National Dairy Council

1
Milk Processor Education Program (MilkPEP) And
Dairy Management Inc. (DMI) School Milk
Vending Test
Evaluating Opportunities for School Milk Vending
September 5, 2001
RESEARCH CONSULTING FOR THE GLOBAL BEVERAGE
INDUSTRY
2
School Milk Vending Test Evaluating
Opportunities for School Milk Vending

• Contents
• Background
• Test Objectives and Methodology
• Key Learnings/Messages from the Test
• Test Results
• Business Models for School Milk Vending
• School Milk Vending Opportunity
• Issues and Hurdles
• Conclusions

3
School Milk Vending Test Evaluating
Opportunities for School Milk Vending

• Contents
• Background
• Test Objectives and Methodology
• Key Learnings/Messages from the Test
• Test Results
• Business Models for School Milk Vending
• School Milk Vending Opportunity
• Issues and Hurdles
• Conclusions

4
Background

• Kids drink milk in schools because they like it,
  but also because they have to
• When there are more attractive options available,
  kids quite often choose them over milk
• School milk has some quite negative
  characteristics and associations for many students

• Why Kids Drink School Milk
• They like milk
• Gatekeeping by school/parents
• Its free with hot lunches
• It goes with certain foods (e.g. sweets)
• Chocolate milk is a treat for young kids

• Why Kids Dont Drink/Stop Drinking School Milk
• Prefer to drink something else
• Milk is sometimes bad/warm
• Absence of preferred flavor/type
• Insufficient portions
• Negative association with school milk
• Unattractive/juvenile packaging
• Package inconvenient
• Free milk perceived lower value vs. other
  beverages

5
Background

• In fact, students rate school milk between
  neutral and bad
• Soft drinks, on the other hand, are rated good

Kids Image of School Milk

• Parents gatekeep
• Soda is sweet
• Variety of flavors
• A treat
• Bubbles and caffeine
• Comes in neat packaging
• Comes in teen-size portion

• Parents push
• White milk is plain, blah
• Limited options
• Only flavored milk is a treat
• No bubbles or caffeine
• Comes in a box
• Comes in child-size portion

Source Kidfacts
6
Background

• The battle is only getting tougher for school
  milk, as soft drink companies target schools as a
  critical market for recruiting life-long, loyal
  customers for an ever-widening variety of
  beverages
• These companies are building relationships with
  school districts through strong sales and
  marketing programs and exclusive contracts
  (pouring rights)

• Revenue generating programs for schools/districts
• Value-added educational tools and scholarships
• A broad range of popular products students want
• Availability throughout the day (through vending)
• Fun, image building promotions
• Attractive display/dispensing equipment free to
  the schools

Competitor Offerings
7
Background

• Kids have identified some very basic improvements
  that they say would make them drink more milk
• However, to date, very little effort has been
  made by schools or the dairy industry to respond
  to these needs

What Kids Say Would Make Them Drink More School
Milk

• If school milk tasted better
• Always fresh
• Always cold
• If milk tasted different
• New flavors e.g. banana, coffee
• New formats e.g. milk shakes
• If school milk looked appealing
• More attractive packaging
• Cleaner packaging
• If milk were more convenient
• In portable/resealable packages
• The right serving size
• Available off the lunch line

Source Kidfacts DDC
8
Background
Historically, regulatory restrictions and
economic considerations have limited milks
profitability in schools, and in turn, have
limited processors ability to/interest in
innovating or marketing
School Milk Realities
Implications

• Legislated regulations effectively limits type of
  products, size and price of milk on the meal
  line, where 90 of school milk is purchased

• Considering only the school meal line, processors
  have seen little incentive to innovate, limiting
  milks volumetric, profit and competitive
  opportunities
• Reinforces students negative experience with
  school milk missed opportunity to build consumer
  interest, loyalty and consumption

• Processor margins on school milk are
  significantly lower than for milk sold through
  other retail channels
• 5-10 of volume, but only 2-5 of profit pool

• No margin to invest in sales, marketing or
  merchandising efforts other milk-selling venues
  get processor focus

• School milk is not competitive with other
  beverages now offered to students in schools in
  terms of availability, variety, convenience,
  packaging and imagery

• Opens door for competitors to buy their way into
  schools and milk loses long-/short-term
  competitive advantage
• Students form life-long consumption habits that
  do not include milk

Source 2000 Processor Audit Beverage Marketing
Corporation
9
Background

• Processors have been innovating outside the
  school arena, however, and the increasingly
  available 16-ounce plastic resealable bottle
  provides a new tool for enhancing school milk
• Vending the plastic value-added milk products
  appears to present a significant opportunity that
  would address student demands and make milk more
  competitive in schools

Vending Addresses Students Concerns
What Kids Dont Like About School Milk . . .
Vending Addresses

• Limited availability only during meals in the
  cafeteria

• Available at meals, throughout the day,
  before/after school
• Machines placed in or outside cafeterias

• Bad packaging small, leaky cartons

• Portable, resalable plastic 16-oz. bottles

• Limited flavors and fat level options

• Wide variety of products

• Perceived as higher-end, cool beverage
• Paying for and selecting product gives kids a
  degree of control

• Image as uncool beverage kids have to take

• Kept very cold until dispensed

• Warm from sitting out on lunch line

• Have to wait in line to get it

• Freedom from waiting in line

10
Background
In fact, in one study of student attitudes,
students said they would be as likely to buy milk
as any other beverage from vending machines
Students That Would Purchase All the Time from
the Vending Machine if Available
Source Beverage Marketing Corporation DDC
11
Background
However, with milk vending practically
non-existent, and soft drink machines pervasive,
students have not had the opportunity to purchase
vended milk
Presence of Competitive Beverages in School
Vending 2000
Penetration
Source Beverage Marketing Corporation FMS
12
Background

