Household Panel Data Reference

1
Household Panel Data Reference
2
Household Panel Data Analyses
The types of analyses that can be done with
Household Panel data can be broadly classified
to address the following issues

• Understanding Basic Purchase Behavior
• Market Summary Reports
• Purchase Dynamics
• Deal Measures
• Identifying Key Target Groups
• Demographics
• Purchase Behavior
• Heavy/Light
• Loyal
• Deal Sensitive

• Evaluating New Product Introduction
• Trial Repeat
• Source of Volume
• Other Diagnostics
• Combination Purchasing
• New/Lost/Retained Buyer Flow Analysis
• Shopping Expenditures
• Market Structure/ Segmentation Study

3
Basic Applications

• Market Summaries
• Purchase Distributions (H/M/L)
• Demographic Profiles

4
Market Summary Report

• Applications/Questions
• How many households purchase this category ? My
  brand ?
• Is my brand at a disadvantage in terms of
  purchase size or purchase frequency ?
• How deal-reliant are my buyers ?
• How deal-reliant are my competitive buyers ?
• How have things changed over time ?

• Report Model
• Provides either a comprehensive snapshot or a
  dynamic look at consumer purchasing behavior for
  a category, its major segments and brands.
• Aids in identifying appropriate marketing
  strategies and tactics.

5
Market Summary Basics
Consumer behavior data complements sales and
share measurement data. Sales of Buyers
x Amount Purchased per Buyer
Penetration x Buying Rate
Occasions per Buyer x Units per Occasion
(Purchase Frequency x Purchase Size)
6
Household Measure Definitions

• Penetration - of total U.S. households making
  at least one purchase of the specified item.
• Volume per Buyer - (Buying Rate) How much the
  average buyer purchases in a given time frame.
• Volume per Occasion - Average volume purchased
  per household each time a purchase was made.
• Occasions per Buyer - Average number of times the
  specified item was purchased.
• Share of Requirements - (Loyalty) Percent of the
  specified items buyers total category volume
  accounted for by the specified item.
• Elapsed days - (Purchase Cycle) Average number of
  days between purchases among buyers of the
  specified item who made two or more purchases.
• of One Time Buyers - of all the specified
  items buyers who purchased the item only once.

7
Household Measure Definitions

• Volume on Total Deal - Proportion of all volume
  purchased on a consumer perceived coupon,
  feature, display, TPR.
• Volume with Coupon - of volume purchased on a
  deal specified by panelist as coupon.
• Average Price - Average price paid
• Average Non-Deal Price - Average price paid when
  no consumer perceived deals were present.
• Average Deal Price - Average price paid for
  purchases made on a consumer perceived deal.
• Volume Purchased by Outlet Type - of
  specified items volume that was purchased in
  each outlet.
• Buyers Projected - Number of households
  purchasing the specified item.
• Raw Buyers (N ) - (Sample Size) Unprojected
  panel households purchasing the specified item.

8
Marketing Tools Affecting Consumer Purchase
Components

• Purchase Components
• Penetration
• Buying Rate
• Purchase Frequency
• Purchase Size

• Tools for Change
• Advertising
• Product Offering
• Couponing, Features, Displays
• Trial Packs
• Product Sampling, Distribution
• Competitive Activity
• Smaller Package Sizes
• Couponing
• Displays
• In-Pack/On-Pack Promotions
• Sweepstakes/Contests
• Advertising Alternative Uses
• Two-For Deals
• Bonus Packs
• Trade Deals
• Displays with Larger Sizes

9
Common Errors or Cautions in the Interpretation
of Market Summary Analysis

• As a general rule, none of the base measurements
  are additive !
• Buying rate, purchase frequency and purchase
  size, represent average
• behavior and should not be taken literally !
• Average purchase frequency can not be used to
  calculate purchase cycle.
• Example 365 days divided by average annual
  category frequency of 10 times does not equate to
  a purchase cycle of 36.5 days.
• Reason Average purchase frequency includes
  households who make only one purchase during the
  given time frame and, therefore, do not have a
  purchase cycle. This measure can only be
  calculated among households with a purchase
  frequency of greater than one.
• Loyalty (Share of Requirements) measurements over
  shorter time periods will almost always lead to
  higher figures.
• Penetration-based marketing efforts can have a
  negative impact on loyalty. The more new brand
  buyers that are brought in, the greater the
  likelihood that total brand loyalty will erode.
  The reverse is also true.
• Declines in penetration levels almost always lead
  to increased loyalty levels.

