The global commerce architecture experienced a seismic change in the past decade, moving toward a decentralized and borderless digital economy as opposed to the centralized, brick-and-mortar institutions of the 20th century. First among these transformations was the emergence of dropshipping, a highly process-efficient retail delivery system in which merchants took care of sales and marketing, while third-party suppliers handled stocking and logistics (Gleim et al., 2013; Alsuliman and Elrayah, 2021). This business model required a new kind of labor that was flexible, cost-effective, and able to work across distance while handling complex backend processes. As such, the Philippines became part of the back office of the global internet economy and became a source of Virtual Assistants (VAs) for shop owners in the United States, Europe, and Australia.

Beerepoot and Lambregts, (2015) defined this change as the emergence of the so-called micro-multinational. These dropshipping organizations were different from conventional multinational businesses because formal Human Resource (HR) departments were usually absent, and the businesses were often managed by owners with minimal capital and limited management training. Dropshipping business model has been simplified to be efficient; however its underlying processes are usually weakly designed by humans. Gig economy accelerated productivity more quickly than organizational changes to human resources (HR) led towards management styles that involved converting traditional job descriptions as employees became freelancers with less oversight through automation systems instead. 

This study was urgent, so it needed to be contextually relevant with respect of how special is our position as a country on world labour markets. This country is an example of being one that was historically known as a BPO hub and had workers who were culturally adapted for western services. Nevertheless, employees leaving office positions at a corporate level for work from home were highly vulnerable targets. According to Graham et al. (2017) digital labor platforms (DLPs) such as Upwork and OnlineJobs.ph increased access to global work but also tended to commodify labor, deprive workers of bargaining power, and generate a race to the bottom in terms of wages and working conditions. 

A major gap noted by researchers is that there are conflicts among western management styles with local Filipino working attitudes/values. Philippine HR practice is traditionally associated with the concept of genuine care, or Malasakit, and warm organizational culture, which has been practiced in established companies like Sun Life (Alfonso, 2023). Malasakit is based on an interdependent relationship between employers/employees; employees are viewed like members belonging to their companies' families whereas they respond by being loyal towards them. On contrary sharply, the e-commerce world is highly transaction oriented. These groups often adopted roe (result only work environment) that valued productivity over interaction. It is usually at odds to pakigkasam (social harmony). According to Cultural Dimensions Theory, the high-context communication characteristic of Filipino culture, where meaning is often based on non-verbal expressions and relational connection, did not always fit Western store owners who were low-context and direct instructional. Because there were no physical places for people's interactions so they communicated by writing messages; this created a contractual relationship that wasn't as friendly with customers leading towards high rates of turnover among va's. HR vacancies of hiring/selection were similarly transformed by way of drastic shift away from traditional credentials based hiring practices. The academic qualification of an individual was usually considered to be like some kind of filter for employment within our country's business world locally. However, Sullivan, (2013) also noted that this has changed with the rise of skills-based hiring. The study also discovered an aspect known as ‘trial by fire' hiring for dropshippers; candidates have been tried out with work assignments rather than qualifications and interview processes alone. The trend was developed due to issues with technology that prevented employees from checking their skills remotely, at an off-hours time zone. It's hard for sure if they can handle things like managing a dashboard of shopify products orders shipping refunds ads etc while also being able to do some back-end work as well. Lack of direct mentoring exacerbated it as most va's needed knowledge about system operation problems that had no hands-on training from a manager at their stores either (Sami, 2021).

In this context, work samples and trial tasks became substitutes for traditional interviews because they allowed employers to observe practical ability, speed, accuracy, and problem-solving under pressure. In his research on online labor markets Horton, (2017) observed the increasing reliance on work samples or trial tasks among employers instead of interviews to filter job applicants.

Research Problem

This was in spite of the economic importance of the VA industry which still had a sharp documentation gap. Conventional HR theories are tailored to the centralized corporate organizations, and only a limited number of studies have analyzed the concept of gigified HR of the micro-enterprises. This paper aimed to fill this gap by examining the employee lifecycle of the Filipino VAs within the dropshipping niche on an end-to-end basis. The misfitting of the highly technical requirements of the industry and the informal, primitive HR practices was the main issue.

