Establishing brand image

There are a number of limitations connected to establishing Wood Wood’s brand image through the use of the BCM approach. The following section discusses research design limitations in terms of the BCM sample as well as the mapping and aggregation stage.

7.2.1 BCM SAMPLE

The BCM sample size is fairly small, including 38 respondents in the elicitation stage and 21 participants in the map-ping stage. This poses a limitation, as the sample size might be too small to uncover all nuances of the brand image in the consensus map, as it is unlikely that a sample of this size will provide sufficient generalizability to establish the Wood Wood brand image in relation to the gap. However, the gap is established for this particular sample, which can provide an indication of the full picture.

Another limitation in terms of sample is that respondents were friends and acquaintances. This suggests that certain

demographic factors such as age, education and income level may be somewhat similar, providing little insights to the entirety of Wood Wood’s consumer population. The vast majority of BCM participants are within the age group 21-30, which decreases the generalizability of the Wood Wood brand image. However, this age group is Wood Wood’s main target group and it could be added that if the brand does not resonate with this age group, it does not resonate with any age group.

Another limitation in terms of BCM participants is the level of brand familiarity in the sample. John et al. (2006) ar-gues that participants level of brand expertise is a defining factor for the complexity of BCMs. Experts have a larger amount of brand associations and association links, stronger association links, more hierarchical links (third-order etc.), as opposed to brand novices. Brand familiarity is utilized as the primary factor for participant selection in the mapping stage. The brand familiarity measure includes four groups: 1) I know of Wood Wood, but have never pur-chased their products, 2) I have purpur-chased Wood Wood products 1-4 times over the past two years, 3) I have purpur-chased Wood Wood products 5+ times over the past two years, and 4) I have purchased other brands through Wood Wood, but not their own products over the past two years. The elicitation stage uncovered that only one respondent belonged to the third category where brand experts are placed. The majority of participants belonged to group 2, followed by participants belonging to group 1. A limitation is therefore, that the brand consensus map is based on participants’

low-medium brand knowledge. This could have been avoided by involving more brand experts in the BCM sample and the consensus map could potentially have included more complex, unique, strong and favorable associations, which could potentially have affected the brand identity and image gap in terms of increasing alignment.

7.2.2 SALIENT ASSOCIATIONS FOR MAPPING STAGE

In terms of associations utilized in the mapping stage, a few specific limitations are relevant to address. In order for an association to be considered a salient association that would proceed to the mapping stage the frequency of that association in the elicitation stage had meet a threshold of 8%. While acknowledging that this threshold is low, none of the salient was able to meet a threshold of 50% that John et al. (2006) employs. However, increasing the threshold, would mean fewer associations would be included in the mapping stage, which would potentially ruin the value of the BCMs. Therefore, the threshold was set according to the number of salient associations in the mapping stage, so that this amount was consistent with John et al.’s (2006) pool of associations. John et al. (2016) included 25 salient associations in the mapping stage and this research study included 26. The low degree of brand knowledge among BCM participants, could therefore, be expressed through the low reoccurrence of associations in the elicitation stage.

John et al. (2006) argues that survey respondents utilized in the elicitation stage needs to be identical to participants in the mapping stage. However, the salient associations were based on the total amount of contributed associations

in the elicitation stage, while the mapping stage only included half of the respondents from the elicitation stage.

This could explain why the association Cool did not appear in the consensus map, despite it being the fourth most frequently mentioned association in the elicitation stage. Moreover, it could potentially have been mentioned by the other half of respondents that was excluded from the mapping stage. Furthermore, John et al.’s first criteria which states that the same sample have to be utilized in both the elicitation and the mapping stage, is inconsistent with the option of utilizing existing market research to elicit salient consumer associations. This option neglects the relation between the salient associations and participants in the mapping stage. However, the two methods in the first crite-ria are deemed equally reliable. Therefore, in order to establish fairness between the two methods the first critecrite-ria should not be included in the BCM approach.

Another limitation, which potentially can influence the brand identity and image gap, is that strictly adhering to the BCM approach might not be beneficial. John et al. (2006) even break their own BCM approach in a test study by allowing the company to add associations to the pool of salient associations in the mapping stage. By allowing Wood Wood the same opportunity to add associations to the pool of associations in the mapping stage, more alignment between the brand identity and brand image may have occurred, since all identity traits would have been included in the pool of associations. However, this would contradict the notion that only consumers salient brand associations are relevant in establishing the brand image, thus increasing the risk of biasness occurring in the mapping stage.

7.2.3 DEVIATION IN MAPPING STAGE

The mapping stage followed the process of John et al. (2006), but did deviate in one aspect. An issue with utilizing the original BCM approach is that loops can occur between associations in BCMs. This becomes an issue as the causality between associations are impossible to determine, meaning which associations triggers other associations in the net-work and a vice versa effect does not necessarily exist. To decrease the complexity and increase the interpretability of the BCM approach, the following rule were added to the mapping stage: Associations can not be linked horizontally, meaning that a second-order association can only be connected to a first-order or a third-order association and not another second-order association. In addition, two individuals conducted the mapping stage separately. This could have impacted BCM results in terms of how the method was presented to participants. This poses a limitation, despite the BCM mapping stage being tested on a test participant, prior to conducting the actual BCMs.

7.2.4 THE ORIGINAL AGGREGATION RULES

One limitation in terms of the brand identity and brand image gap is that Böger et al.’s (2017) aggregation rules may be too strict. On the other hand, following the aggregation rules of John et al. (2006) could potentially have resulted

in the consensus map including a larger amount of salient associations. However, associations would have to meet the inflection point criteria in order to be added to the consensus map. However, if the three associations Cool, Youthful and Sporty met the inflection point criteria, Wood Wood’s brand identity and brand image gap would decrease in size, allowing alignment of five traits and associations: Streetwear, Understated, Cool, Youthful and Sporty, as opposed to the two in this research study. Using John et al.’s (2017) original aggregation rules could potentially have impacted the result of the research study. However, it would require a substantial amount of work to test this implication.