valuing food attributes: experimental auctions and sensory analysis

37
Valuing Food Attributes: Experimental Auctions and Sensory Analysis Marin Bozic - University of Minnesota Presented at the Food Science Department North Carolina State University, June 11, 2012

Upload: jeri

Post on 25-Feb-2016

42 views

Category:

Documents


1 download

DESCRIPTION

Valuing Food Attributes: Experimental Auctions and Sensory Analysis. Marin Bozic - University of Minnesota Presented at the Food Science Department North Carolina State University, June 11, 2012. Travel back in time to 1992…. 1992: Will consumers accept rBST ?. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Valuing Food Attributes: Experimental Auctions and

Sensory Analysis

Marin Bozic - University of Minnesota

Presented at the Food Science Department North Carolina State University, June 11, 2012

Page 2: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Travel back in time to 1992….

2

Page 3: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

• Smith and Warland (1992) – survey of 11 consumer studies: an average of 57% of respondents views rBST negatively

• Hoban and Kendall (1992) – national telephone survey of 1200 adults: fewer than half accepted biotechnology for increased milk production

• Based on surveys, analysts expected 4-20% decline in consumption of fluid milk upon rBST introduction

• But… should we trust hypothetical surveys?

1992: Will consumers accept rBST?

Page 4: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Compulsion to Drink Experiment• A experiment with

tangible consequences…• Each subject was given

a glass of “rBST milk”• In order to be

compensated, they either had to bid to exchange for “rBST-free milk”, or drink the glass they had

Page 5: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Experiment run from coast to coast

Page 6: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Group dynamics and the effect of information

• 10 bidding rounds where initially conducted.

• After 10th round, a neutral description of rBST was read aloud.

• At round 20, 70% indicated WTP the product at no or small discount

Page 7: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

• Initial consumption reaction to rBST introduction was minimal

• Product differentiation (“rBST free”) emerged to cater to niche markets

• Study by Foltz (2005) – market share for rBST-free in single digits• Today – a dramatic reversal in fluid milk

market

What happened in the end?

7

Page 8: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

1. Toolbox Revisited: Overview of methods food science and marketing/economics uses to understand consumer preferences

2. The Promise of Experimental Auctions (EA)3. Auctions in Action: what have we learned

about dairy products4. New Product Development: When should food

scientists and marketing team work together5. EA with Fluid Milk: Outline of my research

Plan for today

Page 9: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Why we need to understand consumer behavior

1. New product development and pricing strategy

2. Advising public policy

3. Guiding and valuing fundamental research

Page 10: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Revealed vs. Stated Preferences

Page 11: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

• Self-stated WTP

• Contingent Valuation

• Conjoint analysis

• Experimental auctions

Evaluating willingness-to-pay for new goods

Page 12: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

• Do consumers have an incentive to reveal their true preferences?– Lack of interest in the experiment– Desirability bias (social reputation, experimenter

effect)– Strategic behavior

• Solution: Separate what people say and what they must pay.

Would you pay what you say?

Page 13: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

• Consumer choices are consequential. Real money is exchanged for real goods.

• EA are designed to incentivize subjects to reveal their true preferences

• Experimental design penalizes strategic behavior

• Key element: Final price a person must pay is not dependent on what the person indicated as their WTP

The Promise of Experimental Auctions

Page 14: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

A lottery approach (Becker-DeGroot-Marschak Mechanism)

BDM Mechanism has three defining steps:

1) Subjects submit their statement on maximum WTP for a good

2) A price is randomly drawn from a previously explained uniform distribution and announced to experiment participants.

3) All subjects that submitted WTP that turned out to be higher than the randomly drawn price must buy the product, and pay the announced price, not the WTP they indicated.

Page 15: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

A lottery approach (Becker-DeGroot-Marschak Mechanism)BDM Mechanism is incentive compatible.

- If they submit WTP that is too high, they might have to pay for the product more than they value it (if the random price turns out to be higher then their true WTP, but lower then their expressed WTP.)

- If they submit WTP that is lower than their true value they might miss the opportunity to buy the product that was good value-for-money (if the price is drawn lower than their true value, but higher than submitted WTP)

Page 16: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Second-price Auction (Vickrey)Basic Vickrey Second-price auction steps:

1) Participants are assembled in a group, and asked to submit sealed bids for one product.