In particular, the questions that were considered by the study were as follows:

1. Recruitment: What was the process of VAs sourcing and screening, and to what degree does the use of trial tasks substitute conventional interviewing?
2. Onboarding: In a asynchronous are, how was knowledge is being transferred and what is the effect of no mentorship?
3. Performance Management: What was the impact of the utilization of the algorithmic surveillance tools on the perception of the VAs in terms of trust, autonomy, and job satisfaction?

Theoretical Framework

In order to strictly examine the data, the study was based on the convergence of two complementary theories explaining the conflict between remote control and human motivation, namely: Principal- Agency Theory and Self-Determination Theory.

• Fig. 1: Theoretical Framework Diagram.
  
  
  Principal-Agent Theory (Jensen & Meckling, 1976)

  The economic theory is concerned with the connection between a Principal (Store Owner) and an Agent (the VA). The main conflict in this relationship is the “Agency Problem that is created when the Principal cannot confirm what the Agent is exactly doing a frequent problem with remote work

• Application to the Study: To prevent the possibility of the principal being tempted to accept the Principal-agent relationship (shirking) or cheating, the Principal establishes stringent surveillance systems and result-conditional contracts. This theory justifies why the employer is using surveillance instruments such as Time doctor and hard nosed SOPs. It puts the HR practices in a perspective of risk-reduction measures aimed at making sure that the VA will act in the best interest of the store owner.

Self-Determination Theory (Deci & Ryan, 1985)

Whereas Principal- Agency Theory describes the control of the employer, Self-Determination theory (SDT) describes the psychological response of the employee. SDT states that to become motivated and satisfied, workers must have three fundamental psychological needs, namely Autonomy (control over their work), Competence (the sense of being effective), and Relatedness (the sense of being connected to others).

• Application to the Study: The study hypothesizes the control mechanisms that are hypothesized by the Principal- Agency theory to be in operation result in active violation of needs identified under SDT. Equivalent surveillance, Autonomy compromises algorithmic surveillance; on boarding, Relatedness, asynchronous, and isolated undermines Autonomy. This theoretical conflict underlines the reasons as to why VAs are experiencing burnout and transactional loyalty even though they are being paid wages that are competitive.
  • Self-Determination Theory (Deci & Ryan, 1985)

Whereas Principal- Agency Theory describes the control of the employer, Self-Determination theory (SDT) describes the psychological response of the employee. SDT states that to become motivated and satisfied, workers must have three fundamental psychological needs, namely Autonomy (control over their work), Competence (the sense of being effective), and Relatedness (the sense of being connected to others).

• Application to the Study: The study hypothesizes the control mechanisms that are hypothesized by the Principal- Agency theory to be in operation result in active violation of needs identified under SDT. Equivalent surveillance, Autonomy compromises Algorithmic surveil-lance; on boarding, Relatedness, asynchronous, and isolated undermines Autonomy. This theoretical conflict underlines the reasons as to why VAs are experiencing burnout and transactional loyalty even though they are being paid wages that are competitive.

Research Design

The study used a qualitative study design. A study was appropriate because the research focused on a bounded phenomenon: the HR practices experienced by Filipino Virtual Assistants (VAs) working for global dropshipping or e-commerce micro-enterprises, consistent with (Creswell, 2013; Yin, 2018). Thematic analysis was used to identify recurring themes in recruitment, onboarding, performance management, and retention from participant narratives, based on Braun and Clarke.

Participants and Selection Criteria

Purposive max variate sample was utilized for our research so as not find any specializations of va roles clients markets products niches etc. The participants must be filipinos, philippine workers who are at least 21 years old; have worked for at least 6 months as a VA for a foreign-owned dropshipping or e-commerce company; have direct experience with at least two HR-related activities like recruitment screening, onboarding, training, performance monitoring, client communication, compensation practices, and retention issues. The last sample was comprised of 12 Filipino VAs. The sample size was considered acceptable for an exploratory qualitative study because the purpose was not statistical generalization but analytic depth and pattern identification, as described by Creswell and Yin. 

