Observing the COVID-19 pandemic SOPs and sales performance: Evidence from Khyber Pakhtunkhwa, Pakistan

Graphical abstract

Introduction

Women's entrepreneurship is a crucial driver of economic growth and sustainable development creating employment opportunities and reducing poverty. It brings economic prosperity, enhancing the well-being of women, and families through economic independence and empowerment (Lepeley, 2020). These notions are consistent with the United Nations’ 1st Sustainable Development Goal (SDG)- End poverty in all its forms everywhere (UN, 2020). However, most women's entrepreneurial engagements in developing and emerging countries centre on the undocumented sectors lacking legal, social, and benefit protections (Walker et al., 2022). In addition, women's entrepreneurial operations and performance were badly affected by the COVID-19 pandemic. As mentioned by Muhammad et al. (2023), the effects of the pandemic have been prevalent with health, social and financial consequences, including a decline in business revenues and performance (Rodríguez et al., 2024).

In Pakistan, the first Coronavirus-positive case was testified on February 26, 2020, signalling the government to impose lockdowns. Therefore, the masses were informed and educated about the virus through SOPs which is considered the best way to prevent and slow down the viral infection transmission. The COVID-19 pandemic SOPs mainly aimed to compel people and entrepreneurs to protect themselves and others from infection by keeping at least a one-meter distance, wearing a mask, washing their hands, or using sanitiser frequently. Moreover, due to the strict imposition of the lockdowns, people need to follow such SOPs in travelling, business operations, and social activities that may affect women's entrepreneurial sales performance.

Women's entrepreneurship plays a key role in enhancing living standards by alleviating poverty and has got the attention of research scholars, governments, and international institutions. However, Pakistan is among those countries that report a lower rate of female participation in businesses than some other countries. Since most of their business engagements are particularly exposed to shocks (Baird & Hill, 2020) from declining profitability and the fragile nature of operating in the informal sector (OECD, 2020) provides practical implications that are exacerbated by the pandemic and cause socioeconomic vulnerability (Vázquez-Martínez et al., 2021). Such pandemic effects vary within the business sectors and region to region. This study holds significance as women home-based entrepreneurs contribute to the well-being of their families paying health, education and other necessary expenditures. Recent studies reported that the pandemic decreased profitability as most women's businesses were exposed to economic shocks (Manolova et al., 2020) negatively affected their business growth. In contrast, another study's findings showed that these small businesses are resilient to crises because of the nature of their business activities largely financed by themselves and their families (Muhammad et al., 2023). Therefore, this study also has a theoretical contribution adding recent findings to the existing literature that how entrepreneurial sales performance is influenced by the SOPs.

From an economic point of view, Pakistan has an informal sector that accounts for 76 % of non-agricultural employment, with nearly equivalent percentages of women (77.7 %) and men (75.7 %) working in wholesale and retail trade (32.5 %), manufacturing (22.8 %), and community services (16 %) (PBS, 2017–18). The statistics further show that 38.7 % of the entrepreneurial women in the informal sector are engaged in trade activities, followed by manufacturing concerns (Malik et al., 2016). Women's involvement in such economic activities shows variation in business types comprising stitching and tailoring, dairy, hand embroidery, cosmetics, cloth, beauty parlours, grocery, etc. This study limited its analyses to dairy products, cloth, and cosmetics, which are considered established and emerging businesses (Muhammad et al., 2021a).

Several women's businesses were negatively affected during the pandemic lockdown period (Rodríguez et al., 2024; Jaim, 2021). On the other side, the COVID-19 pandemic's effect on the sales and income of micro-businesses varies, with some businesses doing relatively well, whereas others have come to ruin (Jaim, 2021; Muhammad et al., 2022). To restrict the virus spread the big markets in rural-urban areas were closed due to the imposition of lockdowns, except only shops selling essential goods. Furthermore, the district and local administrations were very active in implementing pandemic SOPs, officials fined and sealed those shops and markets that dealt in non-essential and luxury goods, except medical and grocery/vegetable stores remained open. Women in these businesses attempted to give a local focus on offering products to sustain sales but were compelled to follow/observe the guidelines/SOPs issued by the government. Such SOPs in travelling restrict entrepreneurs’ free movement to purchase material or attend business meetings. Moreover, in a sociocultural environment, customers are sensitive to the behaviour of women entrepreneurs and how they are dealt with. In addition, women entrepreneurs also need to follow SOPs in attending social activities representing their behaviours towards others. Therefore, how these SOPs affect women home-based entrepreneurs, this study attempts to examine such relationship.

The existing body of knowledge reports a diverse finding regarding entrepreneurial income and growth during the COVID-19 pandemic. However, to the best of the authors' knowledge, the impact of the COVID-19 pandemic SOPs on women's sales performance in Khyber Pakhtunkhwa has not been fully explored. Therefore, this study aims to fill this gap at the household level focusing on women home-based entrepreneurs operating in the sociocultural environment facing the issue of gender equality, access to finance, low social and economic empowerment etc. This study set the following objectives: 1) To examine the impact of observing the COVID-19 pandemic SOPs (travelling, dealing customers, and social activities) on women entrepreneurs' sales performance. 2) To understand how the relationship between SOPs and women entrepreneurs' sales performance is affected by other factors. 3) Does heterogeneity exist when examining the impact of SOPs on sales performance using sub-samples?

