Mapping the landscape of productivity in construction research: an umbrella review of trends, gaps and future directions

The global construction industry has been grappling with persistently low productivity levels for decades. This challenge is particularly pronounced when viewed against the backdrop of other economic sectors that have achieved substantial productivity gains over the same period (Barbosa et al., 2017). In Australia, for example, construction productivity has been in steady decline since 2013 (Australian Constructors Association, 2023), reflecting a broader international pattern exemplified by similar declines in the United States since 1987 (Rathnayake and Middleton, 2023). This sustained reduction represents a significant barrier to the industry's capacity to address contemporary challenges.

The imperative to resolve this productivity challenge has intensified due to converging external pressures. Climate change and housing affordability demand both rapid decarbonisation of the built environment and substantial expansion in housing supply (Salisu et al., 2026). Simultaneously, geopolitical instability has created volatility in material costs and supply chain reliability. These factors compound existing inefficiencies and transform productivity improvement from a desirable goal into an existential necessity for the industry's sustainable future (Mischke et al., 2024).

Central to addressing this challenge is a clearer understanding of what productivity in construction means and how it should be measured. Most research characterises productivity as the ratio of output to input, capturing how efficiently resources are converted into built outcomes (Rathnayake and Middleton, 2023). However, this formulation alone is insufficient. A more comprehensive conceptualisation recognises productivity as the product of two distinct dimensions: efficiency, defined as doing things right by maximising output from available resources, and effectiveness, defined as doing the right things by delivering value to clients and end users (Crawford and Pollack, 2021). This distinction matters because prevailing measurement approaches in construction are dominated by efficiency metrics. The most common is labour productivity, expressed as units of work placed per man-hour, which captures only one side of the productivity equation and confines analysis to on-site activities (Dolage and Chan, 2013; Rathnayake and Middleton, 2023). Upstream stages such as design, off-site manufacturing, and project planning, where decisions with profound consequences for construction-phase productivity are made, remain largely outside the scope of productivity measurement frameworks. Understanding why productivity has stagnated therefore requires not only better data, but a more expansive and coherent conceptual framework.

Recognising the need for improvement, governments worldwide are mobilising resources to tackle this challenge. Across Europe, North America, and the Asia-Pacific, national productivity-related strategies increasingly identify construction as a priority sector for intervention. Notable examples include Singapore's Productivity Solutions Grant (PSG), which supports small and medium enterprises in building digital capabilities, and Australia's Productivity Commission review of Housing Construction Productivity. Numerous systematic (e.g. Adebowale and Agumba, 2023b) and non-systematic literature reviews (e.g. Carson and Abbott, 2012) now examine construction productivity from diverse perspectives, including technological adoption and measurement techniques (Dixit et al., 2019).

This diversity of perspectives has yielded an enriched understanding of specific productivity dimensions. Reviews have examined productivity metrics through digitalisation (Dutra et al., 2025), explored monitoring techniques (Alaloul et al., 2022), and identified the most significant factors affecting construction performance (Lindhard et al., 2025). However, this proliferation of domain-specific reviews has also produced a fragmented conceptual landscape. Terms such as “productivity”, “construction productivity”, and “labour productivity” are used inconsistently and often interchangeably, undermining cross-study comparability and hindering cumulative knowledge building. Furthermore, these reviews have concentrated disproportionately on construction-phase activities, leaving upstream stages such as design, manufacturing, and strategic definition substantially underexplored (Dolage and Chan, 2013). Without systematic comparison across reviews, it remains unclear where findings converge to establish robust evidence or diverge to reveal contested areas requiring further investigation. The absence of meta-level synthesis means that overarching patterns, research trends, and strategic knowledge gaps remain invisible, hindering the development of theory and evidence-based policy.

Addressing this limitation requires an umbrella review, a methodology specifically designed to synthesise findings across multiple literature reviews (Grant and Booth, 2009). Unlike traditional systematic reviews, which examine primary studies, umbrella reviews aggregate and critically evaluate existing literature reviews. This provides a meta-perspective capable of revealing patterns, contradictions, and gaps invisible at lower levels of analysis (Aromataris et al., 2020; Biondi-Zoccai, 2016). With 46 eligible reviews published between 2013 and 2025 alone, the field has reached a level of accumulation where meta-level synthesis is not only useful but necessary. Without it, duplication of effort, definitional inconsistencies, and fragmentation of evidence will persist. No prior umbrella review has mapped the productivity literature in construction, representing a significant gap in the current research landscape.

This study presents the first umbrella review of productivity research in construction, synthesising evidence from 46 reviews published between 2013 and 2025. The review identifies convergent themes, methodological patterns, and research gaps that remain invisible at the level of individual literature synthesis and primary studies. It is particularly timely given growing policy momentum around industrialised construction and platform-based delivery models, where productivity measurement frameworks spanning the full project lifecycle are urgently needed to inform investment decisions and national housing strategies. The study was guided by three interconnected research questions:

1. What is the current landscape of productivity research synthesis in construction?

2. To what extent do research syntheses converge or diverge in their methodological approaches and substantive findings?

3. What are the main knowledge gaps, methodological limitations, and strategic opportunities for advancing construction productivity research?

