Implementation/The Future

1
Implementation/The Future

• A Look Back
• A Look Ahead
• ME as Lifetime Learning

2
A Look Back . . .

• Marketing Engineering is Marketing.
• Marketing Engineering is a means to an end.
• Models require judgement.
• Whole gt sum of parts.
• Software Û rapid prototyping (personal
  modeling).

3
ImplementationA Five Step Process

• 1. Problem identification
• 2. Modeling
• 3. Calibration
• 4. Implementation
• 5. Control/Improvement

4
ImplementationStep 1Problem Identification

• Review symptoms Û diagnose disease.
• Eliminate irrelevant issues include all relevant
  issues.
• Agree on problem with all relevant stake-holders
  (sources of funds, data, implementation
  responsibility) generate shared ownership up
  front.

5
ImplementationStep 2Modeling

• Variables
• Relationships
• Objectives

6
ImplementationStep 3Calibration

• (Assign values to parameters)
• Objective
• Subjective (judgmental)
• Combined (Bayesian)

7
ImplementationStep 4Implementation

• Be opportunistic
• Start simple/keep it simple
• Begin with an end in mind (Not What data do I
  have?)

8
ImplementationStep 5Control/Improvement

• Make it a program,not a project

9
What Can You Do NOW?

• 1. Find a problem
• New Product entry??
• Dumb spreadsheet?
• Resource allocation?
• 2. Build a boxes and arrows diagram.
• 3. Develop a model
• Yourself (off-the-shelf)/ME adaptation/Excel.
• In-house expert (needs expertise to build/use
  model).
• Expert consultant (new model/state-of-the-art
  methodology).
• 4. Calibrate Evaluate Implement Control.

10
Marketing EngineeringA Look Ahead

• Data warehousing
• Online analytic processing (OLAP)
• Intelligent marketing systems
• Simulations
• Groupware
• Improved user interfaces
• ASP model/Lifetime learning

11
To Sizzle Brand Manager From CoverStory Date 07
/05/89 Subject Sizzle Brand Summary for Twelve
Weeks Ending May 21, 1989
Sizzles share of type in total United States was
8.3 in the CB Juice/Drink category for the
twelve weeks ending 5/21/98. This is an increase
of 0.2 points from a year earlier, but down 0.3
from last period. This reflects volume sales of
8.2 million gallons. Category volume (currently
99.9 million gallons) declined 1.3 from a year
earlier. Sizzles share of type is 8.3up
0.2from the same period last year. Display
activity and unsupported price cuts rose over the
past yearunsupported price cuts from 38 points
to 46. Featuring and price remained at about the
same level as a year earlier.
Components of Sizzle Share Among Sizzles major
competitors, the principal gainers are Shakey
up 2.5 points from last year to 32.6 Private
Label 0.5 to 19.9 (but down 0.3 since last
period) and loser Generic Seltzer 0.7 to
3.5 Shakeys share of type increase is . . .
Components of Sizzle Share Among components of
Sizzle, the principal gainer is Sizzle 64oz up
0.5 points from last year to 3.7 and
losers Sizzle 48 oz down 0.2 to 1.9 Sizzle 32
oz down 0.1 to 0.7 Sizzle 64 ozs share of type
increase is partly due to 11.3 points ride in
ACV with Display vs. year ago.
Xerox 10-98 6 - 8
12
Modeling Technology
Intelligent/automated decision models
Groupware plus decision models
OLAP plus decision models
MarketingEngineeringtomorrow
Stand-alone marketing decision models
General purpose analysis tools (e.g., SPSS, LP
packages)
MarketingEngineeringtoday
Model User
Analyst
Trainedmanager(in marketingengineering)
Novicemanager
Non-managerialemployee
Forecasting
Allocation and optimization
Simulation
An overview of the evolution of Marketing
Engineering to support a wider range of users
and decision tasks using emerging technologies
Explanation
Decision Tasks
Xerox 10-98 6 - 9
13
The New Learning Model

• We are all becoming knowledge workers
• Learning and work are becoming contemporaneous.
• The shelf life of information/knowledge is
  shrinking dramatically
• Learning is becoming a perpetual process.
• Information is becoming freely available to
  learners
• Learning is becoming a democratic process.
• New knowledge/information is being created at
  mind-numbing rates.
• Learning is becoming a discernment process.
• Knowledge is needed just-in-time, not
  just-in-case
• Learning is becoming context-embedded.
• Career moves and job changes are becoming more
  frequent.
• Learning is becoming self-directed.

14
(No Transcript)
15
Our Future Plans forMarketing Engineering

• Continue to test effects of Marketing Engineering
  under controlled experimental conditions
• Further implement ideas/tools in companies
• Professional versions on line
• Continuous improvement of existing models
• Develop new models (e.g., generalized resource
  allocator, sales territory alignment, enhanced
  choice modeling)
• Marketing Engineering on the Net as ASP
• ME/HCV Portalsstay connected---www,mktgeng.com/ww
  w.valueharvest.com