Theoretical context

This study is well-versed in three theoretical approaches: the “multiplier effect” of women's entrepreneurship (Lepeley, 2020), economic empowerment theory (Mapuranga et al., 2021) and pandemic effects. The multiplier effect demonstrates how women entrepreneurs’ satisfaction, happiness, and work-life balance affect women and their family's well-being. For instance, the well-being of their children, households, families, and society. Managing and balancing their entrepreneurial activities and “home shift” brought by the COVID-19 pandemic while these women entrepreneurs had additional responsibilities for caregiving due to the closure of educational institutions and health care demands.

Economic empowerment theory explains women's increasing involvement in entrepreneurial activities leading to economic independence (Bushra & Wajiha, 2015; Yu et al., 2019) and social empowerment (Muhammad et al., 2021b). Moreover, women's income role in a family is indispensable because it enhances the financial position of their families and contributes to the family's economic well-being (Muhammad et al., 2021a). Such home-based activities empower women for their economic participation within the cultural norms, which is 5th set goal (achieve gender equality and empower all women and girls) of SDGs. Hence, this study examines the presumed multiplier effect of women entrepreneurs during the COVID-19 pandemic whilst concentrating on women's economic empowerment derived from entrepreneurial income and sales performance.

In late 2019, the quick emergence of the coronavirus disease (COVID-19) pandemic is an infectious viral disease caused by the SARS-CoV-2. The masses were informed and educated about the virus through SOPs which is considered the best way to prevent and slow down the viral infection transmission. The people got stuck where they were; due to the strict imposition of the lockdowns, which were considered an effective tool to control the virus spread. Such measures have been imposed, resulting in different implications across the world. As a result, the trade between the nations (Shahzad et al., 2023) was suspended for an unspecified period, which damaged the supply chain and brought an acute shortage of necessities due to an imbalance of demand and supply factors which affected the business performance (Maaz & Ahmad, 2022; Castro et al., 2023). It can be concluded that the pandemic profound effects have not been limited to a specific sector only. However, its economic impacts are severe and will take time to recover. Therefore, this study considers how observing SOPs in travelling, dealing customers and social activities affect women entrepreneurs' sales performance amidst the COVID-19 pandemic.

Hypothesis development

The pandemic has multidimensional and substantial effects, impacting every walk of life (Muhammad et al., 2023). The pandemic has tested the medical, transportation, technological, financial, economic, social and industrial sectors. Rahmafitria et al. (2021) used a purposive sampling approach in Indonesia to examine the effect of physical distancing measures during travel. Their study findings showed that government sanctions or restrictions from family and friends strongly and significantly affect the travellers’ decision-making in a collectivist society affecting the hotel, travelling, and transportation industries negatively (Sánchez-Sánchez et al., 2024). This ultimately lowered the demand for agricultural and other products and reported excess supply, especially in rural areas. Su et al. (2022) applied a qualitative research design to interview textile apparel industry practitioners in Bangladesh and China. The study found that the COVID-19 pandemic has disrupted the supply chain depending on firm size, channels used, product type, and sourcing method. This may be due to the supply chain disruption in large sectors having an ultimate impact on the local supply chain that may affect their sales volume (Maaz & Ahmad, 2022). However, due to travel restrictions and to meet the demands of the local people, some of the merchants shifted their businesses to nearby localities meeting the people's demands. In contrast, due to the unavailability of public transport, such women entrepreneurs who use highly paid transportation recorded an increase in the cost, influencing their revenues adversely. Hence, the hypothesis can be inferred that:

H1:Following the COVID-19 pandemic travelling SOPs negatively affect women entrepreneurs’ sales performance.

Customers decide the fate of any business. Nowadays, every business is trying to enhance their customers not only locally but also by expanding their operations worldwide. In the initial stages of business growth, one of the main foci was mass production and sales, but the recent trends have focused on customer orientation. In the era of competition, customers have given value considering the king of the market. To achieve such goals, a public relations department has been established to satisfy customers to register complaints, solve problems, get timely feedback, etc. Specifically, small businesses' customer base mainly focuses on the national and local area markets. Especially for women home-based entrepreneurs, their main customers are mostly women themselves in sociocultural environments (Muhammad et al., 2020). Therefore, the people in the locality become a source of word-of-mouth advertisement where customers promote products/services based on their excellent experiences. Such a strategy is very effective when your entrepreneurial engagements are in the introduction stage and can't bear the huge cost of advertisement. The customers believe shopping locally is safe (Silva et al., 2023) due to the low chance of the virus spreading while observing SOPs. This shows customers are sensitive and can be switched easily due to stiff competition, availability of substitutes, products at low prices, safety behaviour, etc. Moreover, in a sociocultural context, customers are also sensitive and affected by women entrepreneurs' attitudes and behaviours by keeping distance, avoiding shaking hands, etc., following the COVID-19 pandemic SOPs. Hence, the hypothesis can be inferred that:

H2.Following the COVID-19 pandemic SOPs dealing with customers negatively affect women entrepreneurs’ sales performance.

The impact of the COVID-19 pandemic appears to have affected (Clendenning, 2023) women's businesses more quickly than the economic crunch of 2008–09 because of severe restrictions on trading and social activities. The sociocultural context is complex, especially where societal and religious factors dominate. The findings of Dagoudo et al. (2023) revealed that traditional and cultural practices enable the transfer of socially embedded informal knowledge. This shows how women's entrepreneurial activities are supported by informal in-family education, resulting from family members and social learning-specific skills while also experiencing the sensitivity of belonging to the family and society (Muhammad & Ximei, 2022). This indicates that women's business activities and sociocultural norms can't be separated and need parallel attention. However, it differs from region to region and within the culture as well. Especially, in urban areas, the social interaction is less than the rural due to busy work schedules. Mostly in urban areas, people have certain objectives and stay for a short time, while the rural places are the people's identity as, in most cases, their forefathers belong. At some point, social involvement (Sipahi Dongul & Artantaş, 2023) be used to promote business ventures as a strategy to improve business performance.