The issue was addressed by selecting members with diverse products/services areas; va roles; customer bases; philippines presence as well varying levels of expertise/experience. Sample was of va's from clothing & cosmetics; electronic product retailing; home goods & kitchen appliances; pet products; health care facilities (fitness center). Roles were diverse such as a customer service represent-ative; an analyst for products to sell; logistics coordinator of orders; marketer who lists items on his/her website; social media manager, advertising consultant, supplier relations officer, designer for uploading graphics etc.; operation head & admin assistant. Table 1 is a participant profile and selection matrix. Names are replaced by participant code for privacy purposes only. Store name, customer's name, username on a website etc was never disclosed.

Table 1: Participant Profile and Selection Matrix.

The participant profile showed variation across 12 product niches, multiple VA functions, different foreign client markets, and experience levels ranging from six months to more than three years. The profile increased confidence levels as it did not rely on a single position, product line, customer location & drop ship type.

Data Collection

Data were collected through semi-structured interviews lasting approximately 30 to 60 minutes each. This approach is consistent with creswell's view that qualitative interviews allow researchers to capture participant meaning, interpretation, and lived experience. The interviews took place on various platforms such as zoom, google meet & messenger calls etc.

Every interview included information about recruiting procedures; training programs adopted by employees; interpersonal relationships with clients; evaluation of productivity levels; review mechanisms utilized to provide feedbacks; benefits offered as incentives. Interview date, interview type, consents given/not given, transcript filename & participant information is noted on this form. The data collection period was from March 10 to April 4, 2026, as shown on the interview log for the 12 participants.

Table 2: Coded Excerpt Matrix for the 12 VA Interviews.

Data Analysis Approach

The study used Braun and clarke's six-phase thematic analysis process. Then we have read through them for familiarity of information obtained from these interviews. Also, first code was derived out of repetitions as well as significant trends found within respondents' accounts. Secondly, similar code sets have been clustered as initial categories. Secondly, we checked out these topics for accuracy with respect to what was observed from our sample population as well. Secondly every issue has been specified as well related with theory of mind. Lastly, some samples have been utilized for reporting results with a focus on preserving filipino va center's perspective throughout this study.

Hybrid programming approach is employed here. Deductive codes were developed from Jensen and meckling's Principal-Agent Theory and Deci and ryan's Self-Determination Theory. They were also called for instance surveillance, independence capability proximity knowledge inequality power. Also induction coding was permitted as a result of participants' accounts such as no work done during testing period; disappearing after being tested out; obsolete procedure manuals (sop); slack stressors; showing off portfolios for others' view while on camera recording sessions etc.; feeling pressured by cameras when taking photos/videos.

For enhancing analytical clarity, this research employed 3 interrelated matrix models participant profile matrix coded excerpt matrix theme matrix. Exact quotes are connected with codes, themes as well theories. The process addressed reviewer's suggestion of including an organized dataset prior analysis, as well ensuring traceability from participants' records back through results.

Trustworthiness and Ethical Considerations

Following Lincoln and Guba, credibility was maintained through systematic coding, use of participant excerpts, and comparison of patterns across different VA roles and industries. Reliability is ensured by means such as coding system, diary keeping & auditing records. Verification is ensured through an unbroken chain of relationship between transcript, code, theme & interpretation. The transferability issue is resolved through participants' backgrounds that are disclosed but no identifiers are given out of it. The participants are aware of being able to choose whether they want answers as well skipping some part altogether; also anonymity (pseudonym) /code will go for them. None of these details such as company names, customer identifiers & usernames on platforms are disclosed by us.

Table 3: Theme Matrix Used for Analysis.

Data Matrix for Analysis

The following matrix was placed immediately before the thematic analysis to show that the themes were grounded in the 12 interviews and not just a single deep-dive case narrative. It enabled verification of relationships among participants' profiles at an hr level as well as thematic analysis results derived directly out of that information.

