance management system. For such people, analytics doesn't really help...
3. If you don't know where to use the data (i.e., what indicators are meaningful), you are certainly not a candidate for analytics. Most times, the managers are not aware of where and how to use a report. They spend hours in 'defining' the reports and understanding the kinds of indicators. However, to put that in a smooth-flowing process requires a deeper understanding of the analytics process. Most times this is given a lower priority in the organization. The result is that the legacy process prevails as it is within the 'comfort' levels of the company.
4. If you don't have a proper roadmap of how to 'graduate' from one level to another, you are definitely going overboard with the whole concept. Effective analytics requires a stringent control over all the elements viz. data, updation frequency, updation quality, representativeness checks, etc. And each metric needs a 'minimum' amount of data / substrate information to give meaningful results. If you don't have a proper roadmap of when to use what data, how, and why... you are probably going to spend a lot of money in customizing your analytics tool.
Hope these help...
In case of questions, please let me know. Will be glad to answer.