• Vending value-added milk not only addresses
  students needs, but also makes school milk much
  more attractive for processors
• Additionally, it makes schools and parents feel
  better about kids beverage options in schools

Vending Addresses Others Concerns, Too
Key Concerns . . .
Vending Addresses

• Dairy Processors
• School milk is barely profitable

• Processors can be more profitable on value-added
  products

• Price of school meal milk inflexible and
  determined by Federal regulations

• Vending price determined by competitive
  environment

• Cannot afford to spend on marketing programs in
  schools

• Vending machines act as billboard and can
  enhance kids overall image of milk
• Higher margins can support some level of programs

• Kids have a healthy choice with vended milk

• Parents and Schools
• Between meal beverage options limited to soda and
  other less healthy products

13
Background
Perhaps most important is the potential long-term
impact milk vending can have on changing the
image of milk in schools for all constituencies
Vending Can Help Break the School Milk Paradigm
Old Paradigm
New Paradigm

• Milk is cheap -- it comes with the free lunch
• Milk is uncool
• Its for little kids

Students

• Milk is cool
• It comes in great packages and flavors -- just
  like soft drinks
• A great option instead of soda and fruit drinks

• A politically correct way to drive cafeteria
  revenues
• We can push as CSD alternative that kids love
• Provides great educational tools and equipment --
  a helpful partner
• A revenue generator for student programs

• Milk is a way to get money back from the
  government
• An inexpensive, federally regulated commodity
• Milk is key to the nutritional health of kids

Schools

• A vital way to recruit consumers for life
• A profit generator
• A great testing ground for new ideas

Processors

• I cant make any money on school milk
• A necessary evil

14
School Milk Vending Test Evaluating
Opportunities for School Milk Vending

• Contents
• Background
• Test Objectives and Methodology
• Key Learnings/Messages from the Test
• Test Results
• Business Models for School Milk Vending
• School Milk Vending Opportunity
• Issues and Hurdles
• Conclusions

15
Test Objectives and Methodology
Primary Test Objective

• To evaluate the opportunity for milk vending in
  schools

• To understand and quantify the volumetric and
  profit potential for school milk vending
• To understand the operating dynamics of selling
  milk through vending machines
• To identify optimal products, placement,
  merchandising and service levels
• To build vending economic models for dairy
  processors and third-party vending operators

Supporting Objective
16
Test Objectives and Methodology
Several test elements, not part of the initial
test concept, added complexity and forced some
on the fly decisions/changes
Initial Test Concept
Test Realities
75-80 vending machines in four markets
96 machines in five markets
Full, standardized array of SKUs
SKUs differ slightly by market and flavors were
limited dependent on processor capabilities
Open-armed welcome from schools, with no school
commissions offered or required
Bureaucratic and competitive hurdles to get
placements, with commissions demanded in almost
all markets
User-friendly data collection and manipulation
through electronic web-based technology (i.e.
E-vend)

• Operator collected data with limited e-vend
  back-up data not available in desired detail
  (e.g. by day, day-part)
• E-vend not working added expense, time, delayed
  machine operations

• Districts often dictated participating schools,
  and school dictated where machine could be
  located
• Limited dual placements

• Control over machine placements in schools
• Key variable would be dual vender placements

• Vandalism was a major factor in some school
  districts
• Delayed installation in several schools

Vandalism was not considered a significant issue
17
Test Objectives and Methodology

• The five test markets were selected based on
  processor capabilities/interest, third-party vend
  operator capabilities and geographic
  representation
• Vending machines were placed in middle and high
  schools, with dual machine placements in 13 high
  schools

18
Test Objectives and Methodology

• For a number of reasons, including school
  district practices, machine placement was
  staggered over a number of weeks
• Non-installed and non-reporting machines were
  generally due to damaged equipment, vandalism or
  delays due to school decision/approval process

Milk Vending Machine Placements in Test Markets
Percent Reporting for 6 Weeks
Machines Originally Assigned
Placement Completion Date
Machines Placed
Market

• Austin
• Omaha
• Southern California
• LA
• Orange Cnty
• Miami
• Boston
• Total

24 16 12 12 18 18 100
24 16 10 12 16 18 96
February 7 February 20(1) 1st wave Feb 27 2nd
wave Apr 20(2) 3rd wave May 7(2) April
1 March 28
88 75 79 89 94 85
(1) Additionally machines placed in April (2)
Several machines could not be placed until vandal
guards/cages had been ordered and installed
19
Test Objectives and Methodology

• Eight-five (85) milk venders generated adequate
  data during the test period to be included in
  data analysis
• Threshold operating time required for inclusion
  was six full weeks excluding vacation and other
  unusual weeks (e.g. testing, snow days)

School Milk Vending Test Sample Size By Number of
Reporting Weeks
6-8 Weeks
9-12 Weeks
13-16 Weeks
Total
of Venders 40 24 21 85 of Sample 25 28 47 1
00
20
Test Objectives and Methodology

• Dixie-Narco supplied glass front venders for the
  test
• The vending machines featured the got milk? logo
  and milk mustache celebrities

• Vender Features
• 45 product facings/glass front
• 360 unit capacity
• Advanced refrigeration system
• Temperature guard with automatic shutdown
• Celebrity Graphics
• Tony Hawk
• Dixie Chicks
• Everyone Loves Raymond
• Ricky Martin
• Back Street Boys

