Driving Business The CEO Challenge

1
Data Quality The Effect on Topline Growth
Ramin Eivaz
Vice President, Customer Insights Business
Intelligence
2
The Agenda

• The real IMPACT of data quality

• Marketplace challenges affecting data quality

• PepsiCo IRIs approach to data quality

• The future complexities

3
What is the real
IMPACT
IMPACT
of bad data?
4
What is the real
IMPACT
of bad data?
Superior data is a key contributor to driving
profitable growth.
Inferior data is often the key contributor to
missed opportunities and poor performance.
5
Key Learning from recent Gartner Study

• Through 2007, more than 25 of critical data with
  fortune 1000 businesses will be inaccurate or
  incomplete.
• Important decisions are routinely made based on
  flawed data
• Data problems cause wasted labor and lost
  productivity which directly affect revenue
• Most companies invest significantly in technology
  solutions but these are ineffective because
  underlying data is bad

-611,000,000,000
Source Gartner 2004-2005 Strategic Planning
Report, The Data Warehousing Institute, Yankee
Group, Forrester Research Inc.
6
Incorrect data negatively impacts the entire
organization.
7
Missing EPS targets has serious ramifications on
Wall Street.
Lost Value -512MM
Lost Value -3.3B
Lost Value -2.0B
Lost Value -904MM
8
Marketplace dynamics amplify the challenge to
data quality
9
Media has fragmented
10
Media has fragmented

• Network TV

• Cable TV

• Broadcast radio

• Satellite radio

• Internet

• Print

• Outdoor

• Direct mail

• Movie theaters

• Transit

• Telemarketing

11
Media has fragmented

• Network TV
• Cable TV
• Broadcast radio
• Satellite radio
• Internet
• Print
• Outdoor
• Direct mail
• Movie theaters
• Transit
• Telemarketing

12
Consumers have changed
13
Consumers have changed

• Gender

• Age

• Lifestyle

• Geography

• Attitudes

• Ethnicity

• Income

• Education

• Health

14
Consumers have changed

• Gender
• Age
• Lifestyle
• Geography
• Attitudes
• Ethnicity
• Income
• Education
• Health

Consumer
15
Channels have proliferated
Consumer
16
Channels have proliferated

• Grocery

• Drug

• Mass

• Super centers

• Dollar

Consumer

• Club

• Convenience

• School

• Out-of-home

• On-line

• On-premise

• Vending

17
Channels have proliferated

• Grocery
• Drug
• Mass
• Super centers
• Dollar
• Club
• Convenience
• Work place
• Out-of-home
• On-line
• On-premise
• Vending

Consumer
Channel
18
Marketing approaches are more sophisticated
Media
Channel
Consumer
19
Demands placed upon data are greater now
20
Demands placed upon data are greater now
21
Demands placed upon data are greater now
22
Demands placed upon data are greater now
23
Demands placed upon data are greater now
24
Market projection has become more of a challenge.

• Stratification Factors
• ACV
• County Population
• County profile
• Store Format
• Weekly Volume
• Square Footage
• Demographics

Cross Channel effects are not considered in
sample selection.
25
How does PepsiCoaddress the quality issue with
IRI?
26
The Issues

• PepsiCo needed to be confident in the quality of
  the data that was being used to make critical
  decisions

• We needed to remove as much risk from the
  decision-making process as possible

• Without real confidence, decision-making would
  not be as fast, decisive or bold

• Inability to reconcile between our internal and
  external market data, further eroded our
  confidence in the information.

• Extensive effort was exerted to explain, validate
  and justify the data, which diverted our focus
  from our key deliverables.

27
Data was aligned between PepsiCo and IRI.
28
The result was better decisions and new insights.
OUTPUT

• Understanding the variance
• Raw data - outlier stores, PLUs, highcones,etc.
• Projected data - IRI projection discrepancies,
  sampling variation, etc.

• New insights
• Inventory and Lag times
• Supply Chain transparency

• New benefits
• Improved IRIs overall projection and data
  cleansing
• Highlighted the power of census over sample data
  in the multi channel environment.
• Advanced tracking and diagnostic capabilities

29
Data Lab allows us to identify drivers even
within a small variance.
Least Variance -0.6
30
Key outcomes from Data Lab

• Confidence in data is at its highest
• Ability to articulate quality on a proactive
  basis
• Decision-making has accelerated
• Comfort levels surrounding critical actions are
  increased
• People are accountable data quality is no
  longer an excuse for missed targets

• Data Lab output benefits all clients through the
  identification of outliers, projection
  discrepancies and sampling variation
• Able to resolve data quality issues quickly and
  efficiently
• Client Solutions group spends more time adding
  value and less time on operational details

31
Data Lab elements are also key enablers to our
consumer 360º vision.

• Standardization of data
• Integration of data
• Development of new measures and metrics
• Understanding the relationship among the
  components of the consumer 360º

Manufacturer Shipment
Retail Sales
32
The Challenge Accelerates
33
Data quality will be even more critical as the
usage of data expands.
34
The foundation of an efficient supply chain is
high quality data.
Manufacturer
Brand Marketing
Market Research
Sales Distribution
Manufacturing
Finance
Standardized Processes
High Quality Data
35
Data quality is a key enabler of better
decision-making and a true source of competitive
advantage.
Over 600B is lost each year due to bad data.
How much is YOUR companys share?