The results of this part were obtained out of a code for an interview data set. Analysis is based on key issues derived out of theory as well as interviews with respondents, respectively. Each claim was linked to the coded excerpts in Table 2 and in the Excel coding matrix.

Theme 1: Skill-Signaling Recruitment and Trial-by-Fire Hiring

The first them was about replacing credentials for hiring by showing skills instead. Drop shippers' bosses did n't care much about their education levels but were worried if a va is able to take on things like shopify, zendesk, gorgia, metacampaigns manager & canva. This supported sullivan's view of skill-based hiring and Jensen and meckling's explanation that principals try to minimize uncertainty before hiring agents. VA-01 and VA-02 data are used for showing if a trial task is paid or unpaid, if it is like real work, and whether the participant perceived it as fair assessment or exploitation. From an agency theory point of view, trial assignments enabled agents to evaluate their competence prior to contracting with them. On contrary to that for workers; unrecorded/unclear trial is an unfair process as they were providing their labour first and then getting recognized by employers/ patrons/mentor. An equitable alternative is a suspended work assignment with an attached pay period that also permitted appraisal of their efforts as well.

Theme 2: Asynchronous Onboarding and the Competence Gap

The second theme was onboarding through SOPs, recorded videos, task boards, and self-paced documentation. Bauer, (2010) stated that onboarding helps new employees understand their roles, responsi-bilities, and company culture. Async onboarding addressed an actual issue of working across multiple countries where both owner's as well as va's may be located at various hours. This enabled an alternative vaper for reviewing software/programs with no need of their presence at all times. Nevertheless, it also resulted to an expertise deficit where documents were out-of-date and/or irrelevant for work purposes. VA-03 and VA-04 data are used for identifying types of onboarding materials utilized, whether it was an online course, and any mistakes/uncertainties due to lack of mentorship. This finding extended Deci and ryan, (1985) Self-Determination Theory. Skill wasn't developed just through providing access of resources. Skill needs comments, explanations as well an opportunity to ask questions freely. If an av is educated solely on process scripts then they may be aware of procedures yet lack understanding about their significance. It brought stress when it came to refund requests, chargeback complaints and out-of-stock situations along with upset customers' complaints.

Theme 3: Algorithmic Visibility and Performance Pressure

The third theme was about performance management through digital monitoring. Ball, (2010) defined workplace surveillance as a practice that could support control but also create tension between employer oversight and worker privacy. Certain va's felt that a tracker, screenshot was annoying since its made them feel surveilled constantly. Other people understood these instruments for security, since it was proof that one had been working at an other's company. It is significant as there were two experiences of algoritmization, not just a single one. It was dependent upon if they were coordinating fairly, punishingly. VA-05, VA-06, and VA-07 were used to differentiate chat response time, output KPIs, and screen capture monitoring. Principal-Agent Theory explained why owners relied on monitoring tools in remote teams (Jensen & Meckling, 1976). Yet Self-Determination Theory explained why excessive monitoring reduced autonomy and relatedness (Deci & Ryan, 1985). A veteran of many hours spent on-screen capturing screenshots as well waiting for responses from clients made him/her an adversary to their work ethic when it came time for evaluation. It is a good idea, however, include performance indicators that are based on results as well as regular reviews of their status along side reasonable work hour limits.

Theme 4: Transactional Loyalty, Malasakit, and Retention

The fourth theme was about the quality of the psychological contract. Blau, (1964) Social Social Exchange Theory explained why employees were more likely to reciprocate loyalty when they perceived fairness, support, and respect. Many va's were retained by clients for their work as they got money from it; had freedom of movement anywhere world over. Nevertheless, lack of real concern undermined loyalty. The Filipino va's perceived prompt payment, politeness level; clarity on what is needed from them; real-time monitoring for their needs when they were suffering seriously. On contrary to these factors such as lateness on payment; abrupt firing from company; disappearing after work etc.; blaming others publicly etc.; no response back are signs for a businesslike relation ship only. Evidence from VA-08, VA-09, VA-10, VA-11, and VA-12 expanded the paper beyond the deep-dive case and showed multiple participant voices. The topic explains how, although lean hr was effective initially, it turned out to be costly later on. Bakker and demerouti, (2007) Job demands-resources model indicated that high demands without adequate resources could lead to strain, while Deci and Ryan, (1985) explained why relatedness and autonomy were important for motivation. Workers who weren't valued nor integrated tended be more likely not work for a company; fail at learning its skills and abilities better than others did. Remote sustainable hr needed to be transparent as well relational.