21
Test Objectives and Methodology

• The vended milk was priced at 1.00 across all
  test markets
• 1.00 was in line with competitive vended
  products in similar serving sizes

Typical Beverage Vending Prices in Schools
Item Omaha Boston Austin Los Angeles Miami
20-oz. Plastic Soda 1.00 1.05 1.00 1.00 1.00
Bottled Water 1.00 1.00 1.00 1.00 0.85 Bottle
d Juice 1.00 1.25 1.25 1.25 1.25 Canned
Juice 0.85 0.85 0.80 0.75 0.75 Isotonics 1.2
5 1.25 1.25 1.25 1.00 Bottled
Tea N/A 1.25 1.25 1.25 1.00 Milk
Test 1.00 1.00 1.00 1.00 1.00
22
Test Objectives and Methodology

• Flavors offered vary by market, depending on
  processor product line
• Test parameters required at least three flavors,
  and each market also offered a variety of fat
  levels

Vend Test Product Variety by Market
SKUs Omaha Boston Austin Los Angeles Miami
Whole White ? ? ? ? ? Low Fat White ? ? ? ? ? Skim
White ? Chocolate Whole ? ? ? ? ? Chocolate
Low Fat ? ? Strawberry Low Fat ? ? ? ? Orange
? Coffee ? Total SKUs 5 5 4 6 4
23
Test Objectives and Methodology

• Despite efforts to obtain school participation
  without paying a commission, most of the schools
  were unwilling to forego vending revenue, even
  during the test period
• For the test, however, the schools generally
  accepted lower commissions than would be
  customary
• Commissions ranged from no commissions to 10

Range of Commissions/Subsidies Paid to Schools
for School Milk Vending Test Participation
Test Commission Rate
24
Test Objectives and Methodology
The test was executed through the last half of
the school year 2000/2001
Test Planning and Setup
Analysis
Test

• Monitored test progress/conditions
• Ongoing data collection
• Periodic product mix review/revision
• Student survey
• Began preliminary analysis
• Post-test interviews with processors/ operators

• Designed, purchased and configured vending
  machines
• Identified test market areas
• Identified and recruited participating schools,
  dairies, third-party operators
• Finalized test variables
• Product mix
• Placements
• POS, etc.

• Reviewed data
• Developed projections and volumetric opportunity
• Identified best practices - products, placements,
  merchandising, service
• Built third-party operator business models
• Built dairy business models

July 2000
January 2001
June 2001
August 2001
March 2001
Board Approval
Final Report
Machine Placement
25
Test Objectives and Methodology

• Success of this complex test was dependent upon
  the strong coalition of industry partners
• The team included a number of constituencies
  acting in concert, but with clearly defined roles

School Milk Vending Test Project Team
Maytag Dixie-Narco
MilkPEP
DMI
Project Management Team BMC Bachtelle
Associates MilkPEP/DMI Dixie-Narco
Dairy Processors
Vend Operators
S/R Reps
26
Test Objectives and Methodology

• Key Responsibilities
• Designed strategy and execution plan
• Helped recruit participating processors
• Facilitated partnership of constituencies for
  successful test execution
• Assisted in promotional activities
• Led public relations efforts
• A sponsor of the test
• Provided vending machines
• Helped place and serviced vending machines
• Trained operators on machine use
• In-field trouble-shooting
• Agreed to participate within test parameters and
  consistently provide milk to the packaging and
  variety specifications of the test
• Assisted in-school recruiting/machine placements
• Supplied data
• Worked closely with vending operators to ensure
  appropriate milk supply, etc.

MilkPEP/ DMI
Maytag Dixie-Narco
Dairy Processors
27
Test Objectives and Methodology

• Key Responsibilities
• Recruited schools and placed venders
• Supplied/serviced vending machines
• In-field trouble-shooting, including key school
  contact
• Provided required data for research evaluation
• Communicated continually with project management
  team, and processors to ensure smooth test
  execution
• Assisted in P/R efforts
• Conducted student and SFSD surveys
• Assisted with promotional activities
• Local liaison between all constituencies
• Designed strategy and execution plan
• Recruited participating dairies and vend
  operators
• Managed negotiations with and among participants
• Collected and analyzed data
• Communicated/liased continually with all
  constituencies
• Prepared questionnaires for students, SFSD,
  processors, operators
• In-field trouble-shooting

Vend Operators
S/R Reps
Project Managers BMC/Bachtelle
28
School Milk Vending Test Evaluating
Opportunities for School Milk Vending

• Contents
• Background
• Test Objectives and Methodology
• Key Learnings/Messages from the Test
• Test Results
• Business Models for School Milk Vending
• School Milk Vending Opportunity
• Issues and Hurdles
• Conclusions

29
Key Learnings/Messages from the Vending Test
What We Learned from the School Milk Vending Test

• Kids will eagerly buy milk from vending machines
  in schools
• Vending will be the preferred venue for kids to
  buy flavored milk, especially products they cant
  get on the lunch or a la carte lines
• Flavors out sell white milk nearly 10 to 1
• Milk vending can be a profitable business, for
  both processors and/or vending operators given
  the right product mix, pricing and operational
  components
• Commissions to schools will be likely
• There is no single approach to school milk
  vending
• Not all schools are appropriate for milk vending
• There will be significant competitive response by
  the soft drink players
• There was no significant lunch line
  cannibalization
• Spoilage was not a significant issue/expense

30
School Milk Vending Test Evaluating
Opportunities for School Milk Vending

• Contents
• Background
• Test Objectives and Methodology
• Key Learnings/Messages from the Test
• Test Results
• Business Models for School Milk Vending
• School Milk Vending Opportunity
• Issues and Hurdles
• Conclusions

