Open access peer-reviewed chapter

Logistics Services Satisfaction Survey: SERVQUA

Written By

Marieta Stefanova

Reviewed: February 16th, 2022 Published: March 23rd, 2022

DOI: 10.5772/intechopen.103754

Chapter metrics overview

59 Chapter Downloads

View Full Metrics


The chapter of the study presents the performed analysis of the logistics services satisfaction survey. It has been found that later the non-conformities in the logistics quality management system are identified and corrected, the more serious they are. The analysis is performed using the PLS-PM model captures the causal relationships of the study sample through arrows that start at a latent variable (factor) and point to the measured indicator variables. Results show that expected quality is the most important and effective latent variable. Customers were positive about the satisfaction of their requirements and the ability of logistics service providers to assess their expectations.


  • logistics services satisfaction
  • PLS-PM

1. Introduction

Physical distribution is today’s frontier in business. It is the one area where managerial results of great magnitude can be achieved. And it is still a largely unexplored territory’ Peter Drucker[1].

The later non-conformities in the logistics quality management system are identified and corrected, the more serious they are. Some of the non-conformities in the system may not be detected at all due to the complex requirements for completing and submitting information to the service provider or the control bodies that carry out counter-checks on the case:

  • Process management requires addressing the root causes of non-conformances to minimise their recurrence [2, 3, 4, 5].

  • Significant cost reductions in quality management can be achieved if the system is planned appropriately for the conditions [6] and actual execution of the processes, rather than devoting additional costs to controlling inconsistencies and errors that would occur during the execution of the logistics processes.

  • Quality and risk management are part of every business process [7, 8, 9, 10]; only the application of expert opinions and methods is not applicable in day-to-day operations.

  • The later the risk or non-compliance is discovered, the more difficult it will be to take adequate corrective action and the greater the cost of doing so.

  • It is much more effective to address the causes of potential system failures at the process design stage than to look for methods to detect and correct inconsistencies.

  • Applying the principle of good system planning to prevent errors is fundamental and leads to a reduction in overall costs.

The recipient of logistics services, who is a party to the supply contract, may bring an action against the supplier for breach of any expressed or implied term of the supply contract in relation to a non-conforming service. The supplier is obligated to take all reasonable steps to prevent acts or omissions which are reasonably foreseeable and expected to cause material damage. Any action taken to address operational risks shall be proportionate to the potential impact on the compliance of the logistics services [11].

Results of the satisfaction analysis in logistics services.


2. Analysis model of satisfaction

REBUS-PLS (a response-based procedure for detecting single segments in partial least squares modelling) was developed within the PLS-PM approach by Esposito Vinzi [12]. Many studies detect distinct behavioural differences that are caused naturally by the interactions that take place between the environment and its preferences. The analysis found that such differences often exist in the group studied and the differences clearly define two clusters of behaviour patterns.

The analysis was performed in Excel using the statistical software XLSTAT 2021.4.1.1201 – REBUS [13]. The number of classes was automatically determined during the cluster analysis with a chosen threshold of 95%. Three consecutive clustering operations were established, which created a binary cluster tree (dendrogram) with the root class containing all observations. The results are presented in Figure 1.

Figure 1.

Agglomerative hierarchical clustering (AHC). REBUS-PLS (response-based single segment detection procedure in partial least squares modelling).

In the figure, it can be clearly observed that the first group (right) is more homogeneous than the second (it flatter on the dendrogram). This is confirmed when examining the within-class variance, which is much higher for the second group than for the first. In class 1, satisfaction is mainly explained by the perceived quality and in class 2, satisfaction is also explained by perceived quality but logistics service expectations have a less significant effect.

2.1 Agglomerative hierarchical clustering

REBUS algorithm: Dissimilarity: CM Index. Stop conditions: Iterations = 100 Threshold = 95%. The REBUS algorithm did converge after 99 iterations.

2.2 Analysis for the PLS-PM model of satisfaction

The PLS-PM model (estimated with XLSTAT 2021.4.1.1201-PLS-PM) captures the causal relationships of the study sample through arrows that start at a latent variable (factor) and point to the measured indicator variables. The model specification (measurement model) is presented in Table 1.

Latent variable Latent variableImageExpectationPerceived
Number of manifested variables5252322
ModelModel AModel AModel AModel AModel AModel AModel A

Table 1.

Model specification (measurement model).

