Customer relationship management (CRM) is critical activity of improvising customer interactions while at the same time making the interactions more amicable through individualization. Data mining utilizes various data analysis and modeling methods to detect specific patterns and relationships in data. This helps in understanding what a customer wants and forecasting what they will do.
Using Data mining you can find out right prospects and offer them right products. This results in improved revenue because you can respond to each customer in best way using fewer resources.
Basic process of CRM data mining includes:
1. Define business objective
2. Construct marketing database
3. Analyze data
4. Visualize a model
5. Explore model
6. Set up model & start monitoring
Let me explain above steps in detail.
Define the business objective:
Every CRM process has one or more business objective for which you need to construct the suitable model. This model varies depending on your specific goal. The more precise your statement for defining the problem is the more successful is your CRM project.
Construct a marketing database:
This step involves creation of constructive marketing database since your operational data often don't contain the information in the form you want it. The first step in building your database is to clean it up so that you can construct clean models with accurate data.
The data you need may be scattered across different databases such as the client database, operational database and sales databases. This means you have to integrate the data into a single marketing database. Inaccurately reconciled data is a major source of quality issues.
Analyze the data:
Prior to building a correct predictive model, you must analyze your data. Collect a variety of numerical summaries (such as averages, standard deviations and so forth). You may want to generate a cross-section of multi-dimensional data such as pivot tables.
Graphing and visualization tools are a vital aid in data analysis. Data visualization most often provides better insight that leads to innovative ideas and success.
Source: http://ezinearticles.com/?How-Data-Mining-Can-Help-in-Customer-Relationship-Management-Or-CRM?&id=4572272
Using Data mining you can find out right prospects and offer them right products. This results in improved revenue because you can respond to each customer in best way using fewer resources.
Basic process of CRM data mining includes:
1. Define business objective
2. Construct marketing database
3. Analyze data
4. Visualize a model
5. Explore model
6. Set up model & start monitoring
Let me explain above steps in detail.
Define the business objective:
Every CRM process has one or more business objective for which you need to construct the suitable model. This model varies depending on your specific goal. The more precise your statement for defining the problem is the more successful is your CRM project.
Construct a marketing database:
This step involves creation of constructive marketing database since your operational data often don't contain the information in the form you want it. The first step in building your database is to clean it up so that you can construct clean models with accurate data.
The data you need may be scattered across different databases such as the client database, operational database and sales databases. This means you have to integrate the data into a single marketing database. Inaccurately reconciled data is a major source of quality issues.
Analyze the data:
Prior to building a correct predictive model, you must analyze your data. Collect a variety of numerical summaries (such as averages, standard deviations and so forth). You may want to generate a cross-section of multi-dimensional data such as pivot tables.
Graphing and visualization tools are a vital aid in data analysis. Data visualization most often provides better insight that leads to innovative ideas and success.
Source: http://ezinearticles.com/?How-Data-Mining-Can-Help-in-Customer-Relationship-Management-Or-CRM?&id=4572272
No comments:
Post a Comment