Вы находитесь на странице: 1из 8

Sampel

Penelitian ini didasarkan pada data sekunder. Sampel terdiri dari pelanggan dari jaringan resmi dua merek:
merek premium dan merek ceruk harga menengah. Kedua merek, yang diminta untuk tidak disebutkan
namanya dalam artikel ini, memberikan peringkat dari survei kepuasan pelanggan mereka dan catatan
mereka tentang kunjungan pelanggan ke bengkel mereka dalam 18 bulan sejak survei terakhir. The data for
both brands were gathered using an equivalent procedure and are therefore comparable:
• Customer satisfaction surveys: 6,696 customer surveys for the medium- price brand and 10,092 surveys
for the premium brand. The surveys were conducted by telephone with customers who had visited an
official service workshop during the previous month.
• Records of individual customer visits to official service workshops: 87,612 records in the case of the
medium-price brand, and 1,218,981 records in the case of the premium brand.
The 18-month term was defined considering the average service demand interval in Spain, and
therefore covered a sufficient period of time to expect customers to come back. Customers not coming back
in such a period were considered disloyal.
Research Design
In the first step, a basic analysis of the samples of both brands was realized. Then two multivariate
techniques for the analysis of dependencies were ap- plied: MCA and AID (Santesmases, 2009). Finally, a
PLS-SEM was developed for the premium brand (Henseler, Ringle, & Sinkovics, 2009; Ringle, Wende, &
Will, 2010). Previous to this, to make this possible, principal-component factor analysis was applied to the
service attributes of the after sales service (Santesmases, 2009). The resulting principal components of
service quality were then used for the PLS-SEM model. Three statistical software packages were used for
this analysis: DYANE (Santesmases, 2009), SmartPLS (Ringle, Wende, & Will, 2005), and SPSS (IBM,
2012).
Measurement of Variables
The following variables were used for MCA and AID:
1. Customer satisfaction: a global assessment of satisfaction on a Likert scale
(1 = very unsatisfied, 5 = very satisfied)
A. Garcıa Gonz á lez 251
2. Loyalty toward the workshop: a customer was considered loyal if he or she made one or more visits to
the same workshop after the satisfaction survey 3. Loyalty toward the brand: a customer was considered
loyal if he or she made one or more visits to the manufacturer's retail network after the satisfaction survey
4. Brand: premium or medium-price niche brand 5. Warranty validity: the warranty was valid if the
vehicle was 2 years old or
less at the time of the second visit 6. Type of service: service check or maintenance, mechanical repair,
body
shop repair or others. 7. Customer age in years 8. Customer gender: male or female
The following service attributes or factors were used for the factor anal- ysis of service quality of the
premium brand (all factors were measured using a Likert scale):
1. Workshop's ability to offer a service appointment in a suitable date for
the customer 2. Effort to adapt the service appointment to the customer's requirements
(date and time) 3. Waiting time at the workshop reception 4. Workshop opening days and times 5.
Convenience of workshop location (distance, accessibility, etc.) 6. Parking facilities at the workshop 7.
Comfort of the workshop's waiting area 8. Cleanliness of the workshop's waiting area 9. Equipment in the
waiting area (beverages, magazines, television, Internet
connection, etc.) 10. Guidance by workshop staff on service needs and requirements 11. Friendliness
and behaviour of the workshop staff 12. Credibility of the workshop staff (service advisor) 13. Expertise
and knowledge of the service advisor at the workshop recep-
tion 14. Attention to customer questions 15. Clarification of questions or needs 16. Understanding of
issues related to the vehicle 17. Service fulfilment (time, tasks, etc.) 18. Overall time required to finish the
service 19. Ability to finish the service and deliver the vehicle when promised 20. Explanation of services
done on the vehicle 21. Attention and care when explaining the cost of each service
252 Automotive After Sales Service and Loyalty in Spain
FIGURE 2 Partial Least Squares Structural Equation Model Path Model.
22. Payment process (ease and speed of process) 23. Vehicle cleanliness and overall aspect when
delivered 24. Time required for paying for and collecting the vehicle 25. Ability to identify vehicle
service requirements (diagnosis) 26. Quality of service done on the vehicle (on the technical side) 27.