2) Only the highest bidder gets to buy the product, and he or she must pay the price equal to second-highest bid.

As in BDM mechanism, it’s not possible to “game” the system by strategic bidding.

Page 17: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Variation on the theme• English auction• N-th price auctions• Random n-th price auctions• Multiple rounds with increase in

information/experience with goods, with only one round binding

• BDM/Vickrey with “endowment/upgrade” setup• CLT vs point-of-purchase• Repeated treatments

Page 18: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

What can we use auction bids for?

• Simulating market shares• Simulating own-price and cross-price

elasticities• Identify product attributes that generate

higher bids• Identify consumer classes (factor analysis)

Page 19: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Using Auction Bids: Demand Curves

$0

$2

$4

$6

$8

$10

$12

$14

0% 20% 40% 60% 80% 100%

Pric

e (W

illin

gnes

s-to

-Pay

)

Quantity Demanded (Percent age of Consumers)

GenericGuaranteed TenderNaturalChoiceCertified Angus Beef

Page 20: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

What have we learned about dairy products from experimental auctions?

• Organic is a bag of attributes (Bernard & Bernard, 2009)– Bids for rBST-free +

antibiotics-free = bid for organic

– Is there a point in paying for organic feed? “All-natural” products may have a big marketing opportunity.

Page 21: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

What have we learned about dairy products from experimental auctions?

• Effects of new product introduction– Brand switching– Cannibalization– Category expansion

Stigmatization (Kanter et al, 2009)New products redefine the existing ones.

Page 22: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

What have we learned about dairy products from experimental auctions?

• Sensory acceptability mitigates WTP for animal welfare in yogurts (Italy)– Consumers may be forming

sensory expectations based on production-method labels

– Products promising superior production-methods may not succeed if not matched with superior taste

(Napolitano, 2008)

Page 23: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

• Home-grown vs. “affiliated” values• Does not mimic the consumer purchase

process• Expensive, labor intensive, cumbersome• Off-margin bidders quickly lose interest in

Vickrey auctions• Revealed preference for pricing research –

conjoint analysis still by far the most widely used technique

Criticisms of Experimental Auctions

Page 24: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

New Product Development Process

Page 25: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

How should food scientists and economists collaborate?

• It boils down to this: is anything lost when price-point analysis is separated in product development from sensory analysis? – Cost-impact of important ingredients

• How does ‘liking’ score translate to willingness-to-pay in auctions?

• Can check-all-that-apply (CATA) approach be used to infer determinants of WTP, rather than just acceptability score?

• Does product “story” change how sensory attributes are scored by consumers? Can “story” carry higher WTP if there is no “signature” on sensory attributes?

Page 26: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

UMN-ISU Experimental Auctions with Fluid Milk

Sensory aspect: Do consumers prefer flavor of milk from cartons or plastic jugs?

Economic aspect: Are consumers willing to pay more for a “certified fresh taste” milk? I.e. non-oxidx

Page 27: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Experiment DesignThree-stage second-price auction1) “Blind WTP”

- package, milkfat (2x2)2) Information phase

- information on oxidation read- offer to buy “certified fresh”milk in both type of packaging

3) Sensory phase-forced-choice assessment- auction with tastes revealed

- 2 and 3 interchanged

Page 28: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

I. Pre-information auction

First, each of you has been given a bid sheet in your packet (PINK). On this sheet you will, in a moment, write the most you are willing to pay to for each of the following:

a) skim milk, in paperboard cartonb) skim milk, in plastic jugc) Reduced fat milk, in paperboard cartond) Reduced fat milk, in plastic jug

Page 29: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Subjects split to two treatments

• Scenario A: Information-first experience

• Scenario B: Sensory-first experience

Page 30: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Oxidation information

