Assessment of customer satisfaction with the water supply system in low-income countries: A case of Adama City, Ethiopia

Living conditions for people and the availability of clean water are interdependent. Customer satisfaction is one of the key factors for a future-proof organization or business, and setting goals and targets through strategy planning is essential (Reis & Peña 2000; Al-jazzazi & Sultan 2017). In addition, satisfaction with the delivery of water services to individuals is a general attitude or behaviour towards the customer on the part of the service provider and an emotional reaction to the discrepancy in the realization of a want, need, or objective between what they expect and what they receive (Jorgensen et al. 2009; Pakurár et al. 2019). Inconsistencies exist between consumer satisfaction with the quality of the water and services and compliance with monitoring standards (Denantes & Donoso 2021). The provision of services impacts cost recovery for service providers. A lot of water services in Africa are unable to recover even 50% of their total invoiced amounts in any billing cycle (Kayaga et al. 2004). Delayed bill payments and massive backlogs can severely impair a utility's ability to provide water service (Post et al. 2010).

On the other hand, clients are not served by service providers because of issues in sector governance, potable water coverage, limited maintenance, and discontinuity of service (Grönwall & Danert 2020). Client satisfaction is based on the level of quality of service provided (Saravanan & Rao 2007). It is important to comprehend how perceptions are created and the variables that influence them to minimize the gap between perceptions and realities. As a result, understanding how consumer impressions are created is crucial in the realm of drinking water supply (Luo et al. 2022). Of the customers in six rural kebeles of Woliso district and Southern national nationalities people's region, 51.54 and 43% dissatisfied with water supply services in Ethiopia, respectively (Kassa et al. 2017; Mekuriaw & Gurmessa 2020). According to Adama Water Supply and Sewerage Enterprises (AWSSE), Adama's water supply coverage is approximately 74% and its water loss is approximately 24.7% as of 2022 (Desta et al. 2022). The Adama city government is preparing to upgrade the current sanitation and water supply systems because their lifespan is almost over. To improve water supply services, it is necessary to take customer satisfaction into account. As such, a thorough study is required to provide guidance to the city administration.

The present study aimed to (i) assess the customer's satisfaction with water supply service in Adama town, Ethiopia; (ii) investigate the demographic characteristics of categorical variables; (iii) assess overall customers' satisfaction with service provision; (iv) evaluate the relationship between demographic characteristics and overall satisfaction of respondents; and (v) analyse customer satisfaction and the dimensions of service quality by bivariate correlation analysis and multiple linear regression models. It is AWSSE's first attempt to find customer satisfaction in the study area. This study will serve as a standard for future studies, as well as a resource for decision-makers and policymakers such as the Ministry of Water and Energy, regional water and energy bureaus, and city water supply and sewerage enterprises that can use the study results to take corrective and integrated actions to improve service quality and change customer perceptions.

Adama town is situated about 90 km southwest of the capital city Addis Ababa. The town is located between 8°26′34″ N to 8°36′1″N and 39°13′9″E to 39°21′3″E with an elevation range between 1,590 and 1,770 m.a.s.l (Desta et al. 2022). The settlement of Adama is situated in the Rift Valley on the flat land with ridged topography surrounding it (Figure 1). The town, which has six sub-cities named Aba Geda, Bole, Boku, Dabe Soloke, Denbela, and Lugo, has a total area of 110 km². When the most recent growth area is taken into account and added to the master plan, the city's overall size grows to 136.7 km² (Bulti & Abebe 2020). The city experiences precipitation on an annual average of 1,371, 629, and 920 mm, respectively. Adama city's average annual temperature is 21 °C, and it is located in a sub-humid tropical zone (Asfaw 2007).

According to CSA (2007), Adama had a total population of 220,212 of whom 108,872 were men and 111,340 were women. The projected population of Adama city is 374,332 with 89,127 households in the year 2021. Adama has six sub-city administrations made up of 18 kebeles. A total of 75,859 people in the city use the water service in total. According to Adama Water Supply and Sewerage Services, the enterprise head stated that the city's most recent water supply coverage is 78%. The city has been utilizing a surface water treatment plant with a rapid sand filter. The distribution system can be divided into five distinct pressure zones due to the city's topographic layout. As shown in Table 1, it contains three more reservoirs functioning as boosting at three booster stations, together with eight service reservoirs of varying capacities, to address the communities living in a relatively elevated area of the city (Bokku Shaenaen, Dhaekaa Adi, and Hangatu). The city's water supply system has around 560 km of pipeline with sizes ranging from DN 40 m to 600 mm installed (Table 1, Figure 1).

