Title: Key Strategies In Marketing
1 Key Strategies In Marketing
Developing Analyzing Strategies in Uncertain
Environments
2Marketing Strategy Framework
Competitor Analysis
Company Analysis
Customer Analysis
Complementors Analysis
Targeting
Positioning
Marketing Mix
External Environment
Performance Metrics
3The Effect of Uncertainty on Decision-making
- Perceptions of uncertainty often dominate
decision-making - Perceptions of risk vary with the individual
- Perceptions of risk vary with the situation
- Managers tend to be risk-averse when decision
will be reviewed - Managers tend to be risk-averse when things are
going well - Perceptions of risk vary with the range of
outcomes - Managers tend to overweight potential losses
- Yet, most managers often assess Uncertainty by
gut feel - 96 of managers alter or reject risk assessments
provided to them (March and Shapira 1987)
4Strategy Under Uncertainty
- 1) Clear Enough Future
- Mature markets, low-technology
- 2) Alternate Futures
- Discrete events forthcoming from one uncertainty
source technology, competitive, market - 3) Range of Futures
- High levels of uncertainty from multiple sources
- 4) True Ambiguity
- Many sources of uncertainty that interact
Courtney, Kirkland, and Viguerie 1997
5How Firms Should Respond to Uncertainty
- 1) Shape the Future
- Big bets -- big risks
- May come from large or small firms
- 2) Adapt to the Future
- No-regrets moves
- 3) Reserve the right to play
- Options, no-regrets moves
- Purpose is to transform big-bets into no-regrets
moves
6How Firms Actually Respond to Uncertainty
- 1) Head in the Sand
- 2) Toe in the Water
- 3) Take the Tiger by the Tail
7How Firms Actually Respond to Uncertainty
Point estimates of strategy outcomes such as
profits, sales, etc. Sensitivity analysis to
account for risk Full range of outcomes
associated with a major strategy
8Systematic Ways to Evaluate Full Range of Outcomes
- There are some useful tools to help assess risk
- Decision trees
- Sensitivity analysis
- However, these can be inaccurate or difficult to
use - Dont inspire managerial confidence
- Dont capture interactions between multiple
variables - Often used in an unsystematic way
We need a better way to estimate strategic risk!
9La Shampoo
For the look and feel of France
10La Shampoo -- Financials
Brand Building Strategy Price 2.25 Variable
Cost .95 CM 1.30 Fixed Costs 30M Sales
Volume 28M Net Profit 6.4M
Which is the better option?
11La Shampoo -- A closer look at the numbers
- Actual value of Discount strategy could be
affected by - Competitive response
- Channel acceptance of new positioning
- Response of loyal customers
- Response of price sensitive customers
- Actual value of Brand Building strategy could
be affected by - Competitive response
- Channel acceptance of new positioning
- Promotional expenses (advertising, sales
promotion) - Promotional quality
12La ShampooOutcome distribution -- Discount
Strategy
Most likely Value 3.0 million Range of
Outcomes 1.8 million to 4.2 million
13La ShampooOutcome Distribution -- Brand
Building Strategy
Most likely Value 6.4 million Range of
Outcomes -1.6 million to 9.0 million
14Risk Analysis Process
15Putting the Players in the Process
Decision Makers
Data Gathering
Modeling
Implement.
Framing
The Analysts
16Stage 1 Framing the Decision
- Framing is simply the way that individuals think
about a decision - Example The proverbial half full (half empty)
glass of water - One of the most critical factors in
decision-making - Proper framing involves developing a common
understanding of the decision, its key elements,
and the decision-making process
17Stage 1 Framing the Decision
- Signs of a poor decision frame
- Poor quality information is disowned
- Real decision-maker is not involved or doesnt
understand - Manger needed for implementation not included in
process - New issues raised at conclusion
- Assumptions challenged after the fact
- Managers need a tool to facilitate framing
process - Influence diagrams
18Two Approaches to Decision-making
Traditional Business Case Approach
Facts Assumptions
Forecast Model with Risk Discount
What will it take for this strategy to be
successful
An Inductive Approach
Cashflow Model with Risk Modeling
Influence Diagram
What should I know to make the best decision?