Oxidation is a process that causes milk to develop flavor defects. Light initiate a chemical reaction in milk that modifies specific components of proteins and fats which results in characteristic off-flavors. Exposure to light for as little as 10 minutes is sufficient enough to cause the defect. The closer the milk is to the light source the quicker the off-flavor develops. Not only is flavor affected, but certain vitamins riboflavin and vitamin A are susceptible to light induced degradation.  Oxidation can be also be induced by metals like copper or iron. The off-flavor has been described as "cardboard" or mouth-drying. Milk packaged in paperboard does not allow this process to occur.  With that information in mind, we ask you to place your bids on the 4 packages in front of you…

Page 31: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Sensory test

Skim milk, from paperboard carton 119Skim milk, from plastic jug 287Reduced fat milk, from paperboard carton 078Reduced fat milk, from plastic jug 113

Subjects are given two additional samples to try, that would be introduced to them as “CERTIFIED FRESH” skim and reduced fat milks. In reality those would be just skim and reduced fat milks from cartons, but that information would not be revealed to subjects. [this part is done after post-sensory auction for sensory-first treatment; and before post-sensory auction for information-first treatment]

Page 32: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Scenario A: II. Post-information instruction

First, each of you has been given a bid sheet in your packet (PINK). On this sheet you will, in a moment, write the most you are willing to pay to for each of the following:

a) skim milk, in paperboard carton, regular b) skim milk, in plastic jug, regular c) reduced fat milk, in paperboard carton, regular d) reduced fat milk, in plastic jug, regular e) skim milk, in paperboard carton, certified fresh taste f) skim milk, in plastic jug, certified fresh taste g) reduced fat milk, in paperboard carton, certified fresh

tasteh) reduced fat milk, in plastic jug, certified fresh taste.

Page 33: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Scenario A: III. Post-sensory auction

First, each of you has been given a bid sheet in your packet (PINK). On this sheet you will, in a moment, write the most you are willing to pay to for each of the following:

a) skim milk, in paperboard carton, regular b) skim milk, in plastic jug, regular c) reduced fat milk, in paperboard carton, regular d) reduced fat milk, in plastic jug, regular e) skim milk, in paperboard carton, certified fresh taste f) skim milk, in plastic jug, certified fresh taste g) reduced fat milk, in paperboard carton, certified fresh

tasteh) reduced fat milk, in plastic jug, certified fresh taste.

Page 34: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Scenario B: II. Post-sensory auction

First, each of you has been given a bid sheet in your packet (PINK). On this sheet you will, in a moment, write the most you are willing to pay to for each of the following:

a) Skim milk, paperboard carton, tastes like sample 119b) Skim milk, paperboard carton, tastes like sample 287c) Skim milk, plastic jug, tastes like sample 119d) Skim milk, plastic jug, tastes like sample 287e) Reduced fat milk, paperboard carton, tastes like sample 078f) Reduced fat milk, paperboard carton tastes like sample 113g) Reduced fat milk, plastic jug, tastes like sample 078h) Reduced fat milk, plastic jug, tastes like s sample 113

Page 35: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Scenario B: III. Post-information instruction

First, each of you has been given a bid sheet in your packet (PINK). On this sheet you will, in a moment, write the most you are willing to pay to for each of the following:

a) skim milk, in paperboard carton, regular b) skim milk, in plastic jug, regular c) Reduced fat milk, in paperboard carton, regular d) Reduced fat milk, in plastic jug, regular e) skim milk, in paperboard carton, certified fresh taste f) skim milk, in plastic jug, certified fresh taste g) Reduced fat milk, in paperboard carton, certified fresh

tasteh) Reduced fat milk, in plastic jug, certified fresh taste.

Page 36: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Analysis

• Do people indicate preference for a milk taste from a particular packaging

• Are consumers willing to pay a premium for “certified fresh taste”?

• What are the drivers of WTP?

• What is the return to R&D for jug-like packaging that would completely eliminate oxidation process?

Page 37: Valuing Food Attributes: Experimental Auctions and Sensory Analysis

Valuing Food Attributes: Experimental Auctions and Sensory Analysis

presented at the North Carolina State University, invited seminar in the Food Science Department, June 11, 2012

Dr. Marin [email protected](612) 624-4746Department of Applied EconomicsUniversity of Minnesota-Twin Cities317c Ruttan Hall1994 Buford AvenueSt Paul, MN 55108

37