The population statistics for Adama city were provided by the Central Statistical Agency. The primary data were collected using closed-ended questionnaires to get the respondents' direct responses. Three elements made up the questionnaire: (a) demographic information about the respondents; (b) an assessment of services' level of customer satisfaction; and (c) semistructured open-ended interviews, which were used to enhance the closed-ended questionnaire's findings. Before data collection, the variables were clearly defined for this study and their accurate and reliable measurement through daily monitoring of data collection was certified. Moreover, employing random sampling techniques helps to minimize bias and improve the generalizability of the findings by obtaining a representative sample from the target population.

The 75 questionnaire is designed about water supply services quality under five dimensions, i.e., tangibility, reliability, responsiveness, assurance, and empathy on top of socio-demography. The questionnaires' dimensions are selected based on an inclusive literature review and expert interview. The questionnaire is prepared in Orromiffa, English, and Amharic versions for the understanding of experts and city people. The questionnaires are rated on five-point Likert scales in a structured format with the oral statements ‘strongly disagree’ and ‘strongly agree’ anchored to the numbers 1 and 5.

Pearson's correlation coefficient and its significance levels are calculated using the bivariate correlations approach. Correlations quantify the relationship between different variables or rank orders. A measurement of the linear relationship is Pearson's correlation coefficient. Even if a link between two variables is exactly linear, Pearson's correlation coefficient should not be used to measure their association. This model uses a continuous variable and examines how customer satisfaction is impacted by service quality (Ahlgren et al. 2003).

• 1.

  Var (έi) = δ2, and the error term should have constant variance (homoscedasticity).

• 2.

  The error term is an independently and identically distributed random variable with a mean 0 and variance δ², i.e., it distributes N (0, δ²).

• 3.

  X's are independent of one another (no multicollinearity).

• 4.

  Explanatory variables and error terms are uncorrelated.

• 5.

  There is no autocorrelation between error terms.

• Linearity: By showing the response variable versus the fitted value in a graphic known as a scatter plot, linearity can be evaluated. The plot pattern must be roughly linear to be linear (Baur 2019). As a result, a plot in Figure 3 in shows a roughly linear relationship between a set of predictor variables and customer satisfaction.

• Normality: A histogram and a pp-plot can be used to test for normalcy. If the histogram's error term distribution is roughly normal (bell-shaped) and the pp-plot's point distribution is centred on the straight line, normality is obtained; if not, it is violated. The pp-plot line for these data was located at 450, as shown in Figure 4. It indicated that the premise of normality was true (Ghasemi & Zahediasl 2012). The result assumption is fulfilled as shown in Figure 4.

• Constant variance (homoscedasticity): Constant variance can be confirmed using the scatter plot of the fitted value versus the standardized residual. There must be no discernible pattern in the plot's distribution of points (Diblasi & Bowman 1997; Flatt & Jacobs 2019). As shown in Figure 5, there was no discernible pattern in the distribution of the dots, which were dispersed all around the mean zero. Thus, the assumption of constant variance was satisfied.

• The absence of multicollinearity: Bivariate correlation analysis and the variance inflation factor (VIF), when the Pearson correlation in bivariate correlation analysis is less than 0.5 and the VIF value is less than 10, can be used to determine the lack of multicollinearity. Otherwise, the presumption that multicollinearity does not exist is broken (Kim 2019). The absence of multicollinearity was assumed to be true since the results shown in Table 2 show that the value of VIF for each predictor variable is significantly less than 10.

• Data reliability test: All study variables have Cronbach's alpha values that are higher than 0.7. This suggests that the study's data are more trustworthy and that the records of the study's variables exhibit strong internal consistency (Elsayed 2012). All of the study's variables were sufficiently dependable, as shown in Table 3.

A statistical technique called the chi-square test is used to compare actual outcomes to predictions. This test aims to determine whether a discrepancy between actual and projected data is caused by chance or by a connection between the variables being examined. The chi-square test is a great option for helping us comprehend and evaluate the relationship between our two category variables as a result (McHugh 2013). The Chi-square test is important for effectively describing the issue based on multiple comparisons.

Customers are satisfied when the services provided match their needs. The effectiveness and aesthetics of infrastructures and water supply facilities are the measurement criteria for customer satisfaction in terms of urban water supply service quality (Ohwo & Agusomu 2018).

The degree of satisfaction is the discrepancy between the performances of the water supply service as expected and as perceived on any spatial scale (Ohwo & Agusomu 2018). Based on the perception of performance water supply quality, and the gap scores between satisfaction and non-satisfaction service aspects, the Customer Satisfaction Index (CSI) measures the degree of satisfaction in dwellers of Adama town.