19The Influence Diagram Shows the Relationships of
all Uncertainties for a Decision
Price Level
New Product Design
Revenues
Sales Volume
NPV
Fixed Costs
Costs
S,G A
Investment
Marginal Costs
It illustrates the relationships among uncertain
quantities and shows their influence on the focal
decision criterion
20Influence Diagrams
- Focus on what we need to learn, rather than what
we already know - Working back from the decision criterion produces
a different type of analysis model than working
forward from available facts - Starts with outcome measure
- Identifies major uncertainties that need to be
understood - Directs efforts toward resolution of key
uncertainties rather than more detailed analysis
of existing data
21What is Influence
- One factor influences another when it is a major
determinant of the possible future outcomes - For example, we know that the volume of next
years sales will be influenced by our
competitors pricing strategy as well as our own
pricing strategy
Competitor Pricing
Expected Sales
Our Price
22Building the Diagram
Decisions
Uncertainties
Calculation
Outcomes
Influence Indicator
23An Example Comparing Two Strategies
Brand Building Strategy
Price Discount Strategy
24LaShampooKey Uncertainties for Brand Building
Strategy
- Advertising
- Management approves base plus contingency
spending - Channel push
- Some retailers will require quarterly slotting
allowances for shelf space - Sales promotions
- Coupon redemption depends on quality of
advertising - Product development costs
- Depends on results of consumer research
- Competitive response
- Affects sales volume
25Influence Diagram for Brand Building Strategy
26Influence Diagram for Brand Building Strategy
Price
Sales Increase
Adv. Quality
Compet. Response
27Influence Diagram for Brand Building Strategy
Price
Sales Increase
Adv. Quality
Compet. Response
28Influence Diagram for Brand Building Strategy
Price
Sales Increase
Adv. Quality
Compet. Response
Market Charact.
29Completing the Diagram
- The objective is to continue dis-aggregating
until we have specific quantities to be estimated - Avoid getting bogged down in minutia
- The influence diagram is NOT the model
- The process of making the diagram is as important
as the result - Creates a common frame of reference to view the
problem - Develops consensus at an early stage
- Can be revised as the analysis team learns more
about the decision
30Influence Diagram
- Final outcome of stage 1 is a listing of experts
to be interviewed - about key uncertainties
Uncertainty Expert Department Prod.
Development Cost Benny Fitz Engineering Advertisi
ng Expense Natalie Drest Marketing Slotting
Allowances Ima Goober Sales Sales
Forecast Mary-Kay Mart Market Research Sales
Promotions Exp. Natalie Drest Marketing
31Stage 2 Data Gathering
32Expert Interviews
- Data gathering involves seeking out the experts
identified in Stage 1 and conducting interviews - Traditional approach attempts to pin the expert
down to a single number - Focus on what must go right for success
- Ad Hoc approach to risk assessment
- Risk analysis approach attempts to define the
full range of values that might occur - Explore the factors that could create extreme
outcomes - Understand what could go wrong / right
33What if There are Two Experts?
- Frequently the decision-makers will appoint two
experts - Often competing factions will want their expert
to have input - How would you handle this?