31
School Milk Vending Test Evaluating
Opportunities for School Milk Vending

• Contents
• Test Results
• The Data
• Student Reaction
• Processor and Third-Party Operator Feedback

32
Test Results The Data

• Average weekly sales by machine during the test
  were 280 for all schools
• High schools achieved higher velocities,
  primarily due to the larger school populations
  compared to middle schools

Average Milk Vender Sales Sales Per Machine Per
Week
Installation Period 1st 3 weeks
Base Line Period After 1st 3 weeks
All Weeks
33
Test Results The Data

• Average weekly sales per school were generally
  higher - 327 per week
• Again, high schools outperformed middle schools,
  in this case nearly two to one consistent with
  school population

Average Milk Sales Per School Per Week for All
Weeks
Average School Population
1,871
905
1,582
34
Test Results The Data

• Most high schools in the test generated weekly
  sales of greater than 200, while the machines
  placed in middle schools had much lower
  velocities
• Student population was the key driver of these
  variations
• As with any test that forces location
  distribution, some locations performed at
  significantly below-average levels and negatively
  skewed overall results

Average Weekly Sales Per Machine
Share of Milk Venders
Average Weekly Sales
35
Test Results The Data

• Excluding the poorest performing schools, each
  market had significantly better results
• This analysis yields a more real world result,
  as poor-performing machines would be pulled from
  operation and placed in higher velocity locations
• For the test, very few schools significantly
  underperformed the market average 10-30 in
  each market

Average Milk Vender Sales Excluding
Poor-Performing Machines Sales Per Machine Per
Week All Weeks
Poor-performing machines had significantly lower
per week sales compared to the market-average
during the base-line test period
36
Test Results The Data

• Looking at the test results on a per capita basis
  reveals somewhat higher per capita consumption
  for middle school students versus high school
• This is consistent with general milk consumption
  patterns by age group, and is largely a result of
  increased beverage options in high schools

Vended Milk Per Capita Consumption Excluding
Poor-Performing Machines Based on All Week Sales
Units Per Week
Poor-performing machines had significantly lower
per week sales compared to the market-average
during the base-line test period
37
Test Results The Data
This means that over a 36-week school year, each
student would consume an average of seven to
eight vended milks in school or an additional
one gallon
Vended Milk Projected Annual Per Capita
Consumption
7.9-8.6
7.2-7.9
7.2-7.9
Units Consumed During School Year
38
Test Results The Data

• The degree of post-installation performance
  drop-off differed significantly by market, too
• Across all markets, baseline sales averaged 64
  of installation velocity - better than expected,
  based on an industry average of 50-60
• In LA Central, where the drop-off was most
  precipitous, there were machine vandalism and
  other unforeseen disruptions

Average Milk Vender Sales by Market Installation
vs. Baseline Performance Sales Per Machine Per
Week
Installation Weeks - 1st 3 weeks
Baseline Weeks - After 1st 3 weeks
Baseline as of Installation
54
58
45
53
87
57
39
Test Results The Data
All test market per caps clustered near the 0.20
units per week average, with the exception of
Boston, which was significantly higher, and
Austin, which was significantly lower
Vended Milk Per Capita Consumption by
Market Based on All Week Sales
Units Per Week
40
Test Results The Data
Vended school milk per caps, as with total
volume, fell off after the initial few weeks of
the test dramatically in some markets
Vended Milk Per Caps by Market Installation vs.
Baseline Performance
Units Per Week
41
Test Results The Data
There were a number of possible reasons for the
wide variation in per capita consumption rates
among markets and among schools within each market

• Economic factors directional volume associated
  with high/low area income levels
• Machine volume if machine access was
  limited/had lower foot traffic
• Machine performance equipment problems had
  significant impact on usage and reduced consumer
  confidence
• Location foodservice programs - expanded
  foodservice (and beverage) options drove lower
  sales volume through milk vender
• Competitive product available at discount in
  Omaha identical product sold a la carte for
  0.80 cut into vending sales

Vended Milk Per Capita Consumption Drivers
42
Test Results The Data

• Dual machine placements made sense in large
  schools
• In 9 of the 13 dual location schools, per caps
  for both machines were in the range of or
  exceeded overall average machine per caps

Dual vs. Single Vender Per Capita
Consumption Baseline Weeks Only (Weeks 4)
Units Per Week Per Machine
Excludes 6 under-performing dual locations
43
Test Results The Data

• Even in schools with low per capita consumption,
  such as in LA, if the schools are large enough
  they will easily support 2 venders or more
• In the smaller Omaha schools, the second vender
  was generally not necessary

Dual Vender School Total Sales Baseline Weeks Only
Average Units Per Week
Miami
LA - Central
Omaha
Boston
Average School Population
2,570 3,285 1,400 1,500
Average Machine Per Caps
0.27 0.14 0.21 0.32
44
Test Results The Data

• From start to finish and consistently across all
  markets, flavored milk outsold white milk 9 to 1
• Chocolate was the best-selling flavor, but all
  the flavors offered performed very well

School Milk Vending Volume by Flavor Share of
Sales by Market
Austin Boston Los Angeles Miami Omaha Total White
10 11 9 9 14 10 Chocolate 73 60
65 51 47 60 Strawberry 18 --- 27 40
29 24 Orange --- --- --- --- 10 1 Coff
ee --- 29 --- --- --- 4 Flavor
Subtotal 90 89 92 91 86 90
45
School Milk Vending Test Evaluating
Opportunities for School Milk Vending