The latent variables assessed in the model are Expectation: customer expectations related to logistics service culture; Quality: the expected quality of logistics services; Value: the perceived value of logistics services; Image: the reputation and tangible indicators of logistics services; Satisfaction: satisfaction derived from the comparison between the expected and actual perception of the logistics service; Complaints: customer complaints arising in the course of logistics service provision, unmet expectations, perceived or actual customer requirements, Loyalty: loyalty is the positive attitude of the customer towards the service provided. The type of model with latent variables is presented in Figures 2 and 3.

Figure 2.

Measurement model with manifested variables.

Figure 3.

Measurement model with values for Corr, R2, Reg.

When running the model, the following criteria were set: Display: Expert and Marketing, Method: PLSPM, Stopping Conditions: Iterations = 5 /Convergence = 0.0001, Confidence Intervals: 95/Bootstrap /Resamplings = 100, Latent Variable Scores: Scale MV, Treatment of manifest variables: Standardised, weights on standardised MV, Values of the first eigenvector, Internal estimation: PLS, Score for latent variables: Standardised.

The first stage in the analysis is the confirmation of the reliability of the results obtained from the study based on the information collected about the data and the model created. The analysis is presented in Table 2.

Latent variableDimensionsCronbach’s alphaD.G. rho (PCA)Condition numberCritical valueEigenvalues
Customer expectations20.0680.7821.0361.0001.035
Perceived quality50.7600.8551.4841.0001.544
Perceived value20.6600.8551.7151.0001.493
Complaints from customers20.9690.9855.6501.0001.939

Table 2.

Composite reliability variance of the outcome against the total variance (Monofactorial manifest variables).

Cronbach’s alpha calculations in the following cases are below 0.7: expectancy, perceived value, and loyalty. For these latent variables, Dillon and Goldstein’s rho is above 0.7 and the first eigenvalue is greater than the second. This leads us to believe that the data collected from the survey can be used for further analysis and interpretation.

A correlation matrix was constructed with the data, presented in Table 3.

6Image 1Image 2Image 3Image 4Image 5Expectation 1Expectation 2Quality 1Quality 2Quality 3Quality 4Quality 5Value 1Value 1Satisfaction 1Satisfaction 2Satisfaction 3Complaints 1Complaints 2Loyalty 1Loyalty 2
Image 11.000−0.113−0.0360.0060.002−0.063−0.128−0.131−0.034−0.084−0.0750.197−0.095−0.013−0.115−0.111−0.176−0.091−0.0810.093−0.091
Image 2−0.1131.000−0.1460.1780.008−0.0630.039−0.1040.0550.090−0.274−0.1190.073−0.058−0.107−0.0730.021−0.0370.0090.051−0.037
Image 3−0.036−0.1461.0000.0440.053−0.217−0.139−0.363−0.1050.0390.1690.0510.0990.225−0.218−0.326−0.331−0.321−0.3170.007−0.321
Image 40.0060.1780.0441.0000.0220.014−0.0930.0950.202−0.0690.0650.1950.230−0.0430.0710.1000.0880.1010.096−0.0860.101
Image 50.0020.0080.0530.0221.0000.0550.0380.050−0.079−0.0670.0910.0280.024−0.030−0.0380.0630.0170.0560.0990.0730.056
Expectation 1−0.063−0.063−0.2170.0140.0551.0000.0350.806−0.056−0.0150.143−0.0690.0470.0640.5860.7670.6810.8420.8170.0170.842
Expectation 2−0.1280.039−0.139−0.0930.0380.0351.0000.153−0.113−0.080−0.027−0.049−0.1100.1240.0700.1920.2950.1520.0990.2260.152
Quality 1−0.131−0.104−0.3630.0950.0500.8060.1531.0000.102−0.1120.1620.032−0.054−0.0340.7640.9650.8600.9440.8910.1090.944
Quality 2−0.0340.055−0.1050.202−0.079−0.056−0.1130.1021.000−0.098−0.0210.0000.1100.036−0.0100.115−0.002−0.0020.0430.117−0.002
Quality 3−0.0840.0900.039−0.069−0.067−0.015−0.080−0.112−0.0981.000−0.240−0.220−0.202−0.128−0.221−0.136−0.127−0.089−0.0540.078−0.089
Quality 4−0.075−0.2740.1690.0650.0910.143−0.0270.162−0.021−0.2401.0000.2490.0670.1490.2640.1060.1080.1390.1240.0490.139
Quality 50.197−0.1190.0510.1950.028−0.069−0.0490.0320.000−0.2200.2491.0000.0640.0830.1570.0440.0050.014−0.0070.0920.014
Value 1−0.0950.0730.0990.2300.0240.047−0.110−0.0540.110−0.2020.0670.0641.0000.493−0.017−0.072−0.157−0.0100.010−0.079−0.010
Value 1−0.013−0.0580.225−0.043−0.0300.0640.124−0.0340.036−0.1280.1490.0830.4931.0000.061−0.041−0.0560.0030.016−0.0480.003
Satisfaction 1−0.115−0.107−0.2180.071−0.0380.5860.0700.764−0.010−0.2210.2640.157−0.0170.0611.0000.7330.7160.7050.6600.0160.705
Satisfaction 2−0.111−0.073−0.3260.1000.0630.7670.1920.9650.115−0.1360.1060.044−0.072−0.0410.7331.0000.8320.9070.8530.0970.907
Satisfaction 3−0.1760.021−0.3310.0880.0170.6810.2950.860−0.002−0.1270.1080.005−0.157−0.0560.7160.8321.0000.8190.7440.1670.819
Complaints 1−0.091−0.037−0.3210.1010.0560.8420.1520.944−0.002−0.0890.1390.014−0.0100.0030.7050.9070.8191.0000.9390.1421.000
Complaints 2−0.0810.009−0.3170.0960.0990.8170.0990.8910.043−0.0540.124−0.0070.0100.0160.6600.8530.7440.9391.0000.1220.939
Loyalty 10.0930.0510.007−0.0860.0730.0170.2260.1090.1170.0780.0490.092−0.079−0.0480.0160.0970.1670.1420.1221.0000.142
Loyalty 2−0.091−0.037−0.3210.1010.0560.8420.1520.944−0.002−0.0890.1390.014−0.0100.0030.7050.9070.8191.0000.9390.1421.000