To address the research questions, an umbrella review was adopted. The volume of domain-specific systematic reviews on productivity in construction has increased substantially over the last decade. This has produced fragmented evidence base that no single systematic review is structurally capable of addressing (Grant and Booth, 2009). A conventional systematic review synthesises primary studies and is well suited to answering focused empirical questions. However, it cannot reveal cross-review patterns and inter-domain contradictions. A scoping review maps the breadth of a literature but does not provide the critical appraisal rigour necessary to assess the methodological quality of accumulated evidence. An umbrella review, by contrast, synthesises existing systematic reviews rather than primary studies, operating at a higher level of evidence aggregation (Biondi-Zoccai, 2016).

This approach offers several distinct advantages. It integrates findings from multiple sub-domains into a single meta-perspective. It identifies convergences and contradictions across existing reviews, providing more reliable evidence base for research and policymaking. It also prevents duplication of effort and clarifies the state of the science at a strategic level (Fusar-Poli and Radua, 2018; Biondi-Zoccai, 2016). These advantages are particularly relevant for exploring productivity in construction research, where insights are distributed across multiple specialised domains, including measurement frameworks, digitalisation, influencing factors, and industrialised construction, and no prior synthesis has mapped the field.

This study followed the Joanna Briggs Institute (JBI) Manual for Evidence Synthesis (Aromataris et al., 2020). The JBI framework was selected as the foundational methodology because it provides the most rigorous and internationally recognised standards specifically designed for umbrella reviews (Biondi-Zoccai, 2016). In accordance with JBI protocols, the methodology followed a structured four-stage process: (1) search strategy development, (2) screening and selection of reviews, (3) data extraction and collection, and (4) data summary and analysis. The specific methods and procedures for each stage are detailed in the following subsections.

Three primary databases were chosen to support the umbrella review: Scopus, Web of Science (WoS), and Google Scholar. The search terms “productivity”, “construction”, and “review” were chosen to align with this review's research questions. Table 1 provides the specific search strings and strategies employed in each database.

The search strategy was developed to identify relevant literature reviews from multiple sources. The search encompassed both peer-reviewed journal articles and grey literature sources, including conference proceedings and book chapters. To maximise coverage, no temporal restrictions were imposed; however, the search was limited to English-language publications.

The initial search was conducted on February 27, 2024, with a subsequent update on September 4, 2025, to capture recently published reviews. The combined searches yielded 531 results (see Figure 1). After removing duplicates (n = 179), 352 records were imported into Covidence (https://www.covidence.org/) for the screening process.

A two-stage screening process was then employed to select relevant reviews. Both rounds adhered to the following inclusion criteria: 1) the record was a complete and published review, 2) it focused on productivity within the construction industry, 3) its primary aim was to conduct a literature review (excluding research papers with a minor literature review component primarily based on empirical/raw data), and 4) it employed a systematic search and selection process, thereby excluding narrative reviews.

Initial screening evaluated titles, abstracts, and keywords against the inclusion criteria. This stage eliminated 235 records, advancing 117 to full-text review. Seven articles were subsequently excluded due to inaccessibility, leaving 110 records for the second round of screening.

Full-text screening incorporated a quality assessment tool using a standardised checklist (see Supplementary Material A) adapted from Aromataris et al. (2020) and AMSTAR-2 (Shea et al., 2017). While AMSTAR-2 is the recognised standard for appraising systematic reviews and reviews of reviews, direct application of the full instrument was deemed inappropriate for this corpus. Developed primarily for clinical and health sciences research, AMSTAR-2 reflects evidence synthesis conventions that differ substantially from those of construction research. Applying its criteria without adaptation would have excluded most otherwise rigorous reviews, not due to poor scholarship, but due to disciplinary differences in research practice. For instance, duplicate study selection and data extraction, standard expectations in clinical reviews, are rarely reported in construction systematic reviews even in high-impact journals (e.g. Alaloul et al., 2022); and the requirement for detailed description of all included studies is neither feasible nor analytically meaningful for bibliometric and mapping reviews analysing evidence exceeding 300 records (e.g. Qi et al., 2024). These characteristics reflect disciplinary norms rather than methodological deficiencies, a distinction acknowledged by Aromataris et al. (2020) in guidance on context-sensitive evidence synthesis. Accordingly, the checklist distinguished between compulsory and optional criteria (See Supplementary Material A). Reviews were considered methodologically sound when they demonstrated clearly articulated research questions or objectives, explicit inclusion criteria, and a transparent and systematic search strategy. Of the 110 records assessed at full-text, 46 satisfied both the inclusion criteria and compulsory quality standards and proceeded to the data extraction phase. Full quality appraisal results per study are reported in Supplementary Material A.

A mixed method approach has been implemented in this study. A standardised data extraction form, based on Aromataris et al. (2020), was developed to systematically capture relevant information from the included reviews. The form collected data across the following key domains:

1. Review characteristics: review and publication type,

2. Search parameters: database coverage and typology of primary studies,

3. Productivity research focus: Productivity conceptualisation, phenomenon of interest (i.e. review's aims and objectives), and project lifecycle phase analysis.

Complementing the structured extraction, a thematic analysis was conducted using NVivo software, following the approach outlined by Miles et al. (2020). Full-text reviews were imported and systematically coded to identify recurring patterns and themes across the evidence base. This inductive coding process facilitated the synthesis of overarching conceptual themes emerging from the collective review findings. To ensure the reliability and trustworthiness of the coding, a coding comparison query was executed in NVivo to measure inter-coder agreement between the two researchers (LGB and SM) using Cohen's kappa. A kappa value exceeding 0.75 was adopted as the threshold for excellent agreement (Landis and Koch, 1977), and an overall kappa of 0.79 was achieved, indicating excellent agreement between researchers. Any divergences were resolved through structured discussion between the researchers until consensus was reached.