• Contents
• Test Results
• The Data
• Student Reaction
• Processor and Third-Party Operator Feedback

46
Test Results Student Reaction

• A self-administered, informal survey was
  conducted toward the end of the test to capture
  student reactions to school milk vending
• While these survey results are not projectable,
  they offer valuable insights into student
  opinions

• Number of Respondents
• 362 High School Students
• 188 Middle School Students
• 550 Total
• Gender of Respondents
• 49 - Female
• 47 - Male
• 4 - No Response
• Grade of Respondents
• 13 Fifth grade students
• 69 Sixth grade students
• 31 Seventh grade students
• 68 Eighth grade students
• 114 Ninth grade students
• 67 Tenth grade students
• 103 Eleventh grade students
• 50 Twelfth grade students
• 35 No response

Post Test Student Surveys
47
Test Results Student Reaction

• Most students with the opportunity used the milk
  vending machines
• Consistent with per cap levels derived from the
  sales data, the share of students buying milk
  from the machines was somewhat higher in middle
  schools than high schools

Students Usage of Milk Vender in Schools
Share of Total Respondents
Source Beverage Marketing Corporation Student
Survey
48
Test Results Student Reaction

• There were a significant number of regular users
  roughly 30 of users purchased milk from the
  venders about once a week or more often
• Less than 20 used the machine only once

Student Frequency of Using School Milk Venders
Share of Total Respondents
Source Beverage Marketing Corporation Student
Survey
49
Test Results Student Reaction

• Most milk sales from venders were incremental for
  school milk if anything, vended milk
  cannibalized competitive beverages much more than
  the school lunch/breakfast line
• Kids said they drank vended milk instead of soft
  drinks, fruit drinks and water

Students Bought Milk Rather Than Other Beverages
Share of Total Respondents
Note Survey question read If the got milk?
vending machine was not in your school what
would you buy instead of milk Source Beverage
Marketing Corporation Student Survey
50
Test Results Student Reaction

• Students were generally satisfied with the milk
  serving size of 16-ounce they definitely did
  not want smaller portions
• However, a number of students would have liked
  more milk per vended serving

Students Reaction to 16-ounce Serving Size
Share of Total Respondents
Source Beverage Marketing Corporation Student
Survey
51
Test Results Student Reaction

• Especially for middle schools, lunch is the most
  popular time for using the milk venders
• High School students used the venders more for
  breakfast and before lunch
• When milk was bought was likely influenced by
  where the vender is placed, however, test data
  showed no clear correlations between location and
  overall sales

When Students Usually Bought Milk from Vender
Share of Users
Source Beverage Marketing Corporation Student
Survey
52
Test Results Student Reaction
Especially in middle schools, vender use
continued throughout the test three quarters of
middle school students had purchased milk within
a week or so of completing the survey, which was
conducted toward the end of test
Last Purchase from Milk Vender
Share of Users
Source Beverage Marketing Corporation Student
Survey
53
Test Results Student Reaction

• Reasons for not using the venders differed
  significantly between middle and high schools,
  with lack of money the key issue for younger kids
  
• High school students stated a dislike for milk as
  a primary deterrent

Students Reasons for Not Using Vender
Share of Students Not Using Vender
Includes buy from lunch line/bring from home
lactose intolerant inconvenient location too
high in fat Source Beverage Marketing
Corporation Student Survey
54
Test Results Student Reaction

• By and large, students had very good things to
  say about the got milk? vending machines
• When asked what they thought of the vending
  program, 84 of the comments from both middle
  school and high school students were positive

Straight From the Students Mouths

• I think that its a very creative idea to
  influence our students into buying milk
• I like it, I wish it was closer to the cafeteria
  because I would buy it more
• Its beautiful and pretty and I love it so much
  Yea for got milk? vending machine
• Its great and its very nutritious for those
  with calcium deficiencies
• Handy Dont have to stand in line
• I like it, especially the Dixie Chicks it gives
  us something else to drink
• Its a great way to call kids attention to
  drinking milk
• I like it sometimes me and my dad get
  strawberry milk after school because its our
  favorite
• I like it it gives us a choice besides soda
• I like it because it gives us more variety and
  more milk

55
Test Results Student Reaction

• Of course, not all of the students comments were
  positive
• Those that were not, primarily focused on the
  cost, the perceived quality and the machine
  itself

Straight From the Students Mouths

• It costs too much for too little better deal
  through lunch line
• All the milk has too much fat in it
• I think its easy to flip over and get free
  milk
• Sometimes there is milk in the machine and after
  you press the button, it says sold out, even
  though theres still milk in there
• I dont think many kids are into buying milk
  from it also, you can get free milk in the lunch
  line
• Milk in the machine for extended period of time
  scares me
• It needs a new picture on the side
• Its good, but I dont use it very much because
  its too expensive
• Its kinda cool, but dumb too because milk is in
  lunch line for 50 cents

56
School Milk Vending Test Evaluating
Opportunities for School Milk Vending

• Contents
• Test Results
• The Data
• Student Reaction
• Processor and Third-Party Operator Feedback

57
Test Results Processor Feedback

• Without exception, all five of the processors who
  participated in the test were pleasantly
  surprised by the results
• Most of the processors are already planning or
  executing programs to expand vending in the
  coming school year

Processors Feedback on School Milk Vending
Participating Processors

• Pleased/surprised with test results 5 of 5
• Plan to continue vending school milk 4 of 5
• With third-party operators 3
• Processor full-service operation 0
• School self-operation 1
• Plan to expand milk product variety in venders 3
  of 4
• Plan to add non-milk products 3 of 4 (a)
• See vending opportunities outside of schools 5 of
  5