Care in servicing the vehicle as required by the customer 28. Spare parts availability 29. Reliability and
trustworthiness of the technical staff
Figure 2 represents the path model and variables used for the PLS-SEM analysis. Latent variables and
observed variables or indicators included the following:
1. Service quality: construct measured based on the main components of the service attributes or factors
of the premium brand, obtained through factor analysis 2. Customer satisfaction: measured through the
customer satisfaction survey 3. Type of service: high competitive services (service check, maintenance,
or body shop repair) or medium-low competitive service (rest) 4. Satisfaction with the vehicle:
measured through the customer satisfaction
survey
A. Garcıa Gonz á lez 253
5. Customer characteristics: construct made up of two observed variables,
gender and age 6. Warranty validity: measured from the registration date and the last repair
date 7. Loyalty to the service workshop: measured as stated previously 8. Loyalty to the brand:
measured as stated previously
RESULTS OF THE RESEARCH
The basic analysis is partially in line with the previous literature, showing higher loyalty and satisfaction
rates for t he premium brand. In contrast with other studies and markets, in which progressive or saddle-
shaped forms are more common (Meyer, 2010), the cross-tabulation shows a regressive- shaped positive
relation between customer satisfaction and loyalty. In any case, the fact is that satisfied Spanish customers
also defect: Only 59% of customers returned to the workshop, although 77% of them reported being
satisfied or very satisfied on the survey. Regarding the results of MCA and AID:
1. Loyalty to the brand: The relatively low levels of explained variance (7.47% for the whole sample,
8.96% for the medium-price brand, and 4.69% for the premium brand) show that other variables not
included in this study have a greater impact on customer loyalty. Of those included, warranty va- lidity is
the most important (4.66% explained variance for the total sample, 7.07% for the medium-price brand,
and 4.01% for the premium brand; it is the first variable in the AID sample partition for all cases).
Consequently, contractual retention is the most effective way to reinforce customer loy- alty. The second
variable for the whole sample is the type of brand (1.51% explained variance), with higher loyalty for the
premium brand customers. Satisfaction is just the third variable in importance for the total sample, with
0.92% explained variance. For the medium-price brand, it ranks fourth with 0.60% explained variance,
after 0.89% for the service type. Regarding the premium brand, satisfaction explains just 0.31% of the
variance and does not turn up at the AID partition. The resting variables have even less of an impact on
loyalty. 2. Loyalty to the workshop: The results are slightly lower than those toward the brand (6.26%
explained variance for the whole sample, 8.26% for the medium-price brand, and 4.00% for the premium
brand). The vari- able with the greatest impact is again warranty validity (3.72% explained variance for
the total sample, 5.67% for the medium-price brand, and 3.17% for the premium brand; it is the first
variable in the AID sample partition for all cases), followed by the type of brand (1.02% explained
variance for the total sample and the second variable in the AID sample
254 Automotive After Sales Service and Loyalty in Spain
partition). Customer satisfaction is the third variable in terms of impact on loyalty for the total sample
(1.20% explained variance). For the medium- price brand, it ranks fourth with 1.01% explained variance,
after 1.24% for the service type. Regarding the premium brand, satisfaction explains just 0.60% of the
variance and ranks second for the AID partition. Enhancing service quality is therefore more beneficial for
the retailers than for the manufacturers.
MCA and AID only applied to 14 out of the 21 hypotheses, making further analysis necessary. To
cover this gap, the 29 service components included in the survey of the premium brand were analysed. I
decided to focus on this brand because its survey was more detailed than the one for the medium-price
brand. The main components of service quality, which reflect the underlying dimensions of service quality,
were defined via factor analysis. These components were then used as indicators for service quality in the
PLS-SEM.
The analysis with four main components explains 62% of the variance after varimax rotation
(Santesmases, 2009). Bartlett's test is significant at p > .0001 (the determinant of the correlation matrix is
0, χ2 with 406 df = 138,750.20, p = .00000):
• Factor 1: Service care to customer—friendliness, expert advice, and image of professionalism: Explains
20.26% of the variance and is linked to Service Components 10, 11, 12, 13, 14, 15, and 16.
• Factor 2: Workshop facilities: Explains 12.22% of the variance and is mainly linked to Service
Components 5, 6, 7, 8, and 9.
• Factor 3: Service availability and waiting times: Explains 10.65% of the variance and is mainly linked to
Service Components 1, 2, 3, and 4.
• Factor 4: Service finalization—end of service and vehicle collection: Ex- plains 19.00% of the variance
and is linked to Service Components 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, and 29.
Customer Loyalty Path Model for the Premium Brand
The assessment of the path model was done in two steps, according to Henseler et al. (2009) and Hair (2012,
2013). The first step was to assess the PLS-SEM measurement (external) models, both reflective and
formative, and the second was to assess the structural (internal) model (see Figure 2).
