Case Study: Gochujang

This scholarly article concerns the modification of a Korean food (Gochujang) to be adapted for the European market. This is a good example of marketing research applied to product adaptation for international markets. Note that the consumer preferences are determined prior to the introduction of this product.

Results

Across all six attributes assessed in this study, the sensory properties of the product (as described by its flavor characteristics), at 43%, was the most important attribute in driving preference. This attribute was significantly higher (p < 0.05) than the next most important attribute (brand at 24%), and reiterates the importance and focus on the product characteristics in determining choices within this market. Brand was significantly (p < 0.05) more important than the price attribute in driving preference (Table 2). Price (12%) and country of origin (11%) were the next most important drivers of preference, with packaging size and material at 6% and 5%, respectively, being relatively less important.

Table 2. Average Attribute Importance.

	All respondents 676%	Male 176%	Female 500%	18–34 y 229%	35–54 y 378%	55–65 y 69%
Flavor description	43	40	44	41	43	48
Brand	24	24	24	25	24	21
Price	12	13	12	14	11	10
Country of origin	11	12	10	11	11	10
Pack size	6	7	5	6	6	6
Pack material	5	5	5	4	5	5
Total	100	100	100	100	100	100

Estimation of the part worth utilities (Fig. 1, Fig. 2) for the total population (

$n = 676$ ) revealed that consumers had the least utility for the hot chili sauce element compared with those variants that were sweeter or were blended with other ingredients. This was encouraging as these latter descriptions approximated the gochujang-adapted based sauces being investigated in this study.

Fig. 1. Multiple comparison test results on zero centred utility values. Letters indicate significant differences (

$p < 0.05$ ) within each attribute for total population (

$n = 676$ ).

Fig. 2. Multiple comparison test results on zero centred utility values. Letters indicate significant differences (

$p < 0.05$ ) within each attribute for total population (

$n = 676$ ).

Across the total sample, respondents derived greater utility from the more familiar brand variants such as Blue Dragon in particular, and to a degree Encona and Wing Yip, than less well-known brands such as Linghams and Conimex. With no UK profile, it is not surprising that the Hot & Joy brand (the proposed gochujang-adapted sauce brand) offered the least utility.

In terms of pricing, the relative utility of different price points followed an expected distribution. From the price point GDP 1.75 onwards, the average utility value declined markedly.

Given the consumer recruitment criteria, that is, being receptive to ethnic and novel foods, perhaps it is not surprising that there is not a wide variation in utility values placed on the elements within the country of origin attribute. The emergence and popularity of Thai cuisine in the UK generally was clearly reflected in these data.

In terms of pack size, a larger pack was preferred over the 150-mL pack, and the glass container was preferred over the squeezable plastic option.

A number of market scenarios were simulated in the conjoint model and share of preference for each of these scenarios was estimated. Seven scenarios were completed. In each scenario, all the product specifications noted in Table 3 remained constant apart from price, which for the Hot & Joy product was tested at each price point (GBP 0.75, GBP 0.95, GBP 1,15, GBP 1.35, GBP 1.55, GBP 1.75 and GBP 1.95).

Table 3. Product Specification Scenarios.

	Brand	Pack size	Pack material	Country of origin	Flavor description	Price (GBP)
Test product	Hot & Joy	220 mL	Glass	Korea	Option 6	Variable
Competitor Product 1	Blue Dragon	220 mL	Glass	Thailand	Option 1	0.95
Competitor Product 2	Wing Yip	150 mL	Glass	Malaysia	Option 3	1.15
Competitor Product 3	Linghams	280 mL	Glass	Malaysia	Option 5	1.35

In these simulations, the share of preference for the Hot & Joy product (the proposed gochujang-adapted sauce brand) declined from around 24% at GBP 0.75 to 13% at GBP 1.95. The largest percentage declines in share of preference occur when price was increased to GBP 1.75 and GBP 1.95 (−13% and −16%, respectively). Moving price from GBP 0.95 to GBP 1.15 also recorded a significant decline in preference of nearly 10%. Respondents seemed less sensitive to price movements in the GBP 1.15–1.55 bracket.

Fig. 3 shows the preference shares between products at Hot & Joy's most competitive price. Given the utility values associated with a well-established UK brand such as Blue Dragon, it is not surprising that preference switched principally to Blue Dragon as the price for the Hot & Joy product became less attractive (Table 4).

Fig. 3. Predicted preference share for Hot & Joy product (price at 0.75p).

Table 4. Test of Proportions for Hot & Joy Product at Different Price Points (p values).

Price (GBP)		0.75	0.95	1.15	1.35	1.55	1.75	1.95
Share means		23.99	22.04	19.88	18.69	17.52	15.29	12.91
	SEs	1.31	1.26	1.21	1.19	1.16	1.09	1.01
23.99	1.31	0.500
22.04	1.26	0.065	0.500
19.88	1.21	0.001	0.040	0.500
18.69	1.19	0.000	0.003	0.161	0.500
17.52	1.16	0.000	0.000	0.023	0.160	0.500
15.29	1.09	0.000	0.000	0.000	0.001	0.024	0.500
12.91	1.01	0.000	0.000	0.000	0.000	0.000	0.012	0.500

SE, standard error.