(a) In at least one district, vend bids may
prohibit inclusion of non-milk products
58
Test Results Processor Feedback

• Processor comments focused on the school milk
  vending opportunity, and the importance of
  keeping the effort alive
• The only real negative comments from the
  processor had to do with machine operation
  primarily e-vend

Straight From the Processors Mouths

• I wish wed done this ten years ago!
• This test served as an awakening for
  single-serve milk through vending
• I doubt the dairy industry would have gotten
  there on its own
• Im surprised that the kids in the schools
  accepted the milk so well. The dairy industrys
  campaigns have really paid off
• As long as we can keep kids tuned into milk,
  this thing can keep going... Vending takes this
  to a whole new level
• The branding opportunities (through school
  vending) will have a lasting effect
• Milk in vending is a hot commodity right now we
  have to strike while the iron is hot
• With todays equipment technology, packaging and
  positive trends for milk, milk is very visible
  now to consumers
• Vending is the best thing to happen to milk
  since the Chug

59
Test Results Third-Party Operator Feedback

• Similar to processors, the third-party operators
  were unanimously pleased with vending test
  results
• However, they clearly see the potential hurdles
  to expanding school milk vending

Third-Party Operator Feedback on School Milk
Vending
Participating Operators

• Test results better than expected 6/6
• Expect lower margins for school milk than average
  vend business 5/6
• Received excellent dairy support 5/6
• Commission issues are major hurdle for future
  programs 4/6
• Plan to dedicate route for school vending in
  future 2/6
• Would like to add more milk options 4/6
• Plan to add non-milk beverages 4/6
• Spoilage was less than expected 4/6
• Fewer than expected vandalism issues 4/6

60
Test Results Third-Party Operator Feedback

• The third-party operators are generally
  optimistic about the future of school milk
  vending
• They hope to improve their margins through lower
  product costs, and commission level containment,
  but understand that high velocities can generate
  efficiencies and higher profitability

Third-Party Operator Views on Future of School
Milk Vending

• Very good future opportunity
• Expect to continue and expand
• Need more dairy support (cost or equipment
  sharing)
• Plan to expand milk vending beyond schools
• Exclusive school soft drink contracts could be an
  obstacle
• Money can be made

61
School Milk Vending Test Evaluating
Opportunities for School Milk Vending

• Contents
• Background
• Test Objectives and Methodology
• Key Learnings/Messages from the Test
• Test Results
• Business Models for School Milk Vending
• School Milk Vending Opportunity
• Issues and Hurdles
• Conclusions

62
Business Models for School Milk Vending

• There are at least four viable business models
  for school milk vending the differences relate
  to who purchases/leases the venders and who
  stocks and services them
• School size and volumetric potential are key
  considerations for determining the appropriate
  model
• Processor margins, infrastructure and school
  relationships are other key variables that can
  effect the choice of model

Third-Party Operation
Processor Full-Service
School Milk Vending Business Models
Processor/School Cooperative Service
School Self-Operation
63
Business Models for School Milk Vending
Third-Party Operator
Pros
Cons

• Processor and school can enter market for
  little/no money low barriers to entry
• Machine filling, maintenance and service are all
  provided by third- party operator
• Exploits expertise and experience of third-party
  operator
• Provides processors entry into schools that have
  vending contracts with third-party operators
• Keeps processors out of equipment business,
  minimizes cost and infrastructure requirements
• Allows for large drop sizes versus small drops at
  a number of schools
• Can get into schools without having school
  contract
• Third-party operators tend to get choice
  locations in schools

• Processor shares profit with third-party operator
• Processor forfeits some control over what goes
  into the venders
• Processor must depend on operator for quality
  control product temperature risk
• Not viable for smaller schools
• Forfeits control over vend price

From processor point of view
64
Business Models for School Milk Vending
Third-Party Operator

• Based on assumptions derived from the vending
  test, third-party operators who purchase and
  service the school milk venders would realize an
  IRR of about 15
• This conservative estimate assumes a commission
  rate of 20 actual commission rate will depend
  on competitive circumstances

Theoretical School Milk Vending Financial Model
for Third Party Operators
Financials
Note This model assumes the third-party operator
is visiting the schools for other machines also
assumes no investment in trucks or additional
routes
65
Business Models for School Milk Vending
Third-Party Operator

• At lower volumes, third-party operation may
  present a more difficult profit scenario
• Lower commission rates and less expensive
  machines are key to profitability for third-party
  operators

Vending Economics for Third-Party Operator With
Variable Vend Price, Volumes, Commission Rates,
Machine Costs
4,500 machines 1.25 vend 20 commission
4,500 machines 1.00 vend 10 commission
10-Year IRR
2,500 machines 1.00 vend 20 commission
4,500 machines 1.00 vend 20 commission
Sales Per Week (Units)
Note Assumes serving requirements increase with
volume other expenses same as detailed on prior
page also assumes no investment in trucks or
additional routes
66
Business Models for School Milk Vending
Processor Full-Service
Pros
Cons

• Processor retains all revenue
• Processor has full control over vender contents
  and quality control
• Good way for processor to build and maintain
  strong school relationships
• Provides immediate channel penetration for any
  new products
• Allows for distribution of processor other
  products
• Control over vend price
• Leverages existing routes and delivery equipment

• Very capital intensive machine purchase,
  vehicles, staff and systems
• Processors may have trouble getting machines
  placed in schools with third-party
  operator/vendor contracts
• Requires significant service/ maintenance
  infrastructure cost and effort intensive
• Small and frequent product drops compared to bulk
  delivery to vend operation
• Dramatic operational learning curve for
  vending-inexperienced processors