Table 3.

Correlation matrix of the study.

The correlation analysis, presented in Table 3, shows the relationship between dependent and independent attributes.

According to Lorentz [14], correlation coefficients do not give a completely accurate picture of the extent to which a trait is a dependent trait in itself and through other traits. In his study, Lorentz proves that there is a limitation in the evaluation of correlation coefficients as these only show the magnitude and direction of the linear relationship between traits without revealing causal relationships. Thus, we performed further analyses based on path coefficients for latent variables.

The measurement model was found to be reliable and fit to the sample because the model fit indices were within the range recommended by Hu and Bentler [15].


3. Impact analysis and contribution of satisfaction variables

Once the measurement pattern has been studied, the structural model can be analysed. All the latent variables explored in the model are related to several other factors with several interpretable solutions. The results obtained for the latent variable satisfaction confirm its interrelationship with the other factors (R2 = 0.681) (Table 4).

SatisfactionR2FPr > F
Latent variableValueStandard errorTPr > |t|f2
Perceived quality0.8620.06014.2590.0002.140
Perceived value−0.0540.036−1.5060.1350.024

Table 4.

Path coefficients for the latent variable: Satisfaction.

The first hypothesis aims to establish the relationship between satisfaction and perceived quality, expectancy, and image. Perceived quality (0.936) was found to have the strongest effect on satisfaction followed by customer expectations of logistics services. The values obtained for the relationship between correlation and path coefficient support the hypothesis as perceived quality is 0.806, expectations are 0.068, perceived value is 0.003 and image is 0.004. Variable importance in the projection (VIP) was calculated, under VIPs (1 Comp / 5% conf. interval). The following chart (Figure 4) summarises the results obtained for the Impact and Contribution of Satisfaction Variables and their advisor VIPs in Figure 5.

Figure 4.

Impact and contribution of satisfaction variables.

Figure 5.

Variable importance in the projection (1 comp / 5% conf. interval).

The study made many assessments of the quality of each stage of customer service according to the most important criteria related to the image and tangibility of the services offered. It was found that customers evaluate the identified indicators of logistics task performance differently. In this study, the questions related to image variables 1, 2, and 3 have a positive relationship with satisfaction, while those related to image variables 4 and 5 have a negative relationship. Factor 4 was found to have the most significant influence on the overall satisfaction pattern and was related to the time between placing the request and receiving the delivery. The results obtained are presented in Table 5.

Latent variableVariables of the manifestExternal weight
ImageImage 10.344
Image 20.146
Image 30.956
Image 4−0.206
Image 5−0.150

Table 5.

Path coefficients for the latent variable: Image.

Customers of logistics services consider problem solving as one of their expectations of the logistics services provided. Expectations for logistics services are related to the knowledge and experience of the persons involved in the process and problems are solved operationally by the designated contact person. The image was found to have the most significant impact on expectations (0.077), the values obtained for the relationship between correlation and path coefficient are: 0.278 and are presented in Table 6.