As indicated in Table 4, in this study, the level of satisfaction with the calibre of the urban water delivery service in Adama city is evaluated using CSI (Alam & Mondal 2019) (Table 4).

• Empathy: It refers to the caring, individualized attention provided to the customer. (It focuses on the four dimensions of service quality such as timely information about the water interruption, clear justification about the water interruption, giving adequate time for water billing and encouraging water supply services.)

• Reliability denotes the ability to perform the service dependably and accurately (it is measured by six parameters, namely, consistent provision of water supply services, timely correction of non-functional water supply system, promising water in the compound by utility, sincerely solving customer problems, the customer receiving the accurate billing, and customers aware about water interruption in advance).

• Assurance is the knowledge and courtesy of employees and their ability to convey trust and confidence. (It focuses on the utility's employees' trustworthy, customer-considered safe water supply, employees' politeness for the customers while providing services, and knowledgeable employees to address customer enquiries.)

• Tangibility or service environment describes the appearance of physical facilities, equipment, personnel, and communication materials. (It is measured in four dimensions, i.e., utility has up-to-date equipment for water supply system work; utility's facilities are feasible to customers; utility's staff are well dressed and appear near and the pipelines are well maintained.)

• Responsiveness designates willingness to help customers and provide prompt service. (It emphasizes utility's attention to customers at the heart, identifying the customers' needs and complaint handling mechanism.)

Residents receive assistance in selecting and then putting into practice a system for gathering and analysing data on their water services. The information was collected by a questionnaire method from 435 households out of 89,127 households in Adama city. Eridadi et al. (2021) studied 60.4% female participants in Sebeta town, Ethiopia. On the other hand, James et al. (2014) noted that 56.3% were female interviewees in Kenya. The number of successfully collected questionnaire reports was 435 (98.86%) out of 440. These results are suitable for the analysis because a response rate of 70% or above is considered satisfactory for survey research (Salant & Dillman 1994). The demographic characteristics of respondents are presented in Table 5.

The information was collected from 275 (63.22%) females and 160 (36.78%) males out of a total 435 participants. Since women are in charge of performing water-related household duties including cooking, washing, cleaning, drinking, and caring for children, which all require water, the statistics with the highest percentage of females will give more accurate information as a majority of water users were frontline users of water supply services (Tessema 2020). The percentage of married and single respondents overall is 360 (82.76%) and 75 (17.24%), respectively. There are also 111 (25.52%) formal and 324 (74.48%) informal occupations. Among these, 296 (68.05%) and 139 (31.95%) of respondents live in their own houses and rent houses, respectively. From 435 respondents, 189 (43.45%) had low income, 174 (40%) had medium income, and 72 (16.55%) had high income, whereas Tessema (2020) narrated that high, medium, and low income were 27 (7.68%), 240 (68.38%), and 84 (23.93%) in Bahir Dar city, Ethiopia, respectively. In this categorical variable, a higher percentage collected the information from the most strategic groups. Three hundred and seventy-four respondents, or 85.98% of the total, reported using personal pipes for water.

In these categorical factors, a majority of comments were gathered from key groups that are extremely closely related to water usage, such as females (63.22%), married (82.76%), informal occupation (74.48%), owners (68.05%), and low income (43.45%), to produce more accurate results. According to Tessema (2020) around 47.29% female customers were dissatisfied with the availability of water in the Bahir Dar city, Ethiopia.

The respondents have a mean age of 37 (37.189) years with a standard deviation of 12.51 in Adama city. This showed that the research questionnaires were properly answered and were mature (Tessema 2020; Denantes & Donoso 2021). Representative households explained that the mean pipe water accessible per day (24 h) was 14.4 h per day with a standard deviation of 5.75 h per day. This indicated that respondents had access to the piped water at least half of the day.

The appearance of any service's physical infrastructure, tools, and staff is referred to as its ‘tangibility.’ The tangible aspects of a water supply service are represented by the physical reservoir, distribution network, collecting sites, and equipment. In terms of tangibility, 31.49% of customers reported being satisfied, 41.84% neutral, 26.67% dissatisfied, and 0.92% severely dissatisfied. This indicates that visible infrastructures of water supply systems are satisfying only 20% of respondents. The perception of the utility was that it had made its tangibility and amount of capacity known. In Adama city, 31.04% of the population expressed dissatisfaction due to the poor and non-existent water supply facilities due to functional and communication limitations. Tessema (2020) described that customers were 60.11% dissatisfied, 29.34% neutral, and 10.53% satisfied with the pressure of water supplied in Bahir Dar city.