In contrast, the pandemic restricted people from participating in social and religious engagements to protect themselves and others from the virus spread. As a result, the religious activities restrictions negatively impacted the lives of the Muslims. As pandemic has affected not only the business sector but the social system as well (Alam et al., 2023). In contrast, such restrictions on social involvement may have some unknown consequences for women entrepreneurs, which need to be explored. Such relationships may vary due to socioeconomic and demographic factors of women entrepreneurs. In addition, the mean of the SOPs was calculated to test the main and interaction effects as reported in Fig. 1. Hence, the hypothesis can be inferred that:

Fig. 1.

Conceptual model of the study.

H3.Following the COVID-19 pandemic SOPs in social activities negatively affect women entrepreneurs’ sales performance.

H4.The causal effect of the COVID-19 pandemic SOPs on entrepreneurs’ sales performance is positively affected by moderators; a) sales satisfaction, b) age, c) education, and d) marital status.

Material and method

Study setting

A quantitative research design was applied to gather insights from women entrepreneurs in two districts of Khyber Pakhtunkhwa Province, Pakistan. Mardan and Charsadda are adjacent districts with a resident population of 2.37 million and 1.61 million, respectively, as reported in Table A1. Due to its hidden nature most of the women's home-based businesses are operating in the informal economy that lacks social security and other benefits. In addition, being aware of the cultural norms, and a resident of the place access to interviews was easy for the study's principal investigator. The study focused on women engaged in cloth, cosmetics, and dairy products. Information was gathered from district administration, training centres, and researcher reconnaissance survey list. The reconnaissance survey aimed to find women home-based entrepreneurs to ensure random selection and minimize sample bias. Finally, we identified 2500 women entrepreneurs from district Mardan and 1500 from district Charsadda.

Sampling procedure

A multi-stage sampling technique was adopted to select the sample size. At the start, Khyber Pakhtunkhwa province was purposively selected as one of the low engagement of women in entrepreneurial activities. In the second step, District Mardan and Charsadda were carefully chosen due to more population and central regions of the province. Due to the commercial and new trend of women's engagement, three business activities; cloth, cosmetics, and dairy product businesses were selected from the reconnaissance survey (Muhammad et al., 2023). Respondents were sampled based on more than one year of experience to compare their present entrepreneurial performance with the past. The probabilistic random sampling technique was used to select the participants from the researcher's survey list. By applying (Yamane, 1967), a sample size of 395 respondents was determined at a 95 % confidence level and ± 5 % margin of error (Eq.1). Details about sample size calculation are presented in (Appendix 1).

Data collection and instruments

The survey was conducted from June to August 2020 using a questionnaire from a minimum of 395 respondents. However, for precision and better representation, data were collected from 672 respondents. During the survey, the COVID-19 pandemic standard operating procedures (SOPs) were strictly observed. Face masks were distributed among respondents to wear and maintain social distancing (Sánchez-Sánchez et al., 2024) when filling out the questionnaire/interview. Female administrators were trained to perform their role systematically and consistently, not to affect their responses to minimize potential bias, and highly supervised by the principal investigator. A questionnaire was designed, composed of open and closed-ended questions with the same contents as a data collection instrument. The questionnaire was pre-tested on 40 women entrepreneurs who were not part of the existing sample for updating and simplifying the wording of the instrument.

Data analysis

The data were analyzed using bivariate and multivariate techniques. The descriptive statistics in Table A2 show the respondents' mean age is 34 years, representing almost 79.8 % married and residing 64.1 % in rural areas. Education as years of schooling is reported with a mean of 5.22. 59.8 % of the collected sample belongs to District Mardan as compared to Charsadda. The respondents' experience reports a mean of almost 6 years. However, the household size comprises 9 members. Most women entrepreneurs belong to the low family financial status, which is 43.8 %. In addition, 93.5 % of the respondents engaged have family support. Moreover, 61.8 % of women entrepreneurs showed an increase in their sales performance by offering products such as cosmetics, cloth, and dairy. In contrast, 50.3 % were satisfied with their sales, irrespective of their increase in sales performance. The pandemic SOPs travelling, dealing customer and social engagements report a mean of 2.68, 3.38, and 3.02, respectively. Moreover, this study used the mean of pandemic SOPs in mechanism analysis.

Reliability analysis of the scales

In empirical studies, an instrument needs to be reliable when multiple items measure a concept, factor, or construct. The reliability of the instrument was established to measure the internal consistency of the scale using Cronbach's Alpha as reported in Table 1. Cronbach's Alpha value lies between 0 and 1, however, the value (< 0.50) is not acceptable. Results showed that the SA construct has very good reliability (> 0.70) while TR and DC constructs have good reliability (0.694) and (0.688), respectively, which are very close to a very good benchmark of (0.70).

Table 1.

Reliability statistics.

Scale	Cronbach's Alpha	No. of Items
TR	0.694	3
DC	0.688	4
SA	0.758	5

Note: TR: Travelling SOPs, DC: Dealing customers, SA: Social activities.

Exploratory factor analysis (EFA)

The varimax rotation method applied principal component analysis (PCA) was used to perform EFA (Exploratory Factor Analysis). Variance in each dimension was assessed to ensure acceptable levels of explanation with the help of the commonality of the scale. We have set the minimum factor loading to 0.50 criteria.

