sppn.info Business Database Marketing Pdf


Saturday, June 1, 2019

PDF | To increase effectiveness in their marketing and Customer Relationship Manager activities, many organizations are adopting strategies of Database. 3 DATABASE MARKETING – A LONG TERM. STRATEGY FOR PRACTICE SUCCESS. The most important thing to understand about database marketing is that. What is with all of the buzz about database marketing? Why are people spending so much time and so much money on downloading this software technology?.

Database Marketing Pdf

Language:English, Spanish, Hindi
Country:Papua New Guinea
Published (Last):
ePub File Size: MB
PDF File Size: MB
Distribution:Free* [*Regsitration Required]
Uploaded by: SEASON

This module will review the key strategies and principles of database marketing. As a result, you will develop a database strategy to achieve both short- and. Database marketing is at the crossroads of technology, business strategy, and Database Marketing Tools: The Basics. Front Matter. Pages PDF. database marketing is very developed and it is following the global trends. Database marketing serves for better understanding of the customers (potential.

The simple steps for predictive modeling: 1. Do a promotion, or use a previous promotion as your base. You will need enough customers in your model.

The rule of thumb is you need about conversions sales 2. Append demographic and behavioral data to your responders and non-responders 3. Add geographic data, besides this appended data, you should add previous download history with your company to your mix of data grouped with every customer record that you plan to use in your model.

Divide your data into two parts A test group of 12, non-responders and responders and a validation group of 12, non-responders and responders. Both groups should have exactly the same type and variety of people. Discard the outliers.

These are customers whose downloads were so unusual that they will distort the outcome 6.

Construct your model. Use a multiple regression model.

Log in to Wiley Online Library

A regression is an equation that describes the relationship between a dependent variable and more than one independent variable.

The dependent variable is the download that the customer made as a result of your promotion. The independent variables are the behavior and appended data listed above. When the model is run, it applies weights or levels of importance to each of the independent variables. A variable importance is a number that indicates how important weighty is each variable in predicting the desired result they bought the product.

For an example of the outcome see Figure on page Determine the weights for each variable Concentrate on using in the model those independent variables that have a high weight in determining the outcome.

Build your model based on the five or six variables that provide the greatest predictive power. In the case of a statistical model in marketing, the algorithm usually includes the computer code that creates a score for each customer or prospect record.

The scores may vary from 95 percent certain to download the product down to 5 percent. Score the validation group. If the algorithm developed for the test group is going to be useful in predicting, it should correctly identify most of the people in the validation group who bought with a high score and those who did not download with a low score. If the algorithm does correctly score the validation group, then you have a successful model, which can be used to predict customer response in your next promotion.

What if it does not work? You either have to redo your model, or give up the whole process as a bad job.

It may be that with the data you have available, a model cannot pick predict the responders. This is very often the case. A general rule of thumb: if the solution does not seem to make sense to you, then it probably does not make sense. Modeling does not always work. Modeling is not magic. It is only a quantification of intuitive logic.

Who will download? After scoring the file of prospects you typically divide the scored file into deciles.

The top decile 10 percent of the file contains those people most likely to download. You probably should not mail deciles You should, however, mail a few 5 percent of each of these low performing deciles just to prove to yourself, and to your management, that the model is still working properly and actually does predict the downloaders correctly.

The differences between CRM and database marketing

You can download compiled names from list brokers based on the model, in that way you only pay for records that have a high downloading score. In a model, behavioral data typically is more powerful than demographic data.

In other words, you can predict better what people are going to do in the future base on what they have already done in the past than by basing your model on who they are age, income, home value.

The power of prior behavior is one reason why it is important in building a marketing database that you keep n your database as much of the customer transaction and promotion history as possible.

Banks have a lot of data about their customers. They can use their database to examine customers who have only checking accounts. From the data they can determine whether our next product should be a home equity loan, an auto loan, a savings account, or mutual funds.

How do the do this? By using a model to see what thousands of their depositors have downloadd mutual funds look like in terms of behavior and demographics, and how the differs form those depositors who have been offered mutual funds but have not bought them.

The model is used to score all their depositors. Those depositors whose scores resemble the mutual fund downloaders are likely targets for a mutual fund promotion. This is an essential read for those interested in database marketing, customer relationship management and customer optimization. This book will become an invaluable reference or text for anyone practicing, researching, teaching or studying the subject. Malthouse, Theodore R. Skip to main content Skip to table of contents. Advertisement Hide.

Database Marketing Analyzing and Managing Customers. Front Matter Pages i-xxiv. Front Matter Pages Robert C. Pages Why Database Marketing? Organizing for Database Marketing. Customer Privacy and Database Marketing.

Customer Lifetime Value: Issues in Computing Customer Lifetime Value. Customer Lifetime Value Applications. Sources of Data. Test Design and Analysis. The Predictive Modeling Process. Statistical Issues in Predictive Modeling. RFM Analysis. Market Basket Analysis. Collaborative Filtering.

Discrete Dependent Variables and Duration Models. Cluster Analysis.You try everything, and see what works best.

Database Marketing

You either have to redo your model, or give up the whole process as a bad job. Challenges and limitations[ edit ] While real-time business intelligence is a reality for select companies, it remains elusive to many as it is dependent on these premises: the percentage of the business that is online, and the degree of level of sophistication of the software. New York , Amacon, DMA, 5 2: Consumer data[ edit ] In to existing customers, more sophisticated marketers often build broad databases of customer information.

DALY, Virginia.