1. Assessment of the External Model
a) Reflective Model
A. Garcıa Gonz á lez 255
i. Internal consistency: The model is consistent if the composite re- liability (ρc) is 0.6 or higher. This
value scores 1 for the three variables, making the model internally consistent. ii. Indicator reliability: The
indicators are assessed as reliable if the absolute standardized component outer loadings are higher than
0.7. This value is 1 for the three variables, making all assessed indicators reliable. aku aku aku.
Convergent validity: To fulfil this criterion, the average variance extracted must be higher than 0.5. The
average variance extracted is 1 for the three variables, fulfilling the requirements. iv. Discriminant
validity (Fornell–Larcker criterion): Each latent vari- able must have a higher variance with its indicators
than with any other latent variable. Because the square root of the average vari- ance extracted of each
construct is higher than its highest correla- tion with any other construct, this requirement is fulfilled. v.
Discriminant validity (cross-loadings): An indicator's outer load- ings on a construct should be higher
than all of its cross-loadings with other constructs. Because no indicator has a higher correla- tion with
any other variable than with its own latent variable, this requirement is fulfilled. b.)Formative Model
i. Nomological validity: The relationships between the formative in- dex and other constructs in the path
model must be strong and significant. This is the case, as supported by prior research. ii. External validity:
The formative index should explain a big part of the variance of an alternative reflective measure of the
focal construct. The only construct that can be assessed (service quality) is 0.908, higher than the
minimum threshold of 0.8. aku aku aku. Significance of weights: The estimated weights of formative
mea- surement models are above the tolerance level and therefore sig- nificant. iv. Multicollinearity: A
variance inflation factor (VIF) higher than 10 (Henseler et al., 2009) or 5 (Hair, 2013) indicates the
presence of harmful collinearity. Any VIF substantially higher than 1 indicates multicollinearity. For the
two constructs, the VIF is not substantially higher than 1 (1.043 and 1.000). 2. Assessment of the
Structural Model
a) Collinearity of the model: The tolerance and VIF (VIF<5) of each set of predictor constructs must be
examined separately for each subpart of the structural model. No relevant levels of collinearity are found.
Service quality and customer satisfaction exceed 4 for both models of loyalty toward workshop and brand,
consistent with the idea of service quality as an antecedent of customer satisfaction. However, the VIF is
lower than 5, so both constructs are kept in the model.
256 Automotive After Sales Service and Loyalty in Spain
FIGURE 3 Partial Least Squares Structural Equation Model Path Coefficients. Nonsignificant relations are displayed in dotted-
line boxes.
b) Estimates for path coefficients: The estimated values for path relation- ships in the structural model must
be evaluated in terms of sign, mag- nitude, and significance (the latter via bootstrapping). The path coeffi-
cients are displayed in Figure 3:
– Satisfaction with the service and the vehicle have no significant relation with loyalty. – Warranty
validity has the highest influence on loyalty to both brand and workshop. – Service quality has a low
impact on customer loyalty to both work- shop and brand, but this low effect is slightly higher to the
workshop than to the brand. On the other hand, service quality does have a big impact on customer
satisfaction. – Type of service has an influence on loyalty to the workshop but not to the brand. –
Customer characteristics influence loyalty to the brand (not to the workshop) and satisfaction. – Older
customers are more loyal. – No indirect effects are detected. – The service finalization process has the
greatest impact on service quality. c) R2 of endogenous latent variables: R2 values of .67, .33, and .19 in
the inner path model are to be considered, respectively, substantial, moderate, and weak. The model of
customer satisfaction is substantial
A. Garcıa Gonz á lez 257
(R2 = .8251), whereas both models of loyalty are very weak (R2 loyalty to brand = .0448, R2 loyalty to
workshop = .0389). d) Effect size ( f 2): Values of 0.02, 0.15, and 0.35 are a gauge for whether a predictor
latent variable has a weak, medium, or large effect, re- spectively, at the structural level. For loyalty to the
brand and to the workshop, only vehicle warranty validity has a relevant effect, but it is rather weak
(0.0418 and 0.0320, respectively). For customer satisfac- tion, only service quality has a relevant effect (a
large effect at 4.5790). e) Prediction relevance (Q2 and q2): If Q2 > 0, the model has prediction
relevance. q2 values of 0.02, 0.15, and 0.35 refer to low, moderate, and high relevance, respectively. In
the studied model, only warranty validity has prediction relevance for loyalty, but it is low (0.0395 to the
brand and 0.0318 to the workshop). Service quality is the only relevant construct with prediction
relevance for customer satisfaction (very high at 4.4373).