Expert 1
Expert 2
50
75
100
25
34Conducting the Interview
- 1) Qualify the expert
- 2) Search for bias (expert or motivational)
- 3) Define quantity precisely (determine units)
- 4) Have expert explain factors behind extreme
optimistic/pessimistic outcomes - 5) Explore likelihood of outcomes (5/50/95
values) - Most likely value
- Symmetry
- Uniformity
- 6) Have expert review big picture and make
adjustments
35Objectives of the Interview
- Identify the appropriate distribution of possible
outcomes - A few distributions will cover most scenarios
- Identify any bias
- Identify drivers of extreme values
- Major distributions include
- Normal / Lognormal
- Triangular
- Uniform
36The Normal Distribution
- Normal distribution is most common
- Not bounded on high or low sides
- Must be symmetric
- Not appropriate for some values that must always
be positive - e.g., unit sales
37The Log-Normal Distribution
- Lognormal distribution overcomes some limitations
of normal - Distribution is always positive
- Can be skewed
38The Triangular Distribution
- Triangular distribution is powerful, and easy to
use - Covers most situations
- Can be skewed
39The Triangular Distribution
- Triangular distribution is powerful, and easy to
use - Covers most situations
- Can be skewed
40The Uniform Distribution
- The uniform is also simple and intuitive
- Useful when there is not a most likely value
41The Custom Distribution
- The custom distribution is a variant of the
uniform - Useful when there is a contingent factor
affecting the decision
42LaShampoo Assessing Key Variables
Distribution Uncertainty Units Function
Assumptions Prod. Dev. Costs Triangular 0 -
1.0M 50 - 1.5M 100 -
2.3M Advertising Exp. Uniform 0 -
12M 100 - 15M Slotting Allowances Triang
ular 5 - 2M 50 - 5M 95 - 8M
Sales Forecast Units Triangular 0 -
23M 50 - 28M 100 - 30M Cons.
Promotions / unit Normal 5 - .20 50 -
.30 95 - .40 Production Costs /
unit Uniform 0 - 0.90 0 - 1.05
43Stage 3 Modeling the Strategy
44Monte Carlo Simulation
- A process that creates an artificial model of the
real world by modeling the distribution of
important and related variables and then sampling
from these distributions many times to estimate
the impact on an outcome variable - Conceptually simple
- Computationally intensive
- Commercial packages only recently available
45The Monte Carlo Principle
Revenues
Profits
Costs
46LaShampoo Brand Building Model
47Stage 3 Primary Outcome
The primary outcome of the modeling stage is a
probability distribution of the focal variable
48Supplementary Analyses
- The modeling approach allows the analyst to probe
the data to answer important questions regarding
the key drivers of the final outcome - Supplementary Analyses include
- Probability analysis
- Quantitative sensitivity analysis
- Analysis of key drivers
- Qualitative sensitivity analysis
49Probability Analysis
- The goal of probability analysis is to provide
insights into the downside risk and upside
potential of a strategy - Probability analysis allows the analysts to
answer strategic questions such as - What is the probability that this strategy will
be profitable? - What is the probability that this strategy will
achieve its target level of profits? - What is the probability that the strategy will
exceed expectations?
50LaShampoo Brand Building Strategy
- What is the probability that LaShampoo will
achieve its target profitability of 8,000,000?
Approx. 25
51LaShampoo Brand Building Strategy
- What is the probability that the brand building
strategy will lose money?
Approx. 2
52Quantitative Sensitivity Analysis
- The purpose of quantitative sensitivity analysis
is to identify the critical variables that drive
strategy outcomes - Essential part of analysis
- Provides insights into sources of downside risk
- Enables decision team to direct attention where
it will do the most good
53LaShampoo Brand Building Strategy
Consumer Promotions Sales Forecast Production
Cost Slotting Allowances Advertising Product Dev.
Costs
-.49 .47 -.44 -.41 -.31 -.09
54Tornado Diagram
55Qualitative Analysis
- The qualitative analysis is intended to yield
insights into the key variables identified by the
quantitative analysis - Relies on the expert interviews
- Purpose is to identify specific actions that
could be undertaken to manage downside risk or
improve upside potential
56Stage 4 Decision / Implementation
57Making the Decision
- The analysis is not the only factor to be
considered in the decision. Other factors
include - Strategic considerations
- Risk environment
- Individual tolerance
- Size of project
- Financial stability of division, company
- Social / political environment
58Implementation
- Implementing a decision means translating the
outcomes of the analysis made in our pseudo-world
into concrete actions in the real world - Good decisions can still be derailed by poor
implementation - Implementation will be broader than the analysis
- New issues will emerge
- But key issues should have already been resolved
59Developing Strategies in Uncertain Environments
- We have talked about a new approach to evaluating
strategies - Its powerful
- Its flexible
- Its thorough
- At this point, you should be able to go out and
do your own analyses - Be cautious, however, this type of modeling can
be subtle - Stick to the basics