Transcript Lecture 6

Working With Databases
Questions to Answer about a
Database System
• What functions the marketing database is expected
to perform?
• What is the initial and continuing investment cost?
• How long it would take to install and to complete
a set of baseline capabilities?
• What would be required, and of whom to recover
and profit from the initial investment?
Key Issues
• One size does not fit all.
• “off-the-shelf” marketing database systems do not
work.
• Possible reasons:
– Marketing systems are dynamic
– Marketing is largely creative
– Based on moving targets of opportunity
Cost Justifying the Investment
• Costs easy to calculate accurately.
• Profits/returns from implementing a marketing system is
not directly observable.
• Marketing database systems affect the bottom line through
marketing programs developed based on them.
• Effects of the database systems depends on the efficiency
of the marketing programs.
• Effects of the marketing programs may take years to
realize.
Cost Justifying the Investment
• Possible return estimation methods:
– Published results of other companies.
– What-if scenarios:
• Find how much of acquisition, retention etc. is
required to recover the cost.
• Do a cost-benefit analysis based on how attainable
the values in the previous step are.
Learning to Live With Imperfect Data
• Data is not either “good” or “bad”.
• Marketing database system users need to
understand that
• Error in data are possible due to entry errors,
processing errors, limited sources etc.
• If the errors are within acceptable limits it is OK.
• Good co ordination and interaction is
required between IT managers and
marketing managers.
Learning to Live With Imperfect Data
• Marketing database systems are developed
because a company now has new marketing
objectives.
• New objectives require restructuring of the
department.
• Personnel may need training, or new
personnel with different skills may need to
be hired.
Managing Expectations
• Expectations about the benefits of the system
needs to be made clear early on.
• During the planning process there should be
agreement on
– Which applications the database is expected to support?
– Anticipated costs, schedules, and benefits.
Objectives, Processes and
Requirements
• Marketing objectives and IT processes and requirements
can be designed sequentially or simultaneously.
• Simultaneous is preferred because
• IT personnel can quickly point out if the objectives of the marketing
department is cost effective and feasible from a database viewpoint.
• Difference arises due to issue regarding flexibility in
access versus standardization across all users.
Marketing Tools
• Marketers should focus on possessing a set of information
delivery and usage channels.
• Use the channels on a needs-basis.
• The common tools marketers need include:
• Research/analytic projects and software
• Standardized reports
• EIS/OLAP
• Ad-hoc query tools
• Analysis Desktop software
• Promotion implementation support software
Business Needs and Functional
Requirements-Queries
• Queries- Ad-hoc questions about the characteristics and
behavior of customers or prospects.
• Type I- Answered directly and fully using the marketing
database system.
• What is the age distribution of purchases of Air Jordon?
• Type II- More complex, require additional analysis beyond
what the database is designed to provide.
• How many same-product mailings are required prior to conversion?
Business Needs and Functional
Requirements-Queries
• Campaign Planning
– Software that assists in finding customers that qualify
for a campaign.
– Identifies the criteria for selections.
– Provides groups of customers in each segment that is
being promoted.
– More sophisticated systems automatically
• Feed campaign specifications
• Update annual marketing plan management system
• Allocate random customers to control group.
Business Needs and Functional
Requirements-Queries
• List selections– select a list of names for a promotion and track
responses
• Event driven promotions– Update process that identifies particular events that
trigger promotions
• Scoring models– Facilitate the creation of statistical models and the
subsequent implementation of model results.
Selecting Names based on Model
Scores
• The database system should
–
–
–
–
–
–
Incorporate model equations
Execute the equation to evaluate customer scores
Store the scores produced by the equation
Sort the file in descending sequence by score
Dividing the file into a number of equal sized groups
Assigning customers to appropriate deciles based on
their scores
– Providing a facility for selecting names based on the
deciles or raw model score
Tracking Promotion Results and
Profiling
• Tracking of promotions is performed to
• Evaluate actual, and expected circulation/contact
quantities, gross responses, net responses, sales
volume etc.
• Profiles of responders and non responders
• Comparisons of test and control groups
• Predicted Vs. actual scoring model performance
• Profiles are based on queries
Tracking Promotion Results and
Profiling
• Uses of profiling customers include:
– Evaluate scoring model results by customer group
– Compare scoring model results across multiple criteria
– Evaluate the demographic and lifestyle differences
between deciles
– Evaluate the characteristics of the responsive and least
responsive segment
• Used to design creative customized messages
based on the differences between groups.
Turnaround Time
• Information requirements should be ordered by
priority.
• Turnaround time is the time between when a job is
submitted to the database and when the results are
obtained.
• Marketers cannot have all the information
immediately.
• Cost of computer hardware needs to be considered
in designing database systems
Developing a source data strategy
• Step 1: Determine value of data for all
groups and applications
- develop data proforma
• Step 2: Evaluate and maximize internal
data-gathering resources
- what data can you get from customers?
- evaluate ways to communicate with
your customers
Developing a source data strategy
• Step 3: Evaluate and maximize external data
gathering resources
• Step 4: Evaluate and maximize custom data
development options
- custom surveys
- research panels
- customer participation promotions
- warranty programs
- relationship enhancement programs
Developing a source data strategy
•
•
Step 5: Confirm the correct mix of source data for
database marketing applications
- whether to maintain data identified in
proforma
- what is bottom line impact of maintaining data.
- did it allow you to
1. Learn about your customers or market
2. Develop better products
3. Upsell or cross-sell product
4. Develop new customers/ keep customers
Step 6: Develop corporate strategy for mkt database