In conclusion, the model meets all of the necessary criteria. The lack of support for the impact of
customer satisfaction (H2 and H4) is driven by the high collinearity between customer satisfaction and
loyalty, which has a big influence on the PLS-SEM results. In this case, the positive results of MCA and
AID are accepted because they are not affected by collinearity.
SUMMARY OF FINDINGS
The variable with the greatest impact on loyalty is the validity of the warranty, followed by service quality
and customer satisfaction. Other variables have low effects, like the type of service, age, and gender.
Although the analysis of the service–loyalty link confirms the existence of a causal relation, it is much
weaker than expected. Regarding the rela- tionship between customer satisfaction and loyalty to both the
workshop and the brand, the cross-tabulation analysis shows that higher satisfaction is linked to higher
loyalty. Nevertheless this relation has a regressive shape (see Figure 1), showing low levels of loyalty for
unsatisfied customers but almost no difference in terms of loyalty between satisfied and very satisfied
customers. This fact is even clearer in the case of loyalty to the workshop, in which the regressive relation
is more acute. Both MCA and AID analysis result in a relatively weak impact of satisfaction on loyalty. In
the case of loyalty to the workshop, the explained variance only reaches 1.20% for MCA, and customer
satisfaction is just the third variable in the AID sample partition. The results for loyalty to the brand are
even lower, with 0.92% explained variance and no sample grouping based on this variable for the AID.
258 Automotive After Sales Service and Loyalty in Spain
Regarding service quality, and considering the PLS-SEM path coefficients as standardized beta
coefficients of an ordinary least squares regression, the weight of service quality in the decision to be loyal
to the workshop and the brand is 8.6% and 7.6%, respectively (with significant relations at p < .01, but with
very low R2 levels). In any case, the service–loyalty link is more complex than it was thought to be a few
years ago, as stated by Liao et al. (2010). It is nonlinear and probably influenced by many other variables
that were not included in this study.
Consequently, H1, H2, H3, and H4 are supported, confirming the exis- tence and influence of the
service–loyalty link. However, H7 and H8 are not supported, dismissing this link as the main driver of
customer loyalty. These results are surely influenced by the long interpurchase cycle for after sales services.
With regard to the different impact of service on loyalty toward the workshop or toward the brand, the
results are higher for the workshop. This gives support to H5 and H6, which predicted a greater impact of
both customer satisfaction and service quality on the retailer, as the retailer is the one having the direct
contact with the customer.
Regarding the differences by brand, the customers of the premium brand are more loyal to the
workshop (the loyalty ratio scores 62% for the premium brand customers vs. 52% for the medium-price
customers) and also to the brand (70% for the premium brand vs. 58% for the medium-price brand).
Nevertheless, both MCA and AID detect a higher influence of customer satis- faction on loyalty in the case
of the medium-price brand. The MCA explained variance for loyalty to the workshop is 1.01% for the
medium-price brand and 0.6% for the premium one. The same applies for loyalty to the brand, with 0.6%
and 0.31%, respectively. For AID, customer satisfaction turns up before as a sample grouping factor. Even
in the case of cross-tabulation, higher sensibility for the medium-price brand is detected. These results
support H9 and H10. These results confirm that premium brand customers differ from customers in other
segments (Bullinger & Van Husen, 2006). They also show that manufacturers are not working according
to their customers' behaviour. On the one hand, big volume manufacturers focus on profit in the after sales
business while their customers are especially sensitive to service. On the other hand, premium
manufacturers work on service enhancements while, as stated by Liao et al. (2010), the impact of such
policies is buffered by brand trust. Although premium brands benefit from higher loyalty rates (like in
Meyer, 2010), the loyalty of their customers is stronger and less affected by service quality.
The analysis of the validity of vehicle warranty shows that this is the variable from Model 1 with the
greatest impact on loyalty. It has the highest explained variance for MCA (3.72% for the workshop and
4.66% for the brand), and it is the first variable in the AID sample grouping. It also has the highest path
coefficients, with 0.176 for the workshop and 0.200 for the brand. The impact is logically higher for loyalty
to the brand, because it is
A. Garcıa Gonz á lez 259
the manufacturer who is providing the warranty, giving support to H11, H12, and H13.