This study further utilised a semi-quantitative content analysis to synthesise existing literature on productivity factors. The methodology involved an iterative process of factor extraction, thematic categorisation, and frequency mapping to identify and rank the most prevalent core themes identified in previous research. 239 factors were considered in this analysis. By quantifying the prevalence of specific factors across diverse studies, the resulting core theme list is grounded in empirical consensus in its impact on construction productivity (Ryan and Bernard, 2003).

Building on the data extraction and thematic analysis, the final stage involved synthesising, analysing, and summarising the findings from the 46 included reviews. This stage comprised two phases: (1) presenting the extracted information in a tabular format and (2) conducting a cross-case analysis to compare the reviews and identify patterns in their findings (Miles et al., 2020). The first phase directly addressed this review's first research question, while the cross-case analysis aimed to answer the second research question. This analysis helped identify key knowledge gaps, methodological limitations, and strategic opportunities for advancing construction productivity research (third question). The first two research questions are addressed in the following section, while the discussion section addresses the third research question.

Despite no date restrictions in the search strategy, all 46 identified reviews emerged from 2013 onward. A marked acceleration occurred after 2020 (Figure 2). The temporal distribution reveals an evolving field: minimal activity during 2013–2018 (0–1 reviews annually), followed by steady growth from 2020–2022 (3–8 reviews), peaking at 10 reviews in 2023. The lower count for 2025 reflects the search date rather than a decline in activity, as publications typically continue appearing throughout the year. This trajectory mirrors increasing interest in conducting systematic reviews within construction research. The post-2020 surge particularly suggests growing recognition of the need for consolidated evidence.

The methodological profile, categorised using Grant and Booth's (2009) typology, reveals a strong preference for qualitative synthesis approaches. Qualitative systematic reviews dominate (n = 26, 56.5%), followed by mapping reviews/systematic maps (n = 8, 17.4%), scoping reviews (n = 7, 15.2%), and meta-analyses (n = 5, 10.9%). The predominance of qualitative approaches and scarcity of meta-analyses likely reflects the heterogeneous nature of construction productivity research. Diverse methodologies and outcome measures challenge quantitative aggregation (Rathnayake and Middleton, 2023). This distribution indicates opportunities for more quantitative synthesis to advance the evidence base and complement qualitative aggregative findings. Supplementary Material B presents a complete list of the included reviews.

To assess the degree of overlap among the included reviews, a Corrected Covered Area (CCA) analysis was conducted. This is a methodological requirement for umbrella reviews to detect potential double-counting of evidence (Pieper et al., 2014). The CCA could only be calculated for 20 of the 46 reviews. The remainder either did not specify their full list of included primary studies or employed bibliometric methods where individual record attribution is not compatible with overlap calculation (see Supplementary Material A). The analysis yielded a CCA of 1.4%, indicating slight overlap across the calculable subset (Pieper et al., 2014). Where overlap was identified, it occurred predominantly among reviews examining factors affecting construction labour productivity. These reviews drew on a shared pool of frequently cited primary studies in this domain. Yet, this overlap did not constitute redundancy. The reviews in question employed the shared evidence base for substantively different research aims. For instance, Hasan et al. (2018) examined productivity factors through a geographical lens across a 30-year evidence base, while Jian et al. (2025) analysed the same factors through a meta-analysis to derive pooled effect sizes. This pattern reflects the depth of the literature on factors influencing construction productivity rather than duplication of effort. Given that the CCA was calculable for less than of the records, this result should be interpreted as indicative rather than definitive.

Although no restrictions were imposed on publication types during the search process, the retrieved records (Supplementary Material C) comprised predominantly academic journal articles (n = 32, 69.6%) and conference proceedings (n = 14, 30.4%), with no book chapters meeting the inclusion criteria.

Analysis of the reviews showed a clear hierarchy of database utilisation. Scopus emerged as the predominant database, used by 32 reviews, followed by Web of Science (n = 20), Google Scholar (n = 9) and Science Direct (n = 6). More specialised databases—including Emerald, Taylor and Francis, IEEE, ASCE, and PubMed—were each utilised by three or fewer reviews, typically serving as supplementary sources to Scopus and Web of Science (c.f., Lindhard et al., 2025).

Five reviews employed exclusive journal-based searches rather than database queries (Dixit et al., 2019; Dolage and Chan, 2013; Hasan et al., 2018; Kirby et al., 2022; Rathnayake and Middleton, 2023). These reviews consistently searched across Construction Management and Economics, International Journal of Project Management, Journal of Management in Engineering, Journal of Construction Engineering and Management, Engineering, Construction and Architectural Management, and Automation in Construction. Methodological transparency varied considerably across these studies. Some authors provided no justification for journal selection (e.g. Dixit et al., 2019). Others provided a clear rationale. For example, Rathnayake and Middleton (2023) explicitly justified their selection of the ten highest-ranked journals based on h-index metrics within construction management.

Journal articles were the dominant source type regardless of the review's publication venue. Conference proceedings were included in 20 reviews (43.5%), typically as supplementary rather than primary sources. Source heterogeneity remained limited, though three reviews expanded their scope: Hasan et al. (2018) and Wong et al. (2021) included books and theses, while Rouhanizadeh and Kermanshachi (2021) failed to specify their included sources, a notable methodological omission.