An important step involved weighing the overall significance of the correlation matrix through Bartlett's Test of Sphericity, which indicates the statistical probability established a significant association among some of its components. The Keiser-Meyer-Olkin (KMO) and Barlett's Test of Sphericity results for the suitability of the factor analysis were found significant, X2(n = 762) =2136.32 (p < 0.001). The KMO Measure of Sampling Adequacy (MSA) value lies between 0 and 1. The value < 0.60 shows the sample is not adequate and remedial actions should be taken.

At last, the factor solution derived from this analysis yielded three factors, which account for 55.50 per cent of the variation. The factor loading results are reported in Table 2. In addition, the scree plot shows eigenvalues against the corresponding principal components as reported in Fig. 2. We have used the eigenvalue 1 for the number of factors solution.

Table 2.

EFA results.

Items	1	2	3
Travelling SOPs
TR1	0.846
TR2	0.693
TR3	0.678
Dealing customers SOPs
DC1		0.843
DC2		0.649
DC3		0.596
DC4		0.704
Social activities SOPs
SA1			0.841
SA2			0.556
SA3			0.589
SA4			0.713
SA5			0.536

Fig. 2.

Showing the number of factors.

Bivariate analysis

Chi-square (χ²) statistics test the relationship and strengths of association between categorical variables. Those variables measured on a continuous scale were classified to test for such relationships. In our case, we have classified data based on two categories of sales performance (increase/decrease). This allows us to compare categories or groups of other variables segregated by the sales performance.

Justification of the empirical models

In the case of quantitative dependent variables, the objective of the regression model is to evaluate its expected or mean value given the value of other independent variables. But when the “y" is in dummy form, then the model's objective is to find the probability of an event or something happening. That's why such models are referred to as probability models, as “y" follows a Bernoulli distribution which takes the value of 0 and 1. To study the influence of the COVID-19 pandemic SOPs on the sales performance of women home-based entrepreneurs. The dependent variable (Y) is binary in nature. Therefore, different models can be used to analyze the data as per the requirements of the study after fulfilling the assumptions of the models. The study has used both the linear probability model (LPM) due to its advantages. The LPM considers the ordinary least square method to estimate the model. Hence, it predicts the probability of an event occurring (in our case, an increase or decrease in sales performance) and reports that the effects of predictors on the probabilities are linear. The functional model established in a simple form is reported in (Eq. (2)).

Where the dependent variable Y is a function of deterministic component βX and non-deterministic component (error term) μ. The expanded functional form is reported in (Eq. (3)) for better understanding and interpretation.

Where; Salesperformancei = 1 reports an increase, 0= no increase; β0 = constant; β1−3= represents the coefficient of the ith predictors (TR= Traveling, DC= Dealing customers and SA= Social activities measured on a five-point Likert scale from 1 to 5; Zi = Control variables including respondents' age, schooling, marital status, location, district regions experience, household size, family financial status, family support, business activities and sales satisfaction; and μ= error term.

To test for moderation, this study presents the two way interaction model (Eq. (4)).

Whereas Yi is the dependent variable (sales performance); β0 = constant; β1−3= represents the coefficient of the ith predictors; X1i = mean SOPs: X2i = moderators (satisfaction, age, education and marital status): X1iX2i = two way interactions: Zi = Control variables; and μ= error term.

This study also used binary logistic regression to check the robustness of the study findings. In our case, the logit function model estimates the relationship among the binary dichotomous dependent variable (increase or decrease in sales performance) and the other binary, ordinal, or continuous explanatory variables by assessing probabilities. (Eq. (5)).

Where; Yi = 1 reports an increase in sales performance, 0= decrease; α = constant; βi= represents the coefficient of the ith predictors; Xi= predictors (travelling, dealing customers and social Activities); Zi = Control variables (mentioned earlier) and μ= error term.

Common method bias

The survey research study is affected by common method bias (CMB), resulting from data collected using the same source (Podsakoff et al., 2012). To address this potential issue by providing reliable and valid results, this study conducted a “Harman single factor test” (Podsakoff et al., 2003). The computed result showed 25.79 % of the total variance explained by a single factor which does not account for a substantial variance confirming that common method variance is not an issue. In addition, we also conducted a multicollinearity test using variance inflation factor (VIF) (Table 4) and the reported values were found less than 5, demonstrating that there is no multicollinearity problem in the model.

Heterogeneity analysis

The study has also tested the heterogenous effects to deepen the analysis for better insights and variation in findings. For instance, the sample was segregated into different categories based on respondents’ location, district region, family financial status, family support, business activities, and sales satisfaction.

Results and discussion

The findings of the chi-square-χ² test of association based on the sales performance decrease/increase were reported in Table 3. A series of cross-tabulations (chi-square test) of sales performance increase/decrease by predictors were estimated first. Women entrepreneurs' age, rural/urban location, marital status, the district they belong to, household size, family financial position, family support, business activities and sales satisfaction were found to be statistically significant. In the age group of 30–34, 35 per cent reported a decrease in sales. Most of the respondents were not formally educated. By taking into consideration marital status, nearly 83 per cent of the women who reported an increase in sales performance were married. Suggesting that women home-based entrepreneurs were necessity-based, helping to support their families and children. In the sales performance (decrease), 70.4 per cent of the respondents reside in rural areas, while in the sales performance increase group, 66 per cent of the respondents belong to district Mardan.

Table 3.

Bivariate association between sales performance increase/decrease and other variables.

