How the Cookie Crumbles: The Case of Gluten-free Cookies

How the Cookie Crumbles:

The Case of Gluten-free Cookies

The Back Story

• Angela Ichwan and her husband Hermanto Hidajat are founders of Arico Natural Foods Company.

• With $500,000 from 14 angel investors, Arico started production in March 2005. Arico set out to produce products that are safe and tasty alternatives for individuals who have a restricted diet.

The Question

• How do we best market our products to meet the specific needs of our traditional customers (persons with celiac disease and families with autistic children) and reach new non-traditional customers?

The Darling of Natural Products

Gluten-Free Products• have no (or very little) gluten, a protein in

wheat and closely related grains.• are the fastest growing segment of the

natural products channel.• increased in sales to $921 million in 2008,

a 16% increase.

The Cookie Market

• The market has been in decline since 2001.• However, the “Good-For-You” segment of the

cookie market is growing.• Low-fat cookies increased by 17.7%.• Gluten-free cookies increased by 30.2%.

• Premium cookies are also on the rise.

The Crumble

• Gluten-free cookies are hard to make well.• Gluten-free cookies do not crumble in the

mouth as gluten-containing cookies do.• Removing gluten requires finding other

flours from beans, rice and the “ancient grains” like amaranth, millet, quinoa, sorghum and teff which do not have gluten.

The Three Consumers

• The Celiac Disease Consumer• The Self- Diagnosed Consumer• The Autistic Consumer

The Model

• What attributes encourage consumers to select the Arico cookie bar over other products?

• In the fall of 2007 and spring of 2008, we conducted a marketing experiment based on a random utility model.

The Model

• The probability of the particular ranking is

=

• The population shares the same utility, we can parameterize the model as

The Model

• where is the utility of choice j, is a set for product attributes, is a set of unknown parameters, and is the random error.

• The probability of a particular ranking is written as…

+

The Model

• where represents the set of attributes of the alternative ranked . This model presentation draws heavily from Calfee, Winston and Stempski (2001).

Pr [𝑈 (𝑟1 )>𝑈 (𝑟 2 )>⋯𝑈 (𝑟 𝑗 ) ]=∏h=1

𝐽 −1 exp 𝛽 ′ 𝑥 (𝑟h )

∑𝑚=h

𝐽

exp 𝛽 ′ 𝑥 (𝑟𝑚 )

The Model

• We asked the participants to rank the nine products from most favorite (1) to least favorite (9).

• The products were different types of chocolate chip cookies with one outside product, an apple.

The Model

• We ran the experiment at Krogers, Day Spring: The Natural Shoppe, and Auburn University campus,

• We had four treatments: no additional information, information on gluten-free diets, a taste test of the Arico cookie bar, and both additional information and a taste test.

The Model

• Each participant had a coupon sheet of the products, which highlighted attributes of the products.

• We also provided posters of the nutritional labels of each product.

• We used an incentive compatibility mechanism ascertain true(r) preferences.

The ModelIndependent Variables Model 1 Model 2 Model 3

Base Model Store EffectStore and All Information

Price -0.17***(0.032)

-0.16***(0.036)

-0.16***(0.033)

Fat -0.021*(0.011)

-0.015(0.013)

-0.022*(0.012)

Gluten-Free 0.23***(0.088)

0.16(0.10)

0.18**(0.093)

Organic -0.20*(0.10)

-0.39***(0.12)

-0.30***(0.11)

Partially Hydrogenated Oils -0.39***(0.11)

-0.32***(-0.12)

-0.35***(0.11)

Effect*Price -0.054(0.077)

-0.12(0.11)

Effect*Fat -0.018(0.028)

0.031(0.039)

Effect*Gluten-Free 0.41*(0.22)

0.77**(0.31)

Effect*Organic 0.90***(0.25)

1.13***(0.36)

Effect*Partially Hydrogenated Oils -0.28(0.26)

-0.40(0.39)

The ModelIndependent Variables Model 1 Model 2 Model 3

Base Model Store EffectStore and All Information

Price -0.17***(0.032)

-0.16***(0.036)

-0.16***(0.033)

Fat -0.021*(0.011)

-0.015(0.013)

-0.022*(0.012)

Gluten-Free 0.23***(0.088)

0.16(0.10)

0.18**(0.093)

Organic -0.20*(0.10)

-0.39***(0.12)

-0.30***(0.11)

Partially Hydrogenated Oils -0.39***(0.11)

-0.32***(-0.12)

-0.35***(0.11)

Effect*Price -0.054(0.077)

-0.12(0.11)

Effect*Fat -0.018(0.028)

0.031(0.039)

Effect*Gluten-Free 0.41*(0.22)

0.77**(0.31)

Effect*Organic 0.90***(0.25)

1.13***(0.36)

Effect*Partially Hydrogenated Oils -0.28(0.26)

-0.40(0.39)

The ModelIndependent Variables Model 1 Model 2 Model 3

Base Model Store EffectStore and All Information

Price -0.17***(0.032)

-0.16***(0.036)

-0.16***(0.033)

Fat -0.021*(0.011)

-0.015(0.013)

-0.022*(0.012)

Gluten-Free 0.23***(0.088)

0.16(0.10)

0.18**(0.093)

Organic -0.20*(0.10)

-0.39***(0.12)

-0.30***(0.11)

Partially Hydrogenated Oils -0.39***(0.11)

-0.32***(-0.12)

-0.35***(0.11)

Effect*Price -0.054(0.077)

-0.12(0.11)

Effect*Fat -0.018(0.028)

0.031(0.039)

Effect*Gluten-Free 0.41*(0.22)

0.77**(0.31)

Effect*Organic 0.90***(0.25)

1.13***(0.36)

Effect*Partially Hydrogenated Oils -0.28(0.26)

-0.40(0.39)

The Probability of First Rank

Product BaseDay Spring

EffectDay Spring and All Information Effect

Kroger and All Information Effect

Apple 15.33% 13.82% 9.67% 15.01%

Arico Chocolate Chip 11.85% 20.77% 24.37% 12.77%

Chips Ahoy Chocolate Chip

11.47% 6.09% 4.42% 10.21%

Chips Ahoy Chocolate Chip Reduced Fat

10.29% 5.48% 3.12% 9.42%

Kroger’s Brand Chocolate Chip

6.40% 2.97% 1.58% 6.11%

Lärabar Chocolate 9.46% 14.68% 24.43% 10.07%Newman’s Own Chocolate Chip

13.09% 17.39% 17.74% 13.83%

Odwalla Bar Chocolate Chip and Peanuts

8.69% 10.46% 8.78% 9.62%

Pepperidge Farm Chocolate Chip

13.43% 8.35% 5.88% 12.95%

The Case Questions

• What does the ranking exercise tell us about the interest in Arico cookies and the attributes customers want when considering snack foods.

• What are the challenges of this choice experiment?

The Case Questions

• Which customers should Arico pursue?• If these consumers are not the traditional

consumers how do you meet the diverging needs of these diverse consumers?