The Case for Analytics & Data Science

Posted on Posted in Analytics and Strategy

It’s a confusing world. People are talking about Analytics and Data Science in all walks of work and every industry. Where do we get started ? Lets take a look …

1. Data Models :- Organise all the data from different systems in your organisation together . Standardise this process and have one repository for all the data you commonly work on. This is the first step to do Analytics on an ongoing basis !!

2. MIS / Data visualisation:-
 Tables, Charts, Trend report
 Drag and drop cubes and pivots
 Conclude on the reports and data you have – Insights will help you take decisions
This is a cool way to dissect and look at your data!

3. Non-Predictive Analytics :-
 Segment and cluster your customers and vendors – which of them are more profitable?
 Thru the door analysis – what is the sourcing quality of your sales and marketing team?
This often answers questions of Leading and Lagging indicators – what influences what else? And how much?

4. Predictive Analytics :-
 Probability models – what is the chance that a customer or employee will attrite?
 Forecasting models / Time Series – Check out the probable revenue and cost numbers for the next few months.

Level II view of the above process:-
Analytics will be used for
1. Understanding data trends through reporting and data visualisation
2. Analysing the key drivers of business results (Profit, Loss, Sales, Customer Satisfaction, and Retention etc.)
3. Improving Operational Effectiveness in terms of process improvement
4. Using Probability of occurrence results and forecasting to decide on future strategies

Methodologies for an organisation to incorporate Analytics :-
1. Create an in-house team from a Top to Bottom approach: – Recruit the Analytics / BIU head and then he recruits the team and inducts the software.
 Pros: – In-house team from scratch, Data sharing with external agency not required
 Cons: – The BIU head’s cost starts from day 1- changing permanent employee will not be easy in-case of discontent with performance. Initial data exploration and reporting will need a reportee. In-house software requirement from day 1
2. Create an in-house team with an External Consultant helping: – External consultant does the initial exploration and work and recruits BIU personnel at required levels as the complexity of work and data availability increase. Initial software requirements can be met by the consulting team/person.

 Pros: – Initial data exploration and reporting can be done by consultant/team, software requirement from day 1 can be handled by consultant – once a desirable output is seen- purchase of software decision can be made, Team can be grown from the middle down and middle up. Ineffective consultancy can be penalised.
 Cons: – External data sharing will happen.

We can say that at times analytics is a long and meandering journey. It is important for us to understand that this journey will require a lot of patience and sustained efforts. This is also a never ending journey .. and the next step can be machine learning.
Machine learning is a new discipline but it will be very important in the future.

Do write in with your thoughts to

Leave a Reply

Your email address will not be published. Required fields are marked *