The consistent reliance on Scopus and Web of Science across both journal and conference reviews (n = 17) underscores the field's dependence on established bibliographic databases for peer-reviewed literature. However, the inclusion of conference proceedings in nearly half the reviews signals recognition of grey literature's value. The broader exclusion of reports from policymakers, industry associations, and consultancies risks marginalising substantial bodies of practical knowledge. In a field where industry-level productivity insights emerge as readily from government inquiries and benchmarking exercises as from academic research, this gap deserves attention.

Three key dimensions were examined to systematically analyse the productivity research focus of the included reviews: the productivity terminology employed, the definition of productivity provided, and the project lifecycle phases under investigation.

The analysis reveals a pronounced emphasis on traditional efficiency metrics (Crawford and Pollack, 2021; Rogers, 1998). A total of 58.7% (n = 27) of reviews adopted this perspective (Table 2). Within this group, approximately one-fifth explicitly operationalised productivity as output-to-input ratios. A similar proportion utilised units of work per man-hour, predominantly in studies examining labour productivity. Three reviews inverted this conventional formula (i.e. input/output), indicating fundamental disagreements about appropriate measurement approaches (Hamza Momade et al., 2023; Yi and Chan, 2014; Ghasemi Poor Sabet and Chong, 2020b).

Only two reviews (4.3%) incorporated both efficiency and effectiveness dimensions (Dolage and Chan, 2013; Nurhendi et al., 2024). This minimal adoption suggests the field has yet to embrace a value-optimisation perspective towards stakeholders' beneficial outcomes (Crawford and Pollack, 2021).

The relationship between productivity terminology and lifecycle phase analysis reveals distinct patterns (Table 3). Reviews employing the generic term “productivity” exhibited substantial heterogeneity. Some explored the entire project lifecycle from planning through operation (e.g. Chowdhury et al., 2019; Dutra et al., 2025), others focused on design-manufacturing-construction analysis to investigate off-site construction productivity (Ghasemi Poor Sabet and Chong, 2020a). Others still examined on-site construction exclusively (Chen et al., 2022; da Barbosa and Costa, 2021). This variability suggests that broader productivity terminology is associated with more diverse analytical scopes. To provide definitional clarity, Table 4 defines each lifecycle phase category and presents representative examples. As shown, phase coverage across all evidence predominantly oriented toward on-site construction activities. Only a small number of reviews extending their analytical scope to upstream stages such as design and engineering, manufacturing, or adopting a whole-lifecycle perspective. This pattern reflects, and arguably reinforces, the field's persistent concentration on construction-phase productivity at the expense of broader project value chain considerations.

By contrast, studies utilising “construction productivity” and “labour productivity” demonstrated greater phase-specific consistency, predominantly focusing on on-site construction activities (e.g. Dixit et al., 2019; Hasan et al., 2018). While this convergence suggests implicit disciplinary consensus about these terms relating to on-site execution, it also reveals a significant blind spot: the failure to acknowledge cross-phase interdependencies. Notable exceptions attempting to bridge this gap included Archchana and Pan (2023) and Ghasemi Poor Sabet and Chong (2020a), which extended analysis to pre-construction phases, recognising that construction productivity is significantly influenced by upstream decisions.

The most concerning finding is the absence of clarity around the definition of productivity and phase-boundary specification across the included reviews. A substantial 37.9% (n = 17) of reviews use productivity terminology without providing explicit definitions. One-third failed to explicitly delineate their phase-specific boundaries. This conceptual absence reveals a troubling assumption that “productivity” or “construction productivity” has a universal meaning. This is particularly problematic given productivity's multi-level nature in construction, spanning macro-industry to micro-task analyses across diverse project phases (Dutra et al., 2025). Without clear baseline definitions, studies cannot be meaningfully compared, synthesised, or critiqued, effectively fragmenting the knowledge base.

Reviews were also classified based on their phenomenon of interest as reflected in each review's aims and objectives (Table 3). Five key thematic groups were identified:

1. Mapping existing literature on productivity, providing an overview of the state-of-the-art (n = 8).

2. Current practices around productivity measurement methods and approaches (n = 2).

3. Methods for productivity monitoring and data capture (n = 8).

4. Factors influencing productivity (n = 16).

5. Drivers and enablers for productivity improvement (n = 12).

The following subsections summarise the findings of each thematic group:

Eight reviews, collectively spanning four decades of scholarly inquiry from 1978 to 2023, have primarily mapped the existing productivity literature landscape within the construction industry. Although these reviews vary in scope, they converge on a central finding: the field has been predominantly concentrated on labour and workforce productivity factors. This labour-centric paradigm has created cascading constraints across multiple dimensions. Theoretically, it has narrowed the conceptual boundaries of productivity research (Dolage and Chan, 2013). Practically, it has limited the development of targeted improvement strategies (Adebowale and Agumba, 2023d). In terms of measurement, traditional labour productivity metrics continue to dominate despite calls for more integrated assessments encompassing organisational and project-level performance (Lee et al., 2023). Equipment resources remain underexplored in productivity discourse and literature synthesis, with Forsythe's (2018) work on crane utilisation a notable exception.