Accessibility, communication, comprehension, care, and specialized attention are all examples of how a service provider demonstrates empathy towards their clients. According to the empathy data in Adama city, 56.55% of consumers were happy (56.32% were happy and 0.23% were very happy) because the employees have shown the consumers a high level of concern. Tessema (2020) discovered that customers were 23.65% dissatisfied, 22.51% neutral, and 21.36% satisfied with scheduling or timely communication on the water supply service in the Bahir Dar city.

To provide assurance, a company must have competent, courteous, credible, reliable, and trustworthy personnel. Clients indicated that they were satisfied with the competence, trust, and confidence of employees in approximately 50.34% of cases. However, 16.55% of respondents were displeased, 12.87% was neutral due to customers not getting sufficient courteous and credible in-service provisions, and 0.23% was severely dissatisfied. Because workers were overburdened and juggling various jobs due to a small number of employees, they sometimes lack personal attention. Tessema (2020) discovered that customers were 28.18% dissatisfied, 28.49% neutral, and 43.33% satisfied in Bahir Dar city.

The results of the study on customer satisfaction may be summed up as follows: customers were substantially more satisfied with assurance and empathy than they were with responsiveness in AWSSE's service offering for their clients. Similarly, the reliability and tangibility affected customer satisfaction with the provision of the services by AWSSE.

Table 6 indicates the chi-square test for independent variables with an overall satisfaction level of customers. Ninety (21%) male respondents and 178 (41%) female respondents were satisfied with AWSSE's services provision. This indicated that female customers were significantly more satisfied than male customers with existing water supply services provision given by AWSSE. A total of 8.16% females and 23.53% males were satisfied with water provision services; however, male customers were more satisfied than female customers (Tessema 2020). Also 221 (51%) married customers and 47 (11%) unmarried customers were satisfied. This reveals that married customers were significantly more satisfied than unmarried customers by AWSSE's services provision to their customers. However, customers with informal occupations were significantly more satisfied than customers with formal occupations. Similar studies conducted at Bahir Dar and Dere Tabor indicated that households who have informal occupation were more satisfied than households who have formal occupations (Tessema 2020; Tekile & Legesse 2023). In the case of educational status, customers with no education were significantly more satisfied than customers who are graduated and post-graduated. Customers living in their own houses were significantly more satisfied than customers living in rental houses. The root cause for their dissatisfaction is that water has been often coming at night. With main water sources, customers who have individual piped water were significantly more satisfied than others (shared water connection, public fountains, and public water taps). On the other hand, customers who used the public fountain and public taps were significantly dissatisfied due to non-functioning. The main concern for their dissatisfaction is that they are not able to get water when they need like customers who have individual water connections. Customers having low and middle income were significantly more satisfied than high-income customers.

According to Jebb et al. (2021), on the five-point Likert scale, a mean score below 3.39 is deemed low, a mean score between 3.40 and 3.79 is thought to be moderate, and a mean score beyond 3.79 is thought to be high. Hence, the result indicated that all mean of variables is below 3.39, which meant that the overall customer satisfaction is low (Table 7). Their dissatisfaction stems from several issues, including water supply disruptions in the daytime, billing problems, and a lack of timely responses to their inquiries.

The CSI indicated that reliability, tangibility, and responsiveness are below 50%, which implies that customers are unsatisfied, whereas empathy is above 50% and assurance is 50%. As shown in Table 8, customers are moderately satisfied with empathy variables. The overall CSI is 41%, which is below 50% but less than Woliso district customer satisfaction of 51.54% (Mekuriaw & Gurmessa 2020). According to Tufa & Abate (2022), the Ejere town's customers level of satisfaction with water supply services was less than 50%, and it denoted that 289 (77.3%) and 85 (22.7%) were dissatisfied and satisfied, respectively, in Ethiopia. Lack of clear and timely communication about service disruptions, billing issues, and other matters related to their water supply caused their dissatisfaction. According to Ramez (2012), the customer satisfaction is below 50%, which implies that customers are unsatisfied. Therefore, the overall CSI result revealed that customers are unsatisfied with water supply service quality due to water coming at night time, billing inaccuracy, and lack of awareness about the water interruption time. Kassa et al. (2017) denoted that South Ethiopia urban water supply customers were 43.27% dissatisfied, 9.83% neutral, and 46.90% satisfied. The overall CSI was lower than 50% (Table 8).

It is an advanced type of statistical analysis that shows the relationship between the variables under study. Regression and correlation have been used to examine the effect of dimensions of service quality on the customer's level of satisfaction (Ali et al. 2021).

The impact of each predictor variable on the response variable is demonstrated via bivariate correlation analysis. Each pair of study variables' correlation coefficients has been computed. The significance level, or the so-called p-value, can be used to evaluate the significance of the correlation between the variables if the correlation value is greater than 0.5 (Leonard 2020).