Study variables	ᵡ²	Decrease		Increase
		n=257	%	n=415	%
Age (Year)	9.27^⁎⁎
22–29		74	(28.8)	88	(21.2)
30–34		90	(35.0)	130	(31.3)
35–39		63	(24.5)	129	(31.1)
40–51		30	(11.7)	68	(16.4)
Schooling	4.41
No formal		87	(33.8)	117	(30.4)
Primary		73	(28.4)	120	(28.4)
Middle & High		59	(23.0)	123	(29.6)
Higher Secondary & University		38	(14.8)	57	(13.6)
Marital status	5.60^⁎⁎
Single		64	(24.9)	72	(17.4)
Married		193	(75.1)	343	(82.6)
Location	07.16^⁎⁎⁎
Urban		76	(29.6)	165	(39.8)
Rural		181	(70.4)	250	(60.2)
District	17.36^⁎⁎⁎
Mardan		128	(49.8)	274	(66.0)
Charsadda		129	(50.1)	141	(34.0)
Business experience	01.30
1–4		81	(31.5)	146	(35.1)
5–7		109	(42.4)	174	(42.0)
8–22		67	(26.1)	95	(22.9)
Household size	39.52^⁎⁎⁎
4–6		95	(37.0)	92	(22.2)
7–8		95	(37.0)	125	(30.1)
9–11		45	(17.5)	95	(22.9)
12–21		22	(08.6)	103	(24.8)
Family financial status	8.46^⁎⁎
Below average		128	(49.9)	166	(40.0)
Average		94	(36.6)	162	(39.0)
Above average		35	(13.5)	87	(21.0)
Family support	55.29^⁎⁎⁎
No		40	(15.6)	04	(01.0)
Yes		217	(84.4)	411	(99.0)
Business activity	145.34^⁎⁎⁎
Cloth		45	(17.5)	234	(56.4)
Cosmetics		53	(20.6)	105	(25.3)
Dairy		159	(61.9)	76	(18.3)
Sales satisfaction	370.62^⁎⁎⁎
No		249	(96.9)	85	(20.5)
Yes		08	(03.1)	330	(79.5)

Source: Author's survey 2020, *p < 0.10; ^⁎⁎p < 0.05, and ^⁎⁎⁎p < 0.01, n = 672: Due to a small number of counts, Fisher Exact test statistics were reported for sales satisfaction and family support variables.

The household size comprised of 4–6 family members showed a decrease, while the increased members of the household size reported an increase in sales performance. Around 50 per cent of the women entrepreneurs who reported a decrease in sales, their family belonged to below average financial status. This indicates that family financial status matters in the case of women entrepreneurship (Muhammad et al., 2021b). Those respondents who received family support reported a 99 per cent increase in sales performance. The cloth businesswomen reported a 56.4 per cent increase, while dairy products showed almost a 62 per cent decrease in sales. It is not necessary that all those women entrepreneurs who reported an increase in sales performance should be satisfied as the findings reported that 79.5 per cent of the entrepreneurs who showed an increase were satisfied with their sales performance.

Baseline model results

This study used the LPM to test the relationship between entrepreneurial sales performance and COVID-19 pandemic SOPs. The SOPs construct comprises three indicators. Such as travelling SOPs, SOPs dealing customers, and SOPs followed in social activities. The results in Table 4 showed that the COVID-19 pandemic travelling SOPs H1 (β = −0.064; SE=0.015; p < 0.01), SOPs dealing customers H2 (β = −0.065; SE= 0.013; p < 0.01), and SOPs observing in social activities H3 (β = −0.035; SE=0.015: p < 0.05) have a significantly negative impact on sales performance. Hence, H1, H2 and H3 are accepted. Our findings support the study of Muhammad et al. (2023) and Rahayu et al. (2023) that women businesses were negatively affected by the COVID-19 pandemic. Interestingly, the effects of dealing customers were found to be more severe than the travelling and social activities. This indicates that customers are very sensitive to the entrepreneurs’ behaviour and can switch very easily.

Table 4.

Impact of pandemic SOPs on the sales performance.

					n=672
Variables	Coefficient	Robust SE	p-value	VIF	Hypotheses decision
Travelling	−0.064	0.015	0.000	1.44	Accepted
Dealing customers	−0.065	0.013	0.000	1.18	Accepted
Social activities	−0.035	0.015	0.017	1.40	Accepted
Controls
Age	0.007	0.002	0.008	2.32
Schooling	−0.003	0.002	0.203	1.61
Marital status	−0.017	0.029	0.570	1.35
Location	0.028	0.025	0.274	1.22
Experience	0.010	0.004	0.034	2.16
Household size	0.009	0.002	0.001	1.12
Family financial status
Below average®
Average	−0.018	0.023	0.446	1.27
Above average	0.008	0.031	0.783	1.26
Family support	0.258	0.040	0.000	1.14
Business activity
Dairy®
Cosmetics	0.408	0.038	0.000	1.83
Cloth	0.447	0.037	0.000	2.36
Sales satisfaction	0.498	0.033	0.000	1.50
District	0.001	0.024	0.957	1.19
Constant	−0.013	0.113	0.907
R^2	0.698

Note: *, ** and *** correspond to the significance level of 10 %, 5 %, and 1 %, respectively. This study reports Huber-White heteroskedasticity robust standard errors. VIF: Variance inflation factor.

Considering the individual-level control variables the findings showed that a unit increase in age, business experience and household size significantly positive impacts on the probability of women entrepreneurs' sales performance. Family support and satisfaction from sales were found to be statistically significant showing that women entrepreneurs received family support and were satisfied with sales increasing the probability of sale performance. This study supports the findings of Gashi Nulleshi (2024) that business experience and support from family positively influence women's businesses. These entrepreneurs were involved in different business activities. The findings reported that compared to the reference category of dairy products, cosmetics and cloth entrepreneurs showed increased sales performance.