The reviews unanimously identify factors affecting productivity as the field's most extensively investigated domain (Dolage and Chan, 2013; Yi and Chan, 2014). Labour and management variables predominate, with skill levels, worker experience, motivation, planning quality, and supervision effectiveness emerging as recurrent themes (Dixit et al., 2019; Lee et al., 2023).

A second area of consensus centres on the accelerating adoption of digital technologies for productivity monitoring. All reviews, including those examining broader productivity (Zarghami, 2024), document growing interest in technologies such as BIM, computer vision, and wearable sensors. However, these innovations remain primarily deployed for labour monitoring and tracking, reinforcing rather than transcending the field's workforce-centric orientation.

The reviews collectively signal a methodological transformation towards data-driven methodological approaches (Dixit et al., 2019). Qi et al. (2024) document the increasing deployment of artificial neural networks, machine learning algorithms, and fuzzy logic systems for productivity prediction and optimisation. Zarghami (2024) similarly suggests the field is experiencing a shift from subjective, survey-based investigations toward quantitative, model-based analytical frameworks. This methodological evolution suggests the field is gradually moving beyond descriptive factor identification toward predictive and prescriptive capabilities.

Only two reviews primarily examined current practices in productivity measurement methodologies. Dutra et al. (2025) conducted a scoping review that analysed productivity metrics using Building Information Modelling (BIM) across the entire project lifecycle. Rathnayake and Middleton (2023) investigated productivity measurement approaches across different levels of analysis and specification, spanning from macro-level economic indicators to micro-level measures including individual projects, activities, and crews/individuals.

Both reviews converge on the traditional economic conceptualisation of productivity as the ratio of output to input and acknowledge that labour productivity remains the predominant indicator during the construction phase. This prevalence likely reflects data collection challenges associated with more comprehensive measures, such as total factor productivity, which necessitates data on multiple input variables (Rathnayake and Middleton, 2023).

A core finding shared by both reviews concerns the absence of standardised productivity measurement frameworks. Dutra et al. (2025) emphasise the need for standardised metrics throughout project lifecycles, incorporating inter-phase dependencies and applicability beyond the construction phase. Rathnayake and Middleton (2023) argue that measurement inconsistencies impede cross-study comparability at the micro-level, a limitation particularly attributable to poorly defined boundaries and unacknowledged methodological constraints within existing studies.

Eight reviews examined techniques for monitoring productivity in construction, with particular emphasis on data capture and processing methodologies (e.g. Alaloul et al., 2022; Lim et al., 2024). Although the majority of reviews focused on automated techniques (e.g. Chen et al., 2022; da Barbosa and Costa, 2021), some also analysed manual methods (e.g. Alaloul et al., 2022; Moohialdin et al., 2020). A consistent theme emerging across these reviews is the imperative to move beyond traditional monitoring practices. Work sampling and crew time utilisation are time-intensive, labour demanding, and susceptible to observational errors and subjective interpretation (da Barbosa and Costa, 2021; Salaheen et al., 2025). Reviews strongly advocate for the adoption of automated solutions, particularly vision-based and sensor-based technologies (e.g. Chen et al., 2022; da Barbosa and Costa, 2021).

Vision-based techniques, predominantly using computer vision and photogrammetry, have demonstrated considerable utility in automated object detection, tracking, and activity recognition (Alaloul et al., 2022). These technologies enhance the efficiency and accuracy of monitoring construction activities, especially in earthmoving operations (Chen et al., 2022). Sensor-based methods, on the other hand, including real-time location systems (RTLS) and kinematic sensors like Inertial Measurement Units (IMUs) and accelerometers, have proven valuable for monitoring resource location, movement patterns, and worker posture (Chen et al., 2022; Moohialdin et al., 2020). Hybrid systems integrating multiple sensor types have emerged as a solution to enhance measurement accuracy while mitigating environmental constraints (Alaloul et al., 2022).

There is strong consensus regarding the potential of BIM and Machine Learning (ML) in advancing productivity monitoring. BIM is widely recognised as a digital representation of construction processes that, when integrated with monitoring technologies, substantially enhances automation and monitoring performance (Archchana and Pan, 2023). ML algorithms have proven valuable for optimising system performance and processing datasets generated by computer vision and sensor technologies (Lim et al., 2024).

Despite the documented advantages of automated methods, significant barriers to widespread implementation remain. Semi-automated processes and high costs continue to hinder broader adoption (Alaloul et al., 2022; Archchana and Pan, 2023). Current research predominantly consists of proof-of-concept studies involving limited samples of workers and equipment, or focusing on discrete construction tasks rather than addressing the complexity of real construction sites (Salaheen et al., 2025; Lim et al., 2024). Wearable technologies present a potentially cost-effective monitoring solution, but their implementation needs careful consideration of social and ethical implications (Chen et al., 2022).

Twelve reviews examined drivers and enablers for enhancing productivity in the construction industry. Drivers were conceptualised as underlying forces propelling productivity growth. Drivers are broad, long-term trends such as industrialisation (e.g. Shang and Chan, 2019), digitalisation (e.g. Chowdhury et al., 2019), and decarbonisation (Adebowale and Agumba, 2026). Enablers, by contrast, represent specific tools, techniques, and approaches that facilitate the realisation of these drivers. They translate strategic objectives into measurable improvements. Notable enablers studies in the reviews include off-site construction for industrialisation (Ghasemi Poor Sabet and Chong, 2020a), augmented reality (AR) for digitalisation (Xu et al., 2024), and sustainable materials for decarbonisation (Adebowale and Agumba, 2026).