ExpectationsR2FPr > F
Latent variableValueStandard errortPr > |t|f2

Table 6.

Path coefficients for the latent variable: Expectation.

The second hypothesis aims to establish the relationship between perceived quality and customer expectations. Expectations towards the logistics services provided also influence the overall satisfaction or service quality. The value obtained for the relationship between correlation and path coefficient is 0.789 and confirms the hypothesis at a p-value <0.05. The results are provided in more detail in Table 7. Therefore, this study proposes that companies that correctly perceive customer needs provide logistics services with better quality. To test the hypothesis, initial data from completed questionnaires were used, customers were asked to directly rate their service provider.

Perceived qualityR2FPr > F
Latent variableValueStandard errortPr > |t|f2

Table 7.

Path coefficients for the latent variable: Quality.

Customers relate their expectations of whether the logistics services meet their requirements (the specification) and whether these are reliable and on time to the perceived value of logistics services (the path coefficient is R2 0.015). Meeting these requirements directly reduces the potential number of competitors and creates a significant competitive advantage (Table 8).

Perceived valueR2FPr > F
Latent variableValueStandard errortPr > |t|f2
Perceived quality−0.1620.164−0.9910.3240.010

Table 8.

Path coefficients for the latent variable: Perceived value.

Customers who receive high-value services will also expect a high level of service. It can be assumed, based on the results obtained, that disappointment with the service quality is likely to encourage customers to seek alternative service providers, even though switching logistics providers will increase costs and cause time wastage. The results that confirm this hypothesis of impact and contribution of variables to the perceived value are presented in Figure 6.

Figure 6.

Impact and contribution of variables: Perceived value.

The results also confirmed the hypothesis that there is a direct relationship between complaints and customer satisfaction with logistics services. Complaints were found to strongly impact satisfaction (path coefficient R2 0.873). Providing quality services that meet the requirements without discrepancies or complaints will achieve greater customer confidence, minimise the risk of competition and ensure customer loyalty and increase the logistics provider’s welfare from potential service revenue. The calculations of the significance coefficients (Table 9) show this hypothesis to be true.

ComplaintsR2FPr > F
Latent variableValueStandard errortPr > |t|f2

Table 9.

Path coefficients for the latent variable: Complaints.

Dynamic changes in the external environment deepen the drive for customer satisfaction, which is strongly influenced by the overall quality of logistics services. Satisfying customers’ service quality requirements influence business performance and promote customer loyalty. Customer loyalty depends on a variety of factors, such as perceived service quality, personal preferences and expectations, social interaction [16], customer experience in dealing with logistics providers, and many other specifically subjective factors. Furthermore, it is necessary to bear in mind that customer loyalty is not an objective assessment of the real situation but an emotional element. The hypothesis that loyalty strongly influences the satisfaction of customer expectations is confirmed by the results obtained for the path coefficient is R2 0.957 (Table 10).

LoyaltyR2FPr > F
Latent variableValueStandard errortPr > |t|f2

Table 10.

Path coefficients for the latent variable: Loyalty.

Customer loyalty is assessed as a consequence of the process and presented with recommendations for the logistics services used in the long and short term. The results confirming the impact hypothesis and the contribution of the loyalty variables are presented in Figure 7.

Figure 7.

Impact and contribution ofloyaltyvariables.

Marketing analyses of the survey data were also conducted to seek recommendations for customer satisfaction management. An IPMA (Importance performance matrix analysis) was performed and the visualised results are presented in Figure 8. This analysis is based on the importance and effectiveness of customer satisfaction and demonstrates the extent of the relationship between the exogenous and endogenous variables that can be derived. The partial least squares modelling method for structural equations with XLSTAT PLS is used to establish IPM.

Figure 8.

Latent variables on a target variable (importance performance matrix analysis – IPMA).

Results show that expected quality is the most important and effective latent variable. A high score was also found for customer expectations related to the logistics services provided. The fastest change in satisfaction can be achieved with the improvement of image impact and expected value. To explore this potential in more detail, incentive diagrams Figures 9 and 10 were constructed. The values in Figure 9 shows how the value of the target latent variable score would change if an improvement of x% is applied to the manifested variable. In the simulation results for the weight for predicting satisfaction (latent variable), the most important variables are perceived quality, expectations, perceived value, and image. The values in Figure 10 show the manner in which the mean value of the target latent variable will change if an improvement of x% is applied to the manifested variable. In the simulation results for the weight for predicting satisfaction (latent variable), the most important manifestation variables are Quality 1, Expectation 1, Image 3, Expectation 2, Quality 4, Quality 3, Image 1, Image 4, Value 1, Image 5, Image 2, Value 1, Quality 2, and Quality 5.