From processor point of view
67
Business Models for School Milk Vending
Processor Full-Service

• Based on assumptions derived from the vending
  test, dairy processors who purchase and service
  the school milk venders could realize a very
  robust IRR of up to 40
• This conservative estimate assumes a commission
  rate of 20 actual commission rate will depend
  on competitive circumstances

Theoretical School Milk Vending Financial Model
for Processor Full-Service
Note This model assumes the processor is
visiting the schools to deliver school milk also
assumes no investment in trucks or additional
routes
68
Business Models for School Milk Vending
Processor Full-Service

• Processor financials are fairly robust for
  full-service, even at relatively low sales volume
  
• BMC believes that few processor marketing
  investments would yield this level of IRR

Vending Economics for Processor Full Service With
Variable Vend Price, Volumes, Commission Rates,
Machine Costs
4,500 machines 1.25 vend 20 commission
2,500 machines 1.00 vend 20 commission
10-Year IRR
4,500 machines 1.00 vend 10 commission
4,500 machines 1.00 vend 20 commission
Sales Per Week (Units)
Note Assumes serving requirements increase with
volume other expenses same as detailed on prior
page also assumes no investment in trucks or
additional routes
69
Business Models for School Milk Vending
Processor/School Cooperative Service
Pros
Cons

• Processors and schools receive higher revenue
  than if third-party operators were involved for
  schools with higher than average beverage
  commissions
• Processor can amortize machine lease costs
  through product pricing
• Processor programs to assist school
  self-operation (e.g. financial help, service) can
  help build school relationships
• Fund-raising mechanism to penetrate student body
• Use of on-site labor
• Processor control over product and strong input
  over pricing

• Requires higher level of hand-holding compared
  to other options
• Processor loses control of product quality and
  may lose control of vending machine contents
• Run the risk that schools will not place required
  focus on maintaining/ stocking venders
• Requires processor to develop service support
  mechanism
• Risk of failing to recover machine cost through
  up-change
• May have to live with under-performing venders
• Processor is financially responsible for
  lease/purchase arrangements

From processor point of view
70
Business Models for School Milk Vending
Processor/School Cooperative Service

• The third model assumes a leasing arrangement
  where the dairy leases the machine and the school
  pays a surcharge for the milk to cover the lease
  expense
• This is a very attractive proposition for the
  schools especially, and for the processor if a
  large number of machines are placed

Theoretical School Milk Vending Financial Model
for Processor/School Cooperative Service

71
School Milk Vending Test Evaluating
Opportunities for School Milk Vending

• Contents
• Background
• Test Objectives and Methodology
• Key Learnings/Messages from the Test
• Test Results
• Business Models for School Milk Vending
• School Milk Vending Opportunity
• Issues and Hurdles
• Conclusions

72
School Milk Vending Opportunity

• The bulk of the school milk vending opportunity
  will come from large population middle and high
  schools per capita consumption is unlikely to
  be sufficient to support venders in very small
  schools
• However, use of the various operating models can
  effectively penetrate all but the smallest schools

Building the Vended School Milk
Opportunity Through Appropriate Operating Models
Large Middle and High Schools
Smaller Middle and High Schools
Total School Milk Vending Opportunity

• Vend Operators
• Processor Full-Service
• Cooperative Service
• School Self-Op

• Processor Full-Service
• Cooperative Service
• School Self-Op

73
School Milk Vending Opportunity

• Based on the PLs from participating processors,
  sales of 180 of milk per week can sustain a milk
  vending machine operated by a third-party
• At 0.20 units per week per capita, most schools
  with 900 or more students would be viable for a
  milk vendor serviced by a third-party operator

Assumption Used In Estimating School Milk
Vending Opportunity

• Estimated sales required for machine viability
• School size necessary to support minimum sales
• Number of middle/high schools with at least that
  population
• Less share of schools that significantly
  underperform average (15)
• Total student population in viable schools
• Average per capita consumption

180 units/week 900 students 8,800 schools 7,480
schools 12.5 million 0.20 units/week
74
School Milk Vending Opportunity

• Among these larger schools, BMC believes the milk
  industry could sell more than 90 million 16-ounce
  units of milk through venders per school year
• These calculations are based on very conservative
  estimates, including the assumption that there
  would be no sales during the summer weeks

Estimated Vended Milk Opportunity for Larger
Schools
12.5 million students 0.20 units/week 2.5 milli
on units/week 36 weeks/school year 90 million
units/school year 8 units/gallon 11.3 million
gallons/school year
_at_
x
_at_
75
School Milk Vending Opportunity

• Through processor full-service, cooperative
  service or school self-operation, BMC estimates
  schools with as few as 400 students could
  profitably sustain a milk vending machine
• Vending machines in smaller schools could also be
  sustained by inclusion of non-milk, higher margin
  beverages

Assumptions Used for Estimating School Milk
Vending Opportunities in Smaller Schools

• Estimated sales required for machine viability
• School size required to support minimum sales
• Number of middle/high schools with 400-900
  students
• Less share that will significantly underperform
  or could not be properly serviced (i.e. higher
  risk) (30)
• Total student population in viable schools
• Average per capita consumption

90 units/week 400 students 12,800 schools 8,960
schools 5.7 million 0.20 units/week
76
School Milk Vending Opportunity

• Among the smaller schools, there is an
  opportunity to sell another 41 million units
  through venders or 5 million gallons
• Absent participation of third-party operator,
  processor could retain all revenue and share with
  schools