Regarding the type of service, the results confirm a slight impact on customer loyalty to the workshop
but not to the brand. Although a small share of the variance is explained through MCA (0.5% to the
workshop and 0.3% to the brand), and this variable appears in third place in the sample grouping for both
types of loyalty, the PLS-SEM path coefficients are only significant for the case of the workshop. Even for
this case, the results are quite low, with a weight of only 3.6% in the customer's decision to be loyal. These
results therefore provide support for H14 but not for H15. Those workshops that are able to attract
proportionally more customers for competitive services (maintenance, service checks, and body shop) are
able to achieve higher levels of loyalty.
Regarding customer characteristics, the analysis was done individually for age, the analysis was done
individually for age and gender in the case of MCA and AID (as independent variables), while both
variables were merged into a single construct in the case of PLS-SEM. Age does have a slight impact on
loyalty, according to MCA and AID, with 0.38% and 0.63% explained variance to the workshop and to the
brand, respectively. In both cases this variable appears in the third level of the AID sample grouping.
Gender, in contrast, does not show a relevant impact on loyalty. Regarding PLS-SEM, the customer
characteristics construct has a weak but significant relation (at p < .01) on loyalty to the brand, with a 4.23%
weight in the customer's decision to be loyal. It also has a slight impact on customer satisfaction, 1.2%
(significant at p < .05), but no impact in the case of loyalty to the workshop. These results support the
impact of age through H16, H17, and H18 but not gender in the case of H19, H20, and H21. Nevertheless,
such a weak impact would not deserve greater attention on either the practical or theoretical level.
In summary, the results of the analysis provide support for H1, H2, H3, H4, H5, H6, H9, H10, H11,
H12, H13, H14, H16, H17, and H18 and do not provide support for H7, H8, H15, H19, H20, and H21.
DISCUSSION
These results serve not only to enrich the academic debate on customer retention management but also to
reallocate corporate resources. Until now, many automotive companies, especially those from the higher
price or pre- mium segment (Saccani et al., 2006), assumed that investing in service was profitable, although
this belief had not been checked sufficiently on an empirical level. The present research shows that the
impact on loyalty of enhancing service is limited and actually weaker than expected. Customers assume
that service quality is a must and will defect if the company under-
260 Automotive After Sales Service and Loyalty in Spain
performs. However, offering service standards above customer expectations will not result in higher loyalty.
This result is driven by the long interpurchase cycle, which softens the effect of service performance.
Consequently, if an automotive service company wants to boost cus- tomer loyalty, the focus has to
be set on other variables rather than offering excellent service. Quality must be ensured and kept at a
reasonable level, but other variables should be influenced to enhance customer retention. Among them, the
vehicle warranty plays an important role in customers' decision making. Customers still think that they are
obliged to visit official workshops during the warranty period, although that is legally no longer the case.
This feeling of contractual bonding influences customer retention. Considering that the vehicle warranty
represents a remarkable cost for ve- hicle manufacturers, an alternative tool with no incremental costs
should be applied, such as warranty extensions or service contracts. Both of these are financial products:
The customer pays a fixed amount and gets in re- turn extended warranty coverage or maintenance for a
defined period of time. They are usually linked with the obligation to visit official workshops and therefore
represent an ideal way to create contractual retention while offering additional profit as a financial product.
IMPLICATIONS FOR MANAGEMENT
The results of this study have managerial implications not only for official workshops and vehicle
distributors but also for their competitors:
• Premium brands: Higher price manufacturers and distributors should focus on contractual bonds to
enhance customer retention. As stated previously, warranty extensions and service contracts are key for this
purpose. Man- ufacturers should take the chance to sell these products together with the vehicle, shifting
from offering a physical product (the vehicle) to selling a bundle of services and products (mobility),
including maintenance, an extended warranty, and even financing and insurance. This would be in line with
the previously mentioned industrial servicization tendency (Ahn & Sohn, 2009). In addition, close control
and follow-up of contracts' expi- rations should be implemented in order to encourage either renovation or
vehicle replacement. This strategy should by no means lead to a deterio- ration of service standards, which
could boost customer defection.
• Medium-price brands: For these brands, service performance has a bigger impact on loyalty. Nevertheless,
investing in improving service is actually not the best alternative for them. First, service has a greater impact
for retailers than for the brand. Second, warranty validity also has a much greater impact on loyalty than
service performance among these brands. Consequently, these brands should follow the same strategy as
the pre-
A. Garcıa Gonz á lez 261
mium brands. This is actually less risky for the medium-price brands; the potential negative impact on brand
trust for them is lower thanks to their brand values.