Figure 9.

Incentive diagram of change in satisfaction value.

Figure 10.

Incentive diagram for average level of satisfaction.

In summary of the analyses, it may be concluded that service providers who want to attract potential service users must first appropriately understand their expectations of service quality as it has a direct impact on customer satisfaction. High-quality logistics services increase service providers’ advantages and customer loyalty. The study showed that customers prefer a company that provides quality services and takes steps to reduce complaints from customers.


4. Conclusion

The study establishes that customers value not only the quality of service but also the expected value of the service provided, the good image of the logistics company is directly related to the competence of the staff who should be personally involved in solving the problems of risks arising from non-conforming supplies or processes.

The study shows that logistics service providers are aware and perceptive of customer expectations and can meet them, as the satisfaction indicators surveyed are more positive than negative. Customers were positive about the satisfaction of their requirements and the ability of logistics service providers to assess their expectations.

The study also proves that logistics firms need to manage the risks of non-compliant logistics services and minimise them, which will be discussed in the next chapter.



A huge thank you to my colleagues at the Department of Management and Logistics for encouraging this project.

I also thank all the logistics companies and users of logistics services who participated in the survey and shared their experiences.

I also thank the “N. Vaptsarov” Naval Academy for opening opportunities and providing access to world-class educational and scientific resources and assistance from distinguished professors with scientific interests in the field of my research.

Thank you to all the organisations that hired me as a consultant and helped me strive for lifelong learning.


  1. 1. Drucker P. The economy’s dark continent. Fortune. 1962;65:265-270
  2. 2. Ask C, Trygged M. Enhancing Problem Solving by Improving the Identification of Root Causes to Avoid Reoccurring Non-conformances. Gothenburg: Chalmers University of Technology; 2016
  3. 3. Lee HL, Whang S. Higher supply chain security with lower cost: Lessons from total quality management. International Journal of Production Economics. 2005;96:289-300. DOI: 10.1016/j.ijpe.2003.06.003
  4. 4. Prokopowicz D, Gołębiowska A. Business Intelligence Analytics Based on the Processing of Large Sets of Information with the Use of Sentiment Analysis and Big Data. Warsaw, Poland: Publishing House SGSP; 2021. pp. 129-154
  5. 5. Zafirova T. Integration of the processes of strategic and crisis Management in the Conditions of crisis. Известия на Съюза на учените – Варна Серия Икономически науки. 2021;10:158-168
  6. 6. Lambert D, Douglas M, James R. Strategic Logistics Management. 4th ed. Homewood: McGraw-Hill Education; 1993
  7. 7. Lai K, Vejvar M, Lun V. Risk in port logistics: Risk classification and mitigation framework. International Journal of Shipping and Transport Logistics. 2020;12:576-596
  8. 8. Zafirova T. Impact of changes in the environment on the strategic aspects of crisis management. Izvestia Journal of the Union of Scientists – Varna Economic Sciences Series. 2021;10:184-192
  9. 9. Thai V. Logistics service quality: Conceptual model and empirical evidence. International Journal of Logistics Research and Applications. 2013;16:114-131. DOI: 10.1080/13675567.2013.804907
  10. 10. Olson D, Wu D. Supply chain risk management. In: New Frontiers in Enterprise Risk Management. 1st ed. Berlin, Heidelberg: Springer; 2008
  11. 11. Craighead CW, Blackhurst J, Rungtusanatham MJ, Handfield RB. The severity of supply chain disruptions: Design characteristics and mitigation capabilities. Decision Sciences. 2007;38:131-156. DOI: 10.1111/j.1540-5915.2007.00151.x
  12. 12. Evans M. Domesday marketing? From an inaugural lecture. Journal of Marketing Management. 1994;10:409-431. DOI: 10.1080/0267257X.1994.9964287
  13. 13. Addinsoft. XLSTAT 2021.3.1 Data Analysis and Statistics Software for Microsoft Excel. Paris, France: Addinsoft; 2021
  14. 14. Lorentz L. Production and body composition traits of broilers in relation to breast weight evaluated by path analysis. Scientia Agricola. 2011;68:320-325. DOI: 10.1590/S0103-90162011000300008
  15. 15. Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal. 1999;6:1-55. DOI: 10.1080/10705519909540118
  16. 16. Kanev D. Commitment as a constraint to the pursuit of self-interest. Economics. 2017;1:3-20

Written By

Marieta Stefanova

Reviewed: February 16th, 2022 Published: March 23rd, 2022