Research foci varied considerably across reviews in this group. Some investigated individual enablers such as artificial intelligence (AI) (Adebowale and Agumba, 2023a) or AR (Xu et al., 2022), while others adopted a more holistic approach, analysing how multiple enablers could advance productivity across the construction industry (Ghasemi Poor Sabet and Chong, 2020b). Despite this diversity, a common analytical framework was evident: most reviews systematically examined the capabilities, applications, and benefits of specific enablers while simultaneously identifying and evaluating implementation challenges (e.g. Chowdhury et al., 2019; Xu et al., 2024).

Digitalisation was the most investigated driver, with AR identified as the primary enabler. Industrialisation ranked second, driven by off-site and lean construction. Ghasemi Poor Sabet and Chong (2020a) examined the interaction between off-site construction and BIM, while Shang and Chan (2019) explored lean construction tools' impact on labour productivity. Decarbonisation received comparatively limited attention, examined in only one review by Adebowale and Agumba (2026). This review explored the productivity implications of sustainable materials through economic, environmental, and social dimensions. While acknowledging the environmental benefits of sustainable materials, it also identified substantial challenges hindering their widespread adoption.

A semi-quantitative frequency analysis was employed to synthesise the existing literature on construction productivity factors, categorising recurring variables into distinct core themes. Sixteen reviews synthesised literature examining specific factors influencing productivity. Fifteen tallied the frequency of identified factors. The number of factors listed across these reviews varied significantly, ranging from five to 35, with 239 factors considered in total. Most reviews did not provide data beyond factor rankings, limiting the possibility of broader analysis involving standard deviations and effect sizes.

One review presented factors grouped by region without an overall classification (Hamza Momade et al., 2023). To ensure methodological consistency, these regional groupings were subsequently unified. Factor counts across all 16 reviews were uniformly weighted to establish an overarching ranking of factors reported in the preceding literature.

Similarities in factor themes were identified across all reviews, regardless of whether reviews analysed labour productivity specifically or construction productivity more broadly (e.g. Hasan et al., 2018; Ardila et al., 2024). The top five core themes identified were: (1) worker motivation and welfare, (2) inadequate or competent supervision and management, (3) labour skills, training and experience, (4) design problems and information management, and (5) financial and payment issues (Table 5). The full list of 26 core themes and 239 extracted factors is presented in Supplementary Material D.

The review of thematic factors revealed a dual analytical focus. Some reviews concentrated on negative factors reducing productivity, while others examined positive factors enhancing it. Despite this difference in orientation, the identified themes frequently converged on the same underlying elements. For example, “supervision and management quality” consistently emerged across both research foci, appearing in negative-focused literature as inadequate supervision (Hasan et al., 2018) or poor management (Hamza et al., 2019), and in neutral/positive-focused reviews as good supervision (Van Tam, 2021), supervisor experience (Ardila et al., 2024), and management style (Adebowale and Agumba, 2023c).

Despite consensus on core factor themes, analysis revealed variations in how different stakeholder groups prioritised productivity factors (Adebowale and Agumba, 2023c; Han et al., 2024). Managers tended to focus on strategic and planning-related factors, including resource planning, management methodologies, and technical documentation (Hasan et al., 2018). Craft workers emphasised immediate operational factors affecting their daily work environment, including site conditions, material accessibility, safety provisions, and equipment utilisation (Adebowale and Agumba, 2023b).

Factor prioritisation also varied considerably by geographic region, raising questions about the reliability and consistency of region-specific findings (Hamza et al., 2019). This divergence was particularly evident when reviews employed different ranking methodologies; specifically, citation frequency versus quantitative effect size. Relying on citation frequency alone may misrepresent factors' actual impact, as illustrated by the underrepresentation of communication deficiencies and rework in some analyses (Jian et al., 2025).

This study presents the first umbrella review to systematically examine the nature and scope of literature reviews on productivity in the construction industry. Guided by three research questions, the findings provided an overview of temporal distribution, publication types, data sources, and research foci. Reviews were further classified based on their phenomenon of interest into five thematic groups. This discussion addresses the third research question by critically examining the main knowledge gaps, methodological limitations, and strategic opportunities for advancing research in this topic.

A core limitation found was the absence of clear terminology and consistent conceptualisation of productivity. The analysis revealed a persistent, narrow focus dominated by traditional efficiency metrics, with 58.7% of reviews operationalising productivity as the output-to-input ratio (Gutierrez-Bucheli et al., 2025). Many studies equate on-site construction productivity with labour productivity, measured as units of work placed or produced per man-hour (e.g. da Barbosa and Costa, 2021). This conflation overlooks the contributions of other resources such as technology, materials, and equipment. Forsythe (2018) offers a well-founded alternative, analysing process productivity using a resources-used approach, yet this remains an exception. As shown in Table 2, the same terms are used interchangeably across the 46 included reviews, efficiency and effectiveness are rarely distinguished, and input-output relationships are operationalised differently across studies and contexts. This finding is consistent with Rathnayake and Middleton (2023), who similarly identified the conflation of efficiency with productivity as a persistent weakness in the construction literature. The present study extends that finding by demonstrating, through systematic cross-review comparison, that this conflation is not confined to individual studies but is structural and field-wide.