Deep-Dive Case Narrative: VA-07

VA-07's deep-dive case also demonstrated how technological filtering of HR practices affected workers' experiences from hiring to work perfor-mance. The candidate's competence is assessed by his/her hands-on service skills as opposed to qualifications that were a sign of competency during an experience-based selection process for drop-shipping jobs. Onboarding was mostly based around passively received information with little active engagement leading to a lack of clarity regarding stores' processes at that time. Digital technologies like clock-in/out logs/screenshots of work done, tickets for issues etc. Deep dive story showed a va's ability to pass assessment tests with little guidance from mentors as well being monitored constantly through social media platforms during their job tenure.

This study examined the HR practices of global dropshipping micro-enterprises through the experi-ences of Filipino VAs. HR was an information-sealed environment for businesses of our industry. There is recruitment, on board training as well but its been moved out to be done through apps/trials/sop management/monitoring dashboard/monitoring application. The strategies facilitated organizational coordination across borders; however, these measures undermined loyalty, independence, ability, and connection as well due to lack of guidance from a superior employee with equitable remuneration for labor performed under compassionate interpersonal relations management techniques. Dropshipping is effective, however unstable. Skills-based hiring found qualified employees; however, unskilled practice assignments seemed to be predatory because these worked out of nowhere with no compensation involved. On-boarding was asynchronous and provided docs; however, mentoring wasn't replaced by asynchrony. Tracking performance helped to recognize working from home; however, over-monitoring caused anxiety as well decreased interest levels internally. Lastly, lack of malasakit altered va client relations from service oriented to commercial ones which increased risk factors for churns.

Paid probation/trial assignments are better for stores, not free sample jobs. It also enables them to evaluate candidate's skill level without wasting their valuable effort on a veteran's part. Hybrid onboarding is needed by stores as well. SOPs and Loom videos may still be useful, but these should be supported by a required first-week live orientation to explain values, expectations, escalation rules, and business context. KPIs for storing companies can be used instead of capturing screens when they are feasible, store managers must do so. 

Ticket resolution rates; errors per transaction; response time; listing correctness of products provided by campaigns etc. Also store managers must plan for visits as well since periodic, weekly, 2-weekly, phone contact will help clarify things better, increase transparency between them & customers. Payschedule, tracking commitments & agreements with clients' privacy policies as well disputes resolution process must too be specified prior to starting a project.

A virtual assistant must have an online presence on top of their job history profile. Sample products listings, customer reply templates, ad-copy samples, dashboard screenshots with sensitive data removed, or workflow maps can better convey competence than credentials in skill-signaling markets (Sullivan, 2013). Also, va's need to define what is going-to-be evaluated on while taking a job. For example “what is your definition of success for me?” ‘will is screenshot your work' ‘do you want a kpi'. Also, va's must have limits of access/monitoring/ payment conditions as per a contract since it reduces ambiguity/confusion between parties.

Future research could also investigate how VA's perceive stores' management styles such as surveillance tools usage, informal hiring practices, and lack of HR systems for managing remote teams. The quantitative research can be done to explore a correlation of monitoring intensity, burnout, job satisfaction, and turnover intention of remote Filipino VAs. Further comparative studies can be conducted on differences between e-commerce industries, customers' countries of origin, virtual assistants' positions and platforms employed by them.

The author wishes to thank the Filipino Virtual Assistants who participated in this study and shared their experiences of working in global dropshipping and e-commerce operations. Their participation provided the qualitative data used in examining recruitment, onboarding, performance management, and retention practices in remote work settings.

The author declares that there are no commercial, financial, or personal interests that could be construed as a potential conflict of interest in relation to this study.