Designing and evaluating a digital secretary workflow platform for SMEs

Thesis Abstract

This study investigates the design, implementation, and evaluation of a digital secretary workflow platform to enhance administrative efficiency and decision support within small and medium-sized enterprises (SMEs) facing information overload, fragmented communication, and manual process bottlenecks. The problem addressed is the misalignment between evolving secretary duties and existing legacy tools, which constrains task coordination, document management, and scheduling accuracy in fast-paced SME environments. The aim is to design a modular, interoperable platform that automates routine secretary tasks, standardizes workflows, and provides analytics to inform management decisions. Specific objectives include (1) mapping core secretary workflows in SMEs and identifying bottlenecks; (2) developing a reference architecture and user-centered prototypes grounded in human-computer interaction (HCI) and digital workflow theories; (3) implementing a functional platform enhancing scheduling, correspondence management, and task tracking; (4) evaluating usability, adoption, and workflow performance through empirical testing; and (5) deriving evidence-based recommendations for SME deployment and change management. The study adopts a mixed-methods design combining design science research (DSR) with a quasi-experimental evaluation. The population comprises administrative staff and secretaries across 12 SMEs in manufacturing and services sectors. A purposive sample of 36 participants (3–4 secretaries per SME) is recruited for qualitative needs elicitation and usability testing, complemented by a broader sample of 120 staff members for impact assessment. Data collection instruments include semi-structured interviews, contextual inquiry notes, task analysis checklists, System Usability Scale (SUS), NASA-TLX workload assessments, and platform usage analytics. The platform is developed using a modular architecture with microservices for scheduling, document routing, email/communication management, task assignment, and analytics dashboards, aligned with the Technology Acceptance Model (TAM) and Activity Theory to inform interface and functionality choices. Validity and reliability are ensured via triangulation, pilot testing of instruments (n=12), inter-rater coding for qualitative data (Cohen’s kappa >0.70), and pilot runs of the platform (n=20) to refine measurement instruments. Data analysis employs descriptive statistics, ANOVA to compare pre- versus post-implementation workload and time-to-complete tasks, regression analysis to identify predictors of perceived usefulness and ease of use, and thematic analysis for interview data. Descriptive analytics from platform logs (e.g., task completion time, backlog size, user engagement) are complemented by pre-post comparisons over a 12-week evaluation period. Expected findings indicate significant reductions in average task completion time (mean decrease of 28%), lower perceived workload (NASA-TLX reduction of 15 points), and improved scheduling accuracy (error rate drop by 22%). SUS scores are anticipated to scale from 62 to above 78, indicating move toward high usability, with higher levels of adoption among staff who engage in co-design activities. Thematically, it is expected that the platform reduces context-switching costs, enhances document governance, and improves cross-functional communication, while highlighting challenges related to data governance, interoperability with legacy systems, and organizational change resistance. The study contributes to knowledge by operationalizing a design, implement, and evaluate blueprint for digital secretary platforms in SMEs, integrating DSR, TAM, and Activity Theory into a practical methodological framework, and offering a validated reference architecture, set of design patterns, and an evaluation toolkit. It also expands empirical understanding of how automated workflow platforms influence administrative efficiency, information accuracy, and user satisfaction in smaller organizations, where resource constraints critically shape technology adoption. The main conclusion anticipates that a purpose-built digital secretary platform substantially enhances administrative performance and decision support in SMEs when co-designed with end-users, integrated with existing systems, and accompanied by structured change management. Recommendations include (i) adopting a phased implementation with early wins in scheduling and document routing, (ii) investing in data governance and API-enabled interoperability, (iii) establishing governance for access control and audit trails, and (iv) developing ongoing training and change-management programs to sustain Benefits realization in SME contexts. Further research could explore long-term impact on organizational agility and the diffusion of digital secretary platforms across industry clusters.