This efficiency-dominated orientation represents a significant blind spot. Efficiency metrics illuminate how resources are utilised to deliver maximum output. However, they cannot capture the impact of innovation or sustainability on project success. Metrics grounded in effectiveness perspectives could address this gap (Crawford and Pollack, 2021). The Construction Productivity Taskforce (2022) similarly advocates for value-based measurement frameworks that move beyond input-output ratios. Adopting a value-optimisation lens incorporating both efficiency and effectiveness could enable researchers and practitioners to capture dimensions of productivity currently invisible to traditional approaches. However, it is also worth mentioning that selecting appropriate effectiveness metrics presents its own challenges, given the diverse stakeholder perspectives on what constitutes project value (Construction Productivity Taskforce, 2022).

Beyond the efficiency-effectiveness division, terminology itself lacks rigour. Terms such as “construction productivity” and “labour productivity” exhibit phase-specific consistency by concentrating on the construction phase. However, the broader use of the generic term “productivity” lacks rigorous stage definition. This terminological imprecision undermines cross-study comparability and fragments the knowledge base. Researchers implicitly operate with divergent assumptions about what productivity covers and where in the project lifecycle it applies. The conceptual heterogeneity evidenced in Table 2, including divergent definitions, inconsistent terminology, and the conflation of efficiency with effectiveness, points to the absence of a validated and multi-level productivity taxonomy as a critical research gap in the field. Such a taxonomy would systematically distinguish efficiency from effectiveness metrics and define operationalisation at macro, meso, and micro levels of analysis. It would provide the conceptual framework necessary to enable cross-study knowledge building. Its development is proposed as a priority direction for future research, grounded in primary empirical work and validated across diverse construction projects and sectors.

The reviewed literature has been primarily centred on construction-phase productivity. Only limited examples analyse productivity in other stages of the project value chain (e.g. Wong et al., 2021). This limited perspective neglects the substantial influence of upstream stages—such as strategic definition, design, and engineering—on overall project productivity. Dolage and Chan (2013) identified this same lifecycle blind spot over a decade ago, yet the present review demonstrates that it remains unresolved. While one review examined factors influencing design or engineering productivity, none addressed specific metrics for measuring productivity during these stages (e.g. Wong et al., 2021) or the technological advancements that could enhance productivity through improved coordination and planning (c.f., Ghasemi Poor Sabet and Chong, 2020a). Design decisions profoundly shape constructability, coordination requirements, and resource utilisation patterns during execution (Kuzmanovska, 2020). A comprehensive understanding of productivity in the construction industry therefore requires a holistic research approach that considers the entire project lifecycle—from conceptualisation and design through construction, handover, and operation. This lifecycle perspective is particularly pertinent given the industry's growing adoption of Integrated Project Delivery and Modern Methods of Construction, where early coordination and collaborative design are recognised as key determinants of project success (Rahman, 2014).

The quality appraisal conducted for this review (Supplementary Material A) revealed consistent methodological limitations across the included studies. Most reviews did not justify exclusion decisions, did not conduct study selection and data extraction in duplicate, and did not describe all included records in adequate detail. These limitations are not unique to individual studies but reflect disciplinary norms in construction research. The risk, however, is that they constrain the confidence with which findings can be aggregated and compared across reviews.

A notable gap exists in quantitative assessments. The scarcity of meta-analyses is attributable to the inherent heterogeneity of productivity research in construction. Diverse methodologies, definitions, and outcome measures pose significant challenges for quantitative aggregation (Rathnayake and Middleton, 2023). Most quantitative reviews rely on ordinal rankings, which do not permit analysis involving standard deviations and effect sizes. This limits the ability to perform robust meta-analysis across diverse studies. Future research should prioritise the systematic aggregation of quantitative metrics to establish a rigorous empirical baseline. This would enable the transition from descriptive factor identification to predictive modelling of productivity impacts on project performance. Underpinning this heterogeneity is the project-based nature of construction itself (Forsythe, 2018). Unique sites, temporary teams, and bespoke designs generate highly variable contexts that hinder consistent measurement and cross-study comparison.

This variability is not only a methodological challenge but a key characteristic of how the industry has traditionally operated (Styhre and Gluch, 2010). When each project constitutes a unique configuration of site conditions, workforce composition, design specifications, and contractual arrangements, establishing stable baselines for productivity measurement becomes profoundly challenging. The absence of standardised measurement frameworks, identified consistently across the reviewed literature, therefore reflects deeper tensions between the industry's operational reality and the requirements for systematic measurement.

A potential opportunity to address the above limitations lies in platform-based and systems-based approaches that promote industrialisation principles (Styhre and Gluch, 2010). While less investigated in the productivity literature—where industrialisation has been predominantly associated with off-site construction and lean manufacturing principles (Jang et al., 2021)—industrialised construction represents more than a shift in production location. It constitutes a systems approach that challenges traditional project-based business models (Lessing and Brege, 2018).

Drawing on Lessing's (2006) framework, industrialised construction involves multiple interrelated dimensions: long-term thinking and relationships that transcend individual projects; systematic integration of design and production processes; pre-developed building systems and platforms; production planning and control mechanisms; customer focus; and integrated technical systems. Within this framing, production, whether off-site or on-site, is just one element of a broader coordinated system. This systems perspective offers a pathway to address the persistent measurement challenges identified throughout this review. Unlike traditional construction's project-based orientation, which generates the heterogeneous contexts that confound consistent productivity measurement, industrialised approaches stabilise the operational base. Standardised processes, controlled production environments, and systematic planning create the conditions necessary for measurement systems that are consistent and comparable across projects and over time. A platform-based approach provides stable reference points against which productivity can be meaningfully assessed and benchmarked (Robertson and Ulrich, 1998).