Estimated Vended Milk Opportunity for Smaller
Schools
5.7 million students 0.20 units/week 1.14 milli
on units/week 36 weeks/school year 41.0 million
units/school year 8 units/gallon 5.1 million
gallons/school year
_at_
x
_at_
77
School Milk Vending Opportunity

• In total, roughly 130 million units of 16-ounce
  milk could be vended in schools nationally
• BMC believes this is a conservative estimate with
  upside potential based on additional machine
  placements, new flavors/products, etc., that
  would drive incremental sales and higher per
  capita consumption

Preliminary Estimate of Vended School Milk
Opportunity Per Year
Millions of Vended Milk Units
Larger Schools (gt900 Students)
Smaller Schools (400-900 Students)
Total
78
School Milk Vending Opportunity

• While 16.3 million gallons per year is a fairly
  small number relative to total milk volume, for
  schools it represents more than a 4 increase
• Nearly all volume is incremental
• This translates to about 20 million of gross
  profit, or more if processors self-operate

Potential Growth of Schools Milk via Milk Vending
in Schools
(4.0)
Millions of Gallons
Current School Milk Volume
Milk Vending Opportunity
Total School Milk Opportunity
79
School Milk Vending Opportunity

• More important, perhaps, is the life-long
  loyalty, enhanced image and higher usage rates
  that can be engendered by kids early exposure to
  value-added, great-tasting, well-presented milk
  options
• BMC estimates that each group of secondary school
  students exposed to school milk vending would
  consume an additional 1.3 billion gallons of milk
  over the remainder of their lives
• That increase would be sustainable, assuming
  continued effective vending efforts

Potential Increase in Milk Consumption
Over Remaining Lifespan of Each Secondary School
Cohort
1.3 billion
Billions of Gallons
Potential Incremental Consumption
Assumes 70 year lifespan
80
School Milk Vending Opportunity

• Beyond generating incremental milk sales, vending
  is also a great vehicle for brand building and
  image enhancement through the machines themselves
• A full-blown national school vending effort could
  generate as many as 10 billion consumer
  impressions without any additional cost
• An advertiser would pay 12 to 15 million for
  that number of impressions generated through a
  bus shelter program, for example

Vending as An Advertising Medium Annual
Impressions Generated by Full-Blown Opportunity
of Students 18.2 million Impressions/Student pe
r day 3 per year 540 Total Impressions 9.8
billion
Source Bozell New York
81
School Milk Vending Opportunity

• The high milk vending velocities generated in
  schools are not projectable to most potential
  vending locations
• Due to milks perishability, it requires high
  turn-over and high foot traffic areas, unlike
  soft drinks, which can be serviced far less
  frequently
• One high potential opportunity for milk is
  through existing or new refrigerated food
  venders, where milk could be a profit center for
  an otherwise relatively low margin business

• Product perishability/short shelf-life
• High cost, so little price flexibility
• Need for continual temperature control
• Likely not appropriate for all venues, e.g. hotel
  corridors, places where there are not other
  perishables being vended
• Growth through vend operator efforts alone will
  be quite slow nature of the business

Milk Vending Limitations
82
School Milk Vending Opportunity

• While it is very difficult to estimate the total
  opportunity for milk vending across all venues
  based on limited available data, BMC believes
  there is an untapped market that could generate
  incremental milk sales of 55 to 60 million
  gallons per year
• The best likely venues outside of schools are
  business/industry and colleges/universities and
  locations that require foodservice 24 hours/day
  (e.g. hospitals)
• The best vending program in the U.S. the Coke
  system, has 1.5 million vending machines with
  average sales of 825 million gallons per year

Potential Incremental Milk Vending Opportunity
Millions of Incremental Gallons/Year
of Venders
School Opportunity 16,500 16.3 New Out-of-School
Beverage Venders 30,000 19.5 (1) Existing
Refrigerated Food Venders 150,000 15.0 (2) Existin
g Milk Venders 84,000 5.3 (3) Total Incremental
Opportunity 280,000 56.1 million gallons
(1)Assume 100 milk units/week in mixed beverage
machines (2)Milk currently comprises 2-3 of
1.5 billion business assume increase to 10 of
sales (3)Current sales 53 million gallons
annually assume 10 increase as a result of
focused efforts of operators
83
School Milk Vending Test Evaluating
Opportunities for School Milk Vending

• Contents
• Background
• Test Objectives and Methodology
• Key Learnings/Messages from the Test
• Test Results
• Business Models for School Milk Vending
• School Milk Vending Opportunity
• Issues and Hurdles
• Conclusions

84
Issues and Hurdles

• A number of areas of concern related to school
  milk vending were brought to light by the test
• Many of these are faced and handled by vend
  operators in the normal course of business, but
  may be new/unexpected by processors

• Machine Selection
• Vender Placement
• Commission Rates
• Product Pricing
• Machine Contents
• Competitive Response
• Product Perishability/Spoilage
• Vandalism
• Summer Vender Utilization

Key School Vending Issues and Hurdles
85
Issues and Hurdles

• The vend test used the Dixie-Narco BevMax
  machine, primarily because of sponsorship of the
  test by Dixie-Narco
• Initially, a full-graphic backlit machine was
  planned for test use
• The BevMax machine, in addition to having greater
  capacity than essential for school venues, costs
  substantially more than smaller plastic front
  machines

Machine Selection Key Considerations

• Machine cost

Higher capital costs require greater vend
velocities to provide adequate returns milk
vending machine prices can vary by as much as
2,000

• Machine capacity

Surplus machine capacity compels consideration of
multi-product selling larger machines may
require less frequent service

• Graphic appeal

Plain front ma