2. THE VISION
Superior Decisions
Inferior Data
To help our customers traverse the path from Data to Decisions quickly,
effectively and efficiently
3. THE APPROACH
Idea
• Frame the Decision need
Analysis
• Manage the necessary data, models and analysis
Decision
• Support Decision making through insight generation
Execution
• Monitor and track decision execution and effectiveness
Value
• Measure and report value
Full Lifecycle support for a truly Analytical decision making process
4. THE FOUNDATIONS
Content Carrier
• Relevant and • Appropriate
Reliable Information business
application to
deliver the insights
at the right point
Consumption Cost
• Present the insights • All of this at a cost
in a form that is that needs no
easily consumed by justification!
the business
5. OUR CAPABILITIES
Data Infrastructure Modeling and
Management Optimization
We also build and deliver
Our Structured Information
mathematical models custom
Framework (SIF) is a fast
tailored to the customer
implementation data
needs that cover areas like
management solution that has
statistical predictive models,
been deployed at varying
optimization models (LP, MIP)
scales (Large Enterprise to
using Open source tools like
Localized Business Unit)
COIN-OR, R
6. STRUCTURED INFORMATION FRAMEWORK
Source Data Useable Data Relevant & Actionable Business Value through
(Uncertain (Analytical Data analysis better decisions
Hygiene) Store)
Data Staging –
Tested, Filtered, Strategic Modeling
cleansed
Metrics Definition and Collaborative
Analytical Data Store Tracking Decision
(Commercial
Database) Support and
Analysis
Exceptions and Alerts
Data Integrity Operational
Dashboard, Data Dashboards
quality metrics
7. RESOURCE MANAGEMENT – ALLOCATION AND
OPTIMIZATION
Situation / Problem Statement Solution
• Web based Resource Management and
• Customer has multiple Allocation
simultaneous projects with • Captures Resource Cost, Skills and
varying deliverables and Competencies
skill requirements • Project Requirements – Timelines,
Skill requirements and target margins
• Globally distributed work • Global Allocation Optimizer identifies
force, with varying cost and “Best” resource to assign to each project
based on
competencies • Skill / Competency matching
• Manage allocation of • Margin maximization
resources to projects to • Operational constraints
optimize on margins and • MIP Formulation solved using open
source solver
delivery SLA • Identifies Skill and competency gaps for
further Training and Development
8. CUSTOMER ATTRITION AND LOYALTY
Situation / Problem Statement Solution
• B2C customer facing a • Statistical model “Predicts”
Customer Attrition individual customers
challenge propensity to attrite and
• How to identify potential identifies critical factors that
customer attrition and precipitate attrition
initiate preventive • Identifies “High-Risk”
measures customers with clear
callouts of the reason for
the attrition risk
• Define a “Play Book” that
triggers action from the
provider for each High-Risk
customer based on the
identified factors
10. ENGAGEMENT MODELS TO SUIT EVERY
BUSINESS NEED
Consultative Engagement
• Identify problems, opportunities and challenges
• Time and Material based costing
Services Delivery
• Build custom solutions and models to solve critical problems
• Fixed price engagements, Pay-Per-Use
Embedded expertise – Retainer model
• Invest, Build and manage customer specific expertise and knowledge
• Annual subscription model
BOT
• Transfer expertise and personnel to customer for continued in-house support
• To be discussed
11. DELIVERY MODEL GEARED TO UNLOCKING
VALUE
SaaS Ramp up
Only pay for Rapid ROI
delivery consumed
what you use Realization
model solutions
Hosted on
Cost- Consulting Extremely
Pay-Per-Use
Effective and Analytics fast delivery
or
Cloud offerings can cycles
Subscription
infrastructure be added „a ensures
models
– No upfront la carte‟ to rapid ROI
available
costs the solution realization
involved
12. THE PEOPLE
• Anand Srinivasan
• Anand is the founder and CEO and brings several years of experience managing the
R&D and modeling teams at Sabre and Dell.
• He has vast and rich expertise in solving large optimization problems for Fleet and
Crew optimization, Revenue Management etc.