Adopting a fuller definition of industrialised construction as a systems approach also enables productivity measurement to extend beyond the construction phase. Design for Manufacture and Assembly (DfMA) principles, integral to industrialised approaches, explicitly acknowledge the interdependencies between design decisions and production efficiency (Kuzmanovska, 2020). This provides a conceptual framework for the lifecycle-wide productivity metrics that the literature has advocated but struggled to operationalise within project-based paradigms. Industrialisation therefore not only offers productivity improvements through technical and process efficiencies but also enables measurement frameworks needed to systematically evaluate and improve productivity across the building lifecycle.

The findings point to a clear need to move beyond fragmented, project-specific productivity measures. What is required are shared and system-level instruments that enable comparability, learning, and strategic decision-making. A key actionable implication is the development of standardised productivity protocols at a national level. These should be grounded in consistent productivity definitions and lifecycle classifications. Standardised protocols would enable the emergence of industry-wide dashboards within organisations, capable of providing transparent, comparable datasets for both internal and external stakeholders. These dashboards should move beyond narrow labour productivity indicators. Multi-factor metrics spanning design, manufacture, construction, and early operation are needed. In practice, this could be operationalised through BIM-integrated datasets linking design intent, production rates, rework, and resource utilisation across project phases (Archchana and Pan, 2023). For policymakers, such dashboards would replace episodic industry inquiries with continuous, evidence-based monitoring of sector performance. This would enable more targeted interventions, benchmarking across delivery models, and evaluation of initiatives such as industrialised construction and Modern Methods of Construction. For industry clients and contractors, a shared measurement infrastructure would reduce ambiguity in performance assessment. It would also support more informed procurement decisions and enable learning across project portfolios rather than within isolated projects.

At the operational and commercial level, the review highlights that many of the field's persistent measurement challenges are structural rather than technical. They arise from the variability inherent in project-based construction. An important practical implication is that industrialised and platform-based delivery models create the preconditions for stabilised productivity measurement (Styhre and Gluch, 2010). Standardised processes, repeatable building systems, and integrated design-production workflows allow productivity metrics to be normalised over time. This enables firms to distinguish between systemic improvement and project-specific variation. This has direct commercial implications: more reliable productivity data supports earlier and better-informed design decisions. It also enables more accurate cost and programme forecasting, reduces contingency pricing, and clarifies the attribution of value across the supply chain.

A shift toward lifecycle-wide and systems-based productivity assessment enables productivity to be aligned with broader public outcomes. These include reduced material waste, lower embodied carbon, improved workforce wellbeing, and safer working conditions (López-Guerrero et al., 2022). Embedding such metrics into education, professional training, and public procurement frameworks would help recalibrate industry norms. Productivity would then be understood not merely as labour efficiency or construction speed, but as the capacity of the construction system to deliver a high-quality, affordable, and sustainable built environment at scale.

This umbrella review maps the evolving landscape of productivity research in the construction industry. The findings reveal a body of work that has predominantly concentrated on on-site productivity factors. Broader project lifecycle considerations and interdependencies between phases remain largely overlooked. Significant research gaps persist, including conceptual ambiguity, inconsistent definitions, and limited quantitative assessment.

Several limitations of this umbrella review require acknowledgement. Selection criteria may have introduced bias in the reviews included. Only English-language reviews were analysed, which may exclude relevant findings from non-English speaking contexts and limit geographical diversity. The review also focused exclusively on existing literature reviews. Primary studies such as empirical research, experiments, case studies, and surveys were outside its scope.

Despite these limitations, the quality appraisal conducted for this review revealed consistent methodological shortcomings across the included studies. Many reviews fall short of the methodological rigour expected for their stated review type (Grant and Booth, 2009). Approximately half of the reviews searched only one database and did not justify exclusions (see Supplementary Material A). Furthermore, some reviews lacked clarity regarding review types, often conflating scoping and systematic reviews, resulting in methodological inconsistencies (Munn et al., 2018). These shortcomings underscore the need for greater methodological rigour in future productivity research synthesis.

Addressing these gaps requires future research to move beyond narrow perspectives. Developing holistic and standardised conceptualisations of productivity—including both efficiency and effectiveness dimensions—is an essential first step. Strengthening methodological rigour in review design and reporting is equally important (see shortcomings in Supplementary Material A). Quantitative approaches capable of evaluating the impact of key drivers, including industrialisation, sustainability, and digitalisation, would substantially advance the evidence base.

As argued in the discussion, the systems approach characteristic of industrialised construction offers a particularly promising avenue for addressing persistent measurement challenges. Standardised processes, integrated planning and control systems, and platform-based delivery models stabilise operational contexts and create the conditions for consistent, comparable productivity measurement across projects and over time. Prioritising human-centred factors and investigating the twin transition of sustainability and digitalisation will also be critical. Emerging technologies such as artificial intelligence and robotics warrant dedicated investigation. The reviewed literature reveals a notable disparity: decarbonisation has received substantially less attention than digitalisation and industrialisation. Integrating sustainability considerations into productivity frameworks