Consequently, Table 9 shows the relationship between the five service quality dimensions (reliability, tangibility, responsiveness, assurance, and empathy) and customer satisfaction. The correlation coefficients indicated the strength and the direction of these relationships. Reliability and tangibility have moderate positive correlations (0.4730 and 0.5689, respectively) with customer satisfaction, indicating that enhancing these dimensions can be a prime to higher levels of customer satisfaction. Responsiveness has a comparatively weaker positive correlation (0.3559), signifying a less significant impact on customer satisfaction. Both assurance and empathy have reasonable positive correlations (0.6378 and 0.5343, respectively) with customer satisfaction, implying that improving these dimensions can contribute to amplified levels of customer satisfaction. Generally, these correlation findings provide valuable insights into the importance of different service quality dimensions in influencing customer satisfaction.

A typical statistical method suitable for a continuous response is multiple linear regression since customer satisfaction was computed using the mean to make it continuous. The model comprises one continuous response and at least two predictors of any type. Therefore, a multiple linear regression model was used to examine the collective impact of service quality dimensions on customer satisfaction (Makwinja et al. 2019).

Basic Outputs of Multiple Linear Regression Model

The adjusted R² in Table 10 shows the degree of explanation of the dependent variable by the model (Leonard 2020). The result reveals that about 61% (R² = 0.6064) (Table 10) of the variability in customer satisfaction is explained by the set of different dimensions of service quality in the model.

An analysis of variance (F-test) as part of regression analysis shows whether the overall model is significant (Todorov et al. 2020). The significance level in the following table revealed that the regression model was highly significant to fit the data. The cumulative effect of the set of service quality dimensions on customer satisfaction is highly significant, but it does not show which predictor variable makes the model significant, which can be examined by the coefficient table (t-test) in Table 11.

Subsequently, the analysis of variance (ANOVA) table analysis indicated that the regression model's overall significance in predicting customer satisfaction, on service quality dimensions, is highly significant (Todorov et al. 2020). The F-statistic of 80.03 indicates that the dimensions collectively have a significant impact on predicting customer satisfaction. Moreover, the regression model accounts for a substantial amount of the observed variability in customer satisfaction, explaining 61.18% of the total variance. These discoveries verify the effectiveness of the regression model and underline the influential role of service quality dimensions in determining overall customer satisfaction.

The coefficient table, or t-test, of regression analysis, displays the effect of each predictor variable on the response variable. Referring to the significance level of each predictor variable revealed that all dimensions of service quality have significant (p-values 0.05) effects on customer satisfaction at a 5% level of significance (Leonard 2020). Among these tangibility, responsiveness and assurance have a highly significant (p-values 0.001) effect on customer satisfaction (Table 12). Moreover, all of the predictor variables except responsiveness have a positive effect on customer satisfaction, i.e., when the service quality dimensions get improved, the level of customer satisfaction with the water supply service also gets enhanced.

This study of the Adama town water utilities revealed five service quality dimensions through data analysis. Customers' opinions and satisfaction at Adama city were gathered from 435 respondents. To collect information for this study, eight categorical independent variables were considered. To produce more accurate results in these categorical factors, the majority of comments were gathered from key groups that are extremely closely related to water usage, such as females (63.22%), married (82.76%), informal occupations (74.48%), owners (68.05%), and low income (43.45%). Representative households stated that the average amount of pipe water available per day (24 h) was 14.4 h. To assess the status of customer satisfaction levels, Likert scale variables such as dependability, tangibility, responsiveness, assurance, and empathy were used. According to the results of reliability, tangibility, responsiveness, assurance, and empathy, 33.33, 26.67, 16.55, 31.04, and 20.46% were dissatisfied, respectively. According to the results of the Chi-square test of customer satisfaction for demographic characteristics, the mean of all variables is less than 3.39, indicating that overall customer satisfaction is low. The model's set of different dimensions of service quality explains approximately 61% (R² value of 0.6064) of the variability in customer satisfaction. The F-test results revealed that the regression model fit the data very well. The set of service quality dimensions has a significant cumulative effect on customer satisfaction. The overall CSI is 41%, which is lower than the 50% mark. Water supply disruptions at night and during the day, billing issues, inconsistency in supply, unscheduled supply, service failures, and a lack of timely response to inquiries all contributed to dissatisfaction and loss of trust. Due to the small number of employees, workers were often overburdened and juggling multiple jobs, resulting in a lack of personal attention. Addressing these issues necessitates effective communication, prompt resolution of billing issues, technical failure resolution, and responsive customer service.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.