10
Demographic Profiles

• Applications/Questions
• What types of households typically buy my
  products ?
• Is my product positioned (through advertising and
  promotion) to reach these buyers?
• How do the buyers of my brand compare to buyers
  of my competitors ?
• Should I target my advertising to buyers of their
  products ?
• Should I introduce a product similar to my
  competitors to appeal to a different customer
  profile ?
• How has the profile of my buyers changed ?
• Did my line extension bring in a different type
  of buyer ?

• Report Model
• Identifies those demographic groups which make up
  the largest portions of a brands total buyer
  base as well as a disproportionate share of
  sales.
• Provides a complete picture of the
  characteristics for brand buyers and other groups
  of interest.

11
Demographic Definitions
What are the demographic characteristics that we
most commonly utilize in analyses?

• Household Income
• Household Size
• Female Head Age
• Age/Presence of Children
• Female Head Employment
• Female Head Education

• Household Head Occupation
• Race/Origin
• Household Lifestages
• Household Affluency
• County Size
• Census Regions

12
Demographic Profile Definitions

• Household Lifestages
• Young Singles - Households with one member under
  35 years old.
• Childless Younger Couples - 1 member households,
  under 35 years old with no children under 18.
• New Families - Households with children under 6
  years old only.
• Maturing Families - Households with children aged
  6-17 years old.
• Established Families - Households with children
  aged 13-17 years.
• Middle-Aged Singles - Households with one member
  aged between 35-54 years old.
• Middle-Aged Childless Couples - 1 member
  households, aged between 35-54 years old, with no
  children under 18.
• Older Singles - Households with one member over
  55 years old.
• Empty Nesters - 1 member households, aged
  between 55-64, with no children under 18.

13
Demographic Profile Definitions

• Per Capita Income Definitions
• HHs are Defined by Poor Getting By
  Comfortable Affluent
• 1 Member Income 0-9,999 10-19,999
  20-39,999 40,000
• 2 Member Income 0-11,999 12-24,999
  25-49,999 50,000
• 3 Member Income 0-14,999 15-29,999
  30-59,999 60,000
• 4 Member Income 0-19,999 20-39,999
  40-69,999 70,000
• 5 Member Income 0-24,999 25-49,999
  50-99,999 100,000

14
Common Errors or Cautions in the
Interpretation/Use of Demographic Profiles

• Use caution when examining demographic profiles
  among buyer groups that contain less than 200 raw
  (unprojected) sample households.
• Use caution when interpreting results from any
  demographic segment in which the raw buyer count
  is less than 30 households.
• Analyses of indices can be misleading without
  also examining the relative size of a demographic
  segment. High volume or high buyer indices may
  not offer much potential if a segment accounts
  for a small percentage of the population.

15
Comparing Household and Store Level Data

• Projected volume from the household panel can be
  compared to universe estimates from scanning data
  or shipment data.
• While projected volume should not be expected to
  match exactly, the data should be directionally
  similar in terms of brand shares and trends.
• An estimated 10 of Scantrack volume is purchased
  for use outside the home and, therefore, would
  not be reported by the panel.
• The two services are sampling different
  universes
• Scantrack 2MM Grocery Stores
• Household Contiguous U.S. Households
• Scantrack samples scanning stores only
• Household collects purchase data from all stores
• Sampling ratios are more robust for Scantrack
  than Household
• Scantrack 1 in every 11 stores is a sample store
• Household 1 in every 2400 households is a
  panelist (40,000 household panel)

16
Issues to consider when comparing Panel and
Scantrack data

• Scantrack measures all volume moving through
  2MM grocery stores.
• Household measures volume consumed by households
  usually in the home.
• Some consumer purchases never make it into the
  home (i.e. candy bars, canned soda, single-serve
  snacks, travel size HBAs)
• Industrial purchases at retail (laundromats,
  day care centers) would not be captured.
• Basis for volume projection is different
• Scantrack projects to all commodity volume
• Household projects to Total U.S. population based
  on household type incidence.
• Outlet types are self-classified by panelists
  (given guidelines)
• Snantrack sample stores and organizations are
  specifically identified in the Scantrack sample.