• Official workshops: Official workshops benefit from a relatively greater impact of service performance
on loyalty but also from a strong impact of contractual binding (but less than the manufacturers). Therefore,
they could work on service performance in order to differentiate themselves from other retailers, trying to
gain customers for high competitive services (service checks, maintenance, and body shop repairs). At the
same time, their main focus should be to promote warranty extensions and service contracts. They could
even consider developing their own contracts to ensure loyalty to their facilities and not to just any retailer
with the same branding. Such a decision could be financially risky for them but in any case would be worth
analysing.
• Independent workshops and specialized automotive after sales service providers: These companies have
been highly successful recently thanks to their aggressive price policy, low prices, and proximity to the
customer. Consequently, they should keep working along the same line. It would be quite difficult for them
to develop their own line of service contracts or warranty extensions, either for all services or for those in
which some of them are specialized (eg, tires, glass, body shop). Customers are probably more likely to
accept such offerings if they cover all of their service needs, but this would be difficult for these types of
companies because they are either too small and/or specialized in certain types of services.
In summary, the demystification of service quality and customer satisfac- tion as means of increasing
customer retention for Spanish automotive after sales leads the way for a contractual retention strategy,
similar to cell phone communication services. However, any shift in the strategy must be analysed from a
holistic perspective, considering not only the after sales service but also the vehicle business. Insufficient
service quality could have a negative impact on brand trust and eventually jeopardize vehicle sales, which
is the source of the after sales business.
RESEARCH LIMITATIONS
The first limitation of this study, and one of the main ones, is the use of secondary data. The availability of
customer satisfaction data from two auto- motive brands made it possible to benefit from a very big sample.
However, the questionnaires were different and were not perfectly suited to the needs of this investigation.
This fact also limited the available variables and made it unfeasible to apply PLS-SEM to both brands.
Nevertheless, considering the costs of doing such a study from scratch, the study would not have
262 Automotive After Sales Service and Loyalty in Spain
been possible without secondary data. The second limitation is the use of transactional satisfaction ratings,
as the questionnaires were related to just one customer visit to the workshop. Previous studies have
confirmed that accumulative satisfaction has a stronger effect on purchase behaviour than transactional
satisfaction (Bodet, 2008; Meyer, 2010). The third limitation is the participation of just two manufacturers
in the study. Although this is the first time that manufacturers have provided data for such a study in Spain,
the participation of other companies would have enriched the results. Data from big volume brands (eg,
Citroen, Peugeot, Ford, Renault, SEAT, Volk- swagen) and nonmanufacturer automotive service chains
(eg, Midas, Aurgi, Carglass) would have helped to get a broader picture of Spanish automotive after sales
customers and their behaviour. The fourth and last limitation is the concept of loyalty applied in this study.
Given the type of data available, loyalty to both brand and workshop were measured from a behavioural
per- spective, without considering the intentionality of that behaviour. Repeated purchase can be driven by
convenience, use, lack of alternatives, or switch- ing costs (Homburg & Koschate, 2007). Moreover, a
customer can be loyal to several service providers (eg, going for maintenance to a fast fit service chain and
visiting an official workshop for complex repairs), making out of the share of wallet a third dimension of
loyalty to be ideally considered.
FURTHER RESEARCH
The present investigation opens the way for further investigation. First, there is a clear need to research
customer defection. Service quality and customer satisfaction explain just a small portion of customer
retention. Studying de- fecting customers could potentially help to define the reasons for defection (eg, trust
in the service provider, brand trust, variety seeking, workshop proximity, perceived price, actual service
needs, fluctuations in purchase power, vehicle change). This knowledge would enable more effective man-
agement of customer retention. Second, this research is focused on the Span- ish market. Considering the
possible differences in customer behaviour by market, it would be useful to extend the investigation to other
countries. This would help to adapt the central strategies of manufacturers to local customer needs (“think
globally, act locally”), making them more effective. Third, it would be beneficial to study the impact of
service quality and cus- tomer satisfaction on brand trust and eventually the effect of brand trust on vehicle
sales. This research has concluded that the effect of service quality and customer satisfaction on retention
is limited. This finding reduces the need to invest in further improving the service delivered to the customer.
At the same time, a reduction in service quality could have a negative impact on brand trust (Liao et al.,
2010). Such an effect could potentially negatively affect the whole automotive business, both the sales and
after sales sides.

Вам также может понравиться