Interaction effects

The continuous variables were mean-centred to predict sales performance when all predictors were 0. Centring continuous predictors, the interaction term and its associated test results are the same for models using the raw predictors or centred predictors. However, the main differences are the models’ intercept and slope coefficients for the individual predictors.

The main effects of SOPs in Table 5 showed a negative effect on sales performance. However, the interaction effects of SOPs and satisfaction H4a (β = 0.070; SE= 0.039; p < 0.10), and SOPs and marital status H4d (β = 0.091; SE=0.053; p < 0.10) significantly positive representing that observing the COVID-19 pandemic SOPs (equal 0 at its mean) on the sales performance were conditional on satisfaction from sales and marital status (married). Therefore, hypotheses H4a and H4d are accepted. On the other hand, the interaction effects of SOPs and age H4b (β = −0.005; SE= 0.003; p < 0.10), and SOPs and education H4c (β = −0.011; SE=0.006; p < 0.10) on entrepreneurial sales performance were found significantly negative. Hence, hypotheses H4b and H4c are rejected. Our findings support the findings of (Muhammad et al., 2023) and OECD (2020) that the pandemic outbreak has diverse financial implications for entrepreneurs worldwide, but the socioeconomic effects have been mainly severe for women operating in the informal economy.

Table 5.

Interaction effects.

					n=672
Variables	Model (1)	Model (2)	Model (3)	Model (4)	Hypothesis decision
Sops_c	−0.179***(0.027)	−0.148***(0.020)	−0.225***(0.030)	−0.362***(0.046)
Satisfaction	0.517***(0.026)
Sops_c x Satisfaction	0.070*(0.039)				Accepted
Age_c		0.009***(0.001)
Sops_c x Age_c		−0.005*(0.003)			Rejected
Edu_c			0.012***(0.004)
Sops_c x Edu_c			−0.011*(0.006)		Rejected
Marital status				−0.004(0.037)
Sops_c x Marital status				0.091*(0.053)	Accepted
Control	Yes	Yes	Yes	Yes
Constant	−0.263***(0.055)	−0.277***(0.050)	0.276***(0.010)	−0.171***(0.066)
R^2	0.674	0.670	0.240	0.519

Note: *, ** and *** correspond to the significance level of 10 %, 5 %, and 1 %, respectively. This study reports Huber-White heteroskedasticity robust standard errors in parenthesis.

As shown in Table 5, the impact of the SOPs on the sales performance differs according to the level of moderators/interactions reported above. Here the question arises of how it differs, this is not clear due to the positive/negative sign of the predictors and interaction. Therefore, as mentioned by Dawson (2014) the size and precise nature of such effect is difficult to divine after examining the coefficient alone particularly when one or more coefficients move in opposite directions, or the standard deviations of predictors and moderators are very different. Dealing with such issues provides, first, easy interpretation through visualization. For such motive, this study has calculated predicted values of sales performance under different conditions/levels of SOPs and moderators to report the predicted relationship/simple slopes between SOPs and sales performance at different levels of moderators.

The study has taken the one standard deviation above and below the mean as used generally. In our case, we have tested simple slopes to probe the significant relationship between the variable of interest at the given value of below and above the standard deviation. Fig. 3, (A) satisfaction, (B) marital status, (C) age, and (D) education show the plotted slopes.

Fig. 3.

Probing interactions with slopes.

Heterogeneous effect

Regional-based heterogeneity

This study has split the sample data into districts and rural and urban locations that help to better understand and compare the regional variation in the results. The results reported in Table 6 showed that following the travelling SOPs, there is a significant variation between urban 10.3 per cent (p < 0.01) and rural 5.9 per cent (p < 0.01), but no variation was found in district regions. Moreover, dealing customers indicated that there is a significant difference between rural-urban locations and district regions by following pandemic SOPs. In addition, social activities pandemic SOPs, women entrepreneurs in district Mardan 2.9 per cent (p < 0.10) and district Charsadda 5.3 per cent (p < 0.05) also reported heterogeneous effects regarding sales performance. Such differences may be observed due to the socioeconomic characteristics of the people. As reported by Muhammad and Ximei (2022), rural-urban disparity and differences exist among women home-based entrepreneurs due to education levels, access to facilities, population density, income etc.

Table 6.

Regional-based variation.

Variables	1- Location		2- District
	Urban	Rural	Mardan	Charsadda
Travelling	−0.103***	−0.059***	−0.060***	−0.060***
	(0.025)	(0.019)	(0.020)	(0.023)
Dealing customers	−0.092***	−0.046**	−0.049***	−0.113***
	(0.022)	(0.018)	(0.016)	(0.024)
Social activities	0.008	−0.056***	−0.029*	−0.053**
	(0.025)	(0.019)	(0.018)	(0.024)
Controls	Yes	Yes	Yes	Yes
Constant	0.350*	0.011*	0.077	−0.081*
	(0.190)	(0.129)	(0.133)	(0.188)
N	241	431	402	270
R^2	0.669	0.742	0.742	0.656

Note: *, ** and *** correspond to the significance level of 10 %, 5 %, and 1 %, respectively. Huber-White Heteroskedasticity robust standard errors are enclosed in parenthesis. Control variables are the same as reported in Table 4.