• He has also worked with the Supply Chain team at Dell building models for
manufacturing optimization, logistics, lean manufacturing, supply network
optimization, inventory staging and management and pricing for demand shaping
• Anand Srikumar
• Anand Srikumar brings several years of experience in the Banking and Financial
Services industry with GE and Standard Chartered Bank
• He brings expertise in Consumer Credit scoring, Portfolio Risk Management and
Financial Modeling
15. WHAT IS REVENUE MANAGEMENT AND WHY IS IT
RELEVANT?
• The practice of selling the right room, for the right price, to the right customer, through the
right channel
• Industry benchmark standards have established that Implementing a Revenue Management
solution can increase Room Rent Revenues by 3% to 6% with minimal increase in operational
costs
• Revenue increase of 17% have been reported in some cases
16. REVENUE MANAGEMENT COMPONENTS AND
ALGORITHMS
Unconstrained Demand Forecasting
• Observed Historical Bookings is a “end state” of bookings after some bookings have been
rejected due to capacity constraints
• Statistical models Forecast the “Unconstrained” (Without capacity restrictions) demand for
various price points.
• “How many people wanted to buy?” from “How many people bought”
Optimization of Inventory controls
• Optimization algorithm to identify the “Best” price point to sell rooms at
• A Dynamic programming formulation that uses the unconstrained forecast for a given
occupancy day and works backward to determine the optimal selling price on any given day
leading up to the occupancy
• Can be re-optimized periodically to reset prices based on actual observed demand
17. RM SOLUTION – SYSTEM SAMPLE
Forecasted Occupancy for
RM price recommendations various Price Points (Baseline,
for future dates Low Fare, and High Fare)
Target Occupancy
Date
Today
Estimated Benefits of
recommended controls
19. THE PERSONAL COMPUTER MARKET
• Consumers are flooded with choice
• Brands, Models, Configurations, Form Factors, Special Offers
• Guaranteed to find a model (configuration) that meets the need of all but the most
esoteric consumer requirement
• Brand Equity has been significantly diluted
• Informed consumers look to other brands for the right configuration/price
• Extremely critical to hit the right configuration in the market at the right price point
Only possible approach to find the “Sweet Spot” is using Choice Models
20. CONSTRUCT
Group Customers by Segment
• Laptop, Tablet etc.
Choice Set visible to the Segment
• Available makes, models, configurations
Significant Product Attributes
• Memory, Screen Size, HD Capacity, Price etc.
Estimate importance of each attribute
• Statistical Techniques
Predict probability (Market Share) of each possible configuration introduced
into the market
Optimize promotional activity and Marcomm spend to drive customers to high
margin sales
21. MODEL PREDICTED AND ACTUAL SHARES
COMPARISON
50 Feb 12 (Calibration) 60 Mar 12 (Prediction)
40 50
3.4% 40 16.3%
30 Error
30
20
20
10 10
0 0
Acer Compaq Dell HP Lenovo Sony Toshiba Acer Compaq Dell HP Lenovo Sony Toshiba
Forecasted Actual Forecasted Actual
50 40 May 12 (Prediction)
Apr 12 (Prediction)
40 30 12.8%
42.6%
30 Error Error
20
20
10
10
0 0
Acer Compaq Dell HP Lenovo Sony Toshiba Acer Compaq Dell HP Lenovo Sony Toshiba
Forecasted Actual
Forecasted Actual
Apr Data shows large deviation from Prediction – Possibly due to heave promotional activity by Dell
22. HOW TO USE THE MODEL
• Profitability Enhancement
• Clearly identified attributes that impact sales
• Can choose high margin components between “nodes of indifference”
• E.g. Screen Resolution
• Sales Target Setting
• The Model predicts “Fair Share” based on product attributes. Any share capture above that
can be attributed to Sales and Marketing Activity
• MarCom ROI
• Can quantify the uplift generated in sales in a given month due the MarCom Spend.
• Apr Marcom/Promotions resulted in +20% market share for Dell
• Competitive Response
• Evaluate the impact of new product launch or price move by competition and respond
appropriately
Notes de l'éditeur
This definition of RM brings into focus ALL the key components available to the Hotel Revenue Manager to increase profitabilityPriceCustomer SegmentationRoom TypeSales ChannelHotels have reported increasing Revenues (Top Line) by 3%-6% through implementing a RM solution. Some cases have reported increases as high as 17%
A intuitive interface allows the Revenue Manager to analyze any future occupancy date and view optimized recommendations for setting the “BAR” for future dates based on expected occupancy