From the conjoint study, the sensory-associated description (and hence the market positioning of the product) was the single most important attribute in driving preference. It was encouraging that the descriptions most aligned to the proposed gochujang-adapted sauce held appeal across the respondent set. However, even with a strong sensory positioning, it was evident that the brand was also a key attribute in driving preference. The finding reiterated the importance of linking the product to an established brand name in the UK market in order to optimize the chances of success within this competitive sector.

Home placement test

As noted in Table 5, both samples were relatively well received in terms of initial appearance and aroma, with eight out of 10 respondents expressing liking to some degree.

Table 5. Hedonic Summary Statistics.

Variable	Sample	Count			Mean	Percent
Variable	Sample	Like	Neither	Dislike	Mean	Like	Neither	Dislike
OverApp	Barbeque	45	5	7	6.7	78.9	8.8	12.3
OverApp	Chicken	48	6	4	6.9	82.8	10.3	6.9
Aroma	Barbeque	47	2	8	6.8	82.5	3.5	14.0
Aroma	Chicken	45	2	11	6.4	77.6	3.4	19.0
OFlav	Barbeque	39	4	14	6.1	68.4	7.0	24.6
OFlav	Chicken	40	2	16	6.1	69.0	3.4	27.6
Overall	Barbeque	39	4	14	6.1	68.4	7.0	24.6
Overall	Chicken	40	2	16	6.0	69.0	3.4	27.6

OverApp = Overall Appearance; Flav = Flavour; Overall = Overall Liking.

In terms of flavor, results were more polarized, with seven out of 10 respondents expressing liking to some degree for both samples, with the remainder expressing dislike to some degree.

Both samples were criticized to some degree for not being sweet enough, although the Just about right scores (7 out of 10 respondents) indicated reasonable consumer satisfaction with this attribute. However, it was clear that around half of the respondents considered the samples to be too hot to some degree.

The flavor results noted in this study reflected a similar pattern of response recorded in an earlier preliminary consumer study on a range of similar development sauces. This earlier study, which consisted of a pre-recruited central location test, evaluated five sauces among 100 consumers who represented a cross-section of the target market. The results from this earlier study demonstrated a similar split response towards the samples based on both hotness (with some consumers preferring the hotter versions and others preferring the less hot options) and flavor familiarity; in this case a barbecue flavor.

The results in respect of flavor translated through to the results for overall liking, with similar mean scores of 6.1 and 6.0 being recorded for the barbeque and chicken sauces, respectively. By individual usage occasion, the overall liking mean scores generally followed a similar pattern to the overall result. Both prior to and after testing the product on three occasions, consumers were asked to indicate their likelihood to purchase the product on a 5-point scale.

After sampling the product on three occasions, it was interesting to note that the proportion of those claiming that they would buy, broadly remained unchanged for both products. Fishers exact test returned a nonsignificant result in terms of changes in purchase propensity across the three conditions (Would buy; May/May not buy; Would not buy) for each sample (Table 6).

Table 6. Purchase Propensity Summary Statistics.

Variable	Sample	Count	Count	Count	Mean	Percent	Percent	Percent
Variable	Sample	Would buy	May/may not	Would not buy	Mean	Would buy	May/may not	Would not buy
Likely	Barbeque	25	22	10	3.3	43.9	38.6	17.5
Likely	Chicken	26	24	8	3.3	44.8	41.4	13.8
Likely	Barbeque	26	12	18	3.1	46.4	21.4	32.1
Likely	Chicken	28	15	15	3.3	48.3	25.9	25.9

Likely = Likely to buy.

In terms of all meal occasions, the barbecue and chicken sauces were used with evening dinner on 65% and 66% of all occasions, respectively. The next most popular occasion for both sauces was as part of an evening snack on 21% of occasions for the chicken sauce and 17% for the barbecue sauce. Across all occasions, both sauces were used most often as a dip (25% chicken and 22% barbecue), and a stir fry (19% for both sauces). Use within a stew/casserole, marinade or as accompaniment also received a significant number of mentions for both sauces.

For both sauces across all three occasions, the most popular main ingredients used were chicken (> 4 out of 10 of all occasions), vegetables (3 out of 10 occasions) and crisps/tortillas (2 out of 10 occasions).

Respondents were also asked to suggest the best ways for using the sauces. The chicken sauce was considered more as a flavor enhancer by most respondents and therefore best added to stews/casseroles (most often mentioned), as a marinade or as a stir fry, as reflected in a selection of consumer comments noted below. "I prefer to use this as a sauce for cooking, not as a table top sauce or as a dip as it has too much seasoning" (Respondent 16). "As a marinade but remember it is very strong" (Respondent 38). " It was lovely in a stir fry" (Respondent 5). "Was hot and spicy but it was very good. First class" (Respondent 42). Similar comments were also made with respect to the barbecue sauce, although its function was extended to include use as a dip or with a barbecue. "Added to a stir fry, it was lovely" (Respondent 31). "As a marinade the flavor comes through very well" (Respondent 35). "In soups or stews, it's too hot for general use" (Respondent 37). "Excellent. But it could come in a larger size" (Respondent 16).

The comments from respondents indicated that the products were aligned to consumers' typical consumption behavior. From the home placement study, both sauces received similar and reasonable levels of acceptability and consumers' perceived product uses were aligned with intended applications.