Ever wondered if you should increase your product price or decrease your pack size? Among the many options your pricing, strategy, and research teams face in a competitive consumer environment, this is one of the most common. In our webinar we explore ways to optimize your pricing and your product portfolio composition to maximize overall revenue.
SKIM webinar "Product Portfolio and Revenue Optimization"
1. Product Portfolio and Revenue Optimization
Juan Andres Tello Scott Garrison
SKIM Director Americas Today’s webinar host
j.tello@skimgroup.com s.garrison@skimgroup.com
SKIM Webinar May 24, 2012 | “Product Portfolio and Revenue Optimization.”
2. Outline
Motivation for Revenue Optimization (RO)
RO requires a MR shift from insight to forecast
Building blocks of an RO system
a) Consumer behavior models
b) Demand forecasting
c) Constrained optimization approach
Some RO strategies
Delivering optimization results to clients
2
3. Revenue Optimization - Motivation
• Maximize:
Revenue = f(Pricing, Product portfolio composition | Selling channel)
• Turns data into actionable foresight tools for clients
• Determine optimal pricing/portfolio strategy within given constraints
• RO pioneers: fixed capacity industries
4. How to charge the max willingness to pay to each customer?
marginal cost
Demand C d(p)
1,000
A B
0 $5 $10 $15
Price
• Solution price differentiation (sometimes controversial)
5. RO requires a MR shift from insight to foresight
90% of consumer-facing companies have a Consumer Insights (CI)
function in early stages of development (1) or (2)
4
BCG’s CI stages of development 3 Strategic
foresight
2 Strategic organization
insight
Business organization
1 contribution
Traditional team
MR function
MR as an Consumer insight
order-taking as a source of
function competitive
advantage
Source: BCG Consumer Insight Benchmarking (May 2009)
6. Building blocks of a RO system
1. Quantitative models of consumer behavior Choice based Conjoint
(CBC)
2. Demand forecasts Market simulator
3. Constrained optimization tools Search algorithm of optimal
solution within market constraints
7. 1. Choice based Conjoint
• Proven and unbiased research technique to model consumer
preferences and market heterogeneity
• Rooted in Utility Theory (Von Neumann–Morgenstern)
• Preferences estimation process has evolved over time:
1. Aggregate Logit model (one size fits all)
2. Latent class (segmentation)
3. Hierarchical Bayes (individual level)
• Choice task resembles purchase behavior process
9. 2. Market Simulator: from consumer preferences to
market shares, to revenue forecasting
• Volumetric adjustments and
calibrations Input prices
• Ability to test unlimited pricing Change
/portfolio strategies and portfolio
potential competitive reactions composition
Market share
output
• In its simplest form, the
Revenue
simulator is a “show of hands” output
from respondents given a
number of choice options
10. 3. Searching for the optimal: define the feasible space first
Total space
of possible Constrained space of
solutions feasible solutions
Sample of solutions
within constraints
11. 3. Searching for the optimal: define objective function &
apply search algorithm
Max
Revenue
(Optimal
solution)
Revenue
Surface
12. 3. It’s not only about finding the winning solution, but about
the patterns observed
• While the main goal is to uncover the
strategy that maximizes revenue, ask
yourself:
Focus on
• What makes it the optimal solution? Max Rev gain = 8%
upper right
quadrant
• Are there alternate strategies with
different tradeoffs yielding positive
results?
• In this example:
• 80,000 scenarios generated
• 40% yield gains in both revenue and
share. Cluster analysis is used to
further group and interpret
13. Some RO strategies
1. Maximize volume share profitably (capping revenue loss)
Balanced “investment” strategy to grow customer base
2. Maximize revenue while capping volume loss
Ideal situation, not always feasible; will depend on price elasticity
3. Game theory strategies: competitive reactions
14. Delivering optimization results to clients
A few insights for a successful deployment:
• Involve key stakeholders from different functions early in the game
• Plan accordingly
• Kick-off: constraints from every function are expressed and discussed
• Delivery: results are discussed in a workshop style
• Dynamic session
• Create tools that allow clients to interact with the data (e.g. ability to
activate/deactivate constraints, rank and select scenarios)
• Don’t be afraid to show the “raw” data; involve stakeholders in the analysis
• As always, be clear about the model’s assumptions and limitations
15. contact us or follow us online!
Juan Andres Tello Scott Garrison
SKIM Director Americas Today’s webinar host
j.tello@skimgroup.com s.garrison@skimgroup.com