Business activity-based heterogeneity

The study participants were involved in different entrepreneurial activities to support their families. Findings in Table 7, showed that the COVID-19 pandemic SOPs negatively influenced women's entrepreneurial sales. To test for the individual constructs effect, observing the travelling pandemic SOPs, we have found a significant variation between cloth 12.4 per cent (p < 0.01) and dairy 6.6 per cent (p < 0.01) businesswomen. Almost the cloth 5.3 per cent (p < 0.01) and dairy 4.1 per cent (p < 0.05) findings were similar, but a significant variation was found with cosmetics 14.3 per cent (p < 0.01). In the case of social activities, cosmetics and dairy women entrepreneurs experienced a smile variation, reported 6.5 per cent (p < 0.10) and 5 per cent (p < 0.05), respectively. This difference may be due to the residence of the women entrepreneurs. For instance, most of the dairy businesswomen reside in rural areas where the population density is low due to which the pandemic SOPs may not be strictly observed compared to urban areas. Our study's findings align with Liu et al. (2022) that the COVID-19 pandemic has brought product competition which can affect the businesses' sales performance (Huang, 2023).

Table 7.

Business activity-based variation.

Variables	1- Cloth	2- Cosmetics	3- Dairy
Travelling	−0.124***	−0.006	−0.066***
	(0.024)	(0.029)	(0.020)
Dealing customers	−0.053***	−0.143***	−0.041**
	(0.019)	(0.030)	(0.018)
Social activities	−0.014	−0.065*	−0.050**
	(0.021)	(0.035)	(0.024)
Controls	Yes	Yes	Yes
Constant	0.574***	0.686**	0.143*
	(0.164)	(0.336)	(0.153)
N	279	158	235
R^2	0.539	0.659	0.818

Note: *, ** and *** correspond to the significance level of 10 %, 5 %, and 1 %, respectively. Huber-White Heteroskedasticity robust standard errors are enclosed in parenthesis. Control variables are the same as reported in Table 4.

Heterogeneity based on the family financial status

Women entrepreneurs engaged in different activities belong to diverse family financial backgrounds as reported in Table 8. The findings showed a negative influence of the COVID-19 pandemic SOPs on the sales of women entrepreneurs. To study the individual effect of observing the pandemic SOPs, travelling showed a significant variation based on family financial status; below average 6.2 per cent (p < 0.01), average 5.2 per cent (p < 0.05) and above average 8.8 per cent (p < 0.05). Dealing with customer women entrepreneurs reported significant and almost the same results. However, for women whose family financial status is reported as below-average, their social activities have a significantly negative impact on sales performance. Such differences may be aroused due to the size, experience and entrepreneurial activities they are involved in. This shows that entrepreneurial activities are not confined to specific people and their family financial status. Our findings contradict the findings of Miroshnychenko et al. (2024) that family business performance is higher than non-family businesses during the COVID-19 pandemic.

Table 8.

Family financial status-based variation.

Variables	1- Below average	2- Average	3- Above average
Travelling	−0.062***	−0.052**	−0.088**
	(0.023)	(0.025)	(0.035)
Dealing customers	−0.067***	−0.065***	−0.058
	(0.020)	(0.021)	(0.038)
Social activities	−0.049**	−0.016	−0.021
	(0.024)	(0.024)	(0.034)
Controls	Yes	Yes	Yes
Constant	0.346*	0.297*	0.734**
	(0.181)	(0.182)	(0.318)
N	294	256	1225
R^2	0.703	0.737	0.646

Note: *, ** and *** correspond to the significance level of 10 %, 5 %, and 1 %, respectively. Huber-White Heteroskedasticity robust standard errors are enclosed in parenthesis. Control variables are the same as reported in Table 4.

Heterogeneity based on sales satisfaction

A significant variation between satisfaction with sales and no satisfaction with sales was found as reported in Table 9. Women entrepreneurs observing travelling SOPs in case of satisfied with sales 2.8 per cent (p < 0.05) and non-satisfied 8.1 per cent (p < 0.01) showed a significant variation. To examine the individual effect of dealing customers, satisfied women entrepreneurs reported 2.1 per cent (p < 0.05) and non-satisfied 10.9 per cent (p < 0.01). Regarding social activities, the non-satisfied women entrepreneurs reported a significant 7.7 per cent decrease in the probability of sales performance. Generally, if we consider the variation, the non-satisfied women entrepreneurs reported significantly different results than satisfied women entrepreneurs. Even though their sales have increased during the COVID-19 pandemic, they still are not satisfied with their sales performance. Confirming that an increase in sales volume doesn't always lead to sales satisfaction.

Table 9.

Satisfaction and no satisfaction-based variation.

Variables	1- Satisfied		2- Not satisfied
	Coefficient	Robust SE	Coefficient	Robust SE
Travelling	−0.028**	0.011	−0.081***	0.023
Dealing customers	−0.021**	0.009	−0.109***	0.027
Social activities	−0.018	0.013	−0.077***	0.025
Controls	Yes		Yes
Constant	0.917***	0.170	0.652***	0.209
N	338		334
R^2	0.259		0.444

Note: *, ** and *** correspond to the significance level of 10 %, 5 %, and 1 %, respectively. Robust SE; Huber-White Heteroskedasticity robust standard errors. Control variables are the same as reported in Table 4.

Robustness check

In addition to heterogeneous effects, this study further applied binary logistic regression as a robustness check to analyze the econometric model sensitivity to some assumptions. We aimed to cross-validate the study results to uncover the specific relationship among variables. We have used the marginal effects instead of the raw coefficients for model results in Table 10. The results showed that the magnitude of the marginal effects of the individual predictors observing the COVID-19 pandemic SOPs (travelling, dealing customers and social activities) were found the same confirming that our findings are robust in nature.

Table 10.

Robustness check.

			n=672
Variables	Marginal effects	Robust standard errors	p-value
Travelling	−0.045	0.010	0.000
Dealing customers	−0.059	0.010	0.000
Social activities	−0.032	0.014	0.023
Controls	Yes
Constant	−6.266	2.055	0.001
Pseudo R^2	0.754

Note: This study reports Huber-White heteroskedasticity robust standard errors. Control variables are the same as reported in Table 4.

Conclusion and implications

Generally, the COVID-19 pandemic has a multidimensional consequence globally, while women entrepreneurs in Khyber Pakhtunkhwa, Pakistan also experienced the economic effects in terms of sales volume. Findings suggest the resilience, agility and multiplier effects of women entrepreneurs in the face of sociocultural, institutional, and economic constraints faced during the COVID-19 pandemic. Observing the COVID-19 pandemic SOPs negatively affects women entrepreneurs’ sales performance. However, their sales performance varies after considering the interaction effect of age, education, marital status, and sales satisfaction with the heterogeneous effect of regional variation, business activities they involve, family financial position and satisfaction from sales during the pandemic. By receiving moral and financial sustenance from their family members, community and government, these women's economic activities add their part to the development of the region by mainstreaming the state economy, thereby increasing the inclusive progress of the society. Both in the short and long-run, these entrepreneurs will increase productivity, leading to economic growth by generating new employment chances enabling these women's engagement to be further competitive and resilient beyond the current situation of the COVID-19 pandemic. This study adds numerous key contributions, both empirical and theoretical as below.

Theoretical implications

Our study findings provide theoretical implications for a better understanding of the women's entrepreneurial performance influenced by observing the COVID-19 pandemic SOPs. We lend support to the multiplier effect of women's entrepreneurship (Lepeley, 2020) and how the satisfaction, and happiness of women entrepreneurs influence their own and the well-being of their households and society. However, in return, women receive support from their families as a multiplier effect. This shows the multiplier effect is one of the major reasons why women's entrepreneurial engagements aim to contribute to family, and society by enhancing their entrepreneurial sustainability. Findings suggest that observing SOPs negatively affects women's entrepreneurial performance. However, the married marital status and satisfaction from sales act as crucial catalysts showing the resilience of women entrepreneurs amidst the COVID-19 pandemic. Therefore, we add diverse findings to the existing body of knowledge on how small business performance has been affected by the global crisis of the pandemic. Moreover, we also add recent findings supporting the economic empowerment theory that women's involvement in entrepreneurial engagements and employment opportunities leads to economic and social independence (Bushra & Wajiha, 2015; Muhammad et al., 2021a).

Practical implications

Based on the findings, this study provides practical and policy implications for the government and other stakeholders. First, during the unexpected global crises, we present a sound empirical dataset on the experiences of women entrepreneurs. Particularly, women entrepreneurship has been adversely affected by observing the COVID-19 pandemic SOPs (travelling, dealing customers and attending social activities). However, entrepreneurs reported an increase in sales volume that shows that women played a key role by contributing to households’ expenditures in crisis from their home-based businesses. Women support their families by experiencing hardships where their male counterparts (father, brother, husband and sons) are mostly unemployed due to the COVID-19 pandemic lockdowns or closure/restricted business activities. In return, they received support from their family contributing to the multiplier effect of women entrepreneurs. Second, the cosmetics and cloth entrepreneurs were less affected showing the resilience of women's small businesses during the pandemic. Therefore, women's home-based entrepreneurship must be supported by the government in framing friendly policies within the cultural context: 1) Being in home-based entrepreneurship women don't need to go out while serving the local people by offering the products in a socially responsible manner; 2) Women entrepreneurial activities can finance their family expenditure that will reduce the burden on men, especially in developing and low-income countries; 3) These economic activities will help reduce gender discrimination and empower them to have a say in decision-making. Third, home-based economic activities influence women's self-esteem and happiness through flexible business hours by managing work-life balance and enhancing the welfare of their households. Finally, the increase in sales shows that people buy from these entrepreneurs by following SOPs with a minimum risk of virus spread. On the other hand, the increase in sales during the pandemic would enable them to explore more opportunities, access to useful resources, achieving financial independence and self-reliance through entrepreneurship.

Study limitations and future work recommendations

This study carries some limitations. First, the respondents of the study were women home-based entrepreneurs from two districts engaged in informal activities. Second, entrepreneurial performance was measured using an increase or decrease in sales volume perceived by women entrepreneurs. Finally, this study adopted a questionnaire-based survey design in a specific context that lacks in-depth insights to understand the phenomenon.

Future studies may investigate a large dataset including more regions and provinces. More indicators of social capital, financial capital and family support would add new insights to the existing body of knowledge on women's entrepreneurship. In addition, the qualitative and longitudinal research design may add better insights into the long-term impacts of COVID-19 SOPs on women entrepreneurs.

CRediT authorship contribution statement

Said Muhammad: Writing – review & editing, Writing – original draft, Validation, Software, Methodology, Formal analysis, Data curation, Conceptualization. Zhichao Yin: Writing – review & editing, Supervision, Project administration, Methodology, Investigation, Conceptualization. Muhammad Haroon Ur Rasheed: Writing – review & editing, Visualization, Software, Investigation, Data curation, Conceptualization.

Special Acknowledgement

Authors are thankful to the study respondents for their time. We are grateful for the feedback of the editor and anonymous reviewers in improving this study.