DATA APPROACH

Broadly there are 3 spectrums to data led decision making and each of these have a pertinent role in the decision making process. An important takeaway is to consider the problems that can be solved by collecting the appropriate data. 

Each individual set of data can be considered as variables and using some basics of univariate and multivariate processes you build an understanding of these dependent and independent variables. Such modelling helps determine the market share or customer heterogeneity or segmentation benefits, there are more than many outcomes possible  

If you wish to know more then click on the link below https://www.dhirubhai.net/school/wharton-online/

Historical Data Analysis 

Predicting trends based on already available information about the consumer or product. Essential tool for such analysis is regression. You need to identify the variables that need to be measured, basis dependent and independent variables and leave the rest to the regression. You can use excel or other set of software to drive this analysis. There are various metrics that gets highlighted in the post analysis example R-square, P Value etc.. and depending on your need you can choose the most relevant one 

This type of analysis is very useful  to calculate various types of business problems, for example impact of promotion and advertising on sales or price vs demand efficiency or optimal pricing techniques. Further it also helps you in identifying time trends and mktg mix 

Predictive Data Analysis 

What happens when you don’t have historical data to solve any business problem, this is where predictive data analysis becomes very useful. It essentially helps in making decision about and around new products. 

Predictive data analysis is a great tool for managerial usage 

• New product development
• Price elasticity to demand
• Market segmentation  

Conjoint analysis is a great tool to drive predictive data analysis.  Helps build business perspective and decision making process by keeping the consumer on the forefront. For example consumer preferences around your product, its various attributes, the worth of each attribute(part worth), identifying utilities and so on and so forth.

Social Media Analytics 

This is probably the largest pool of real-time data available on consumers. You can build clusters and drive clutter analysis. Use of psychological, sociological and anthropological factors combined with demographic information to identify market segments with a propensity to favor soe products/characteristics over another 

Major Variables to play with 

• Demographics
• Geographic
• Purchasing approaches
• Personal characteristics 

Forecasting demand & sentiment - UGC in itself is a huge repository of various types of information and sentiment. NLP methods are used to derive such analysis 

Build approaches to convert qualitative data to quantitative data from different kinds of blogs and reviews. For example co-occurence of a brand on a specific blog signifies that in the consumer mind these brands are compared together.  This helps in building consideration set of a customer 

• These are some basics that I learned, please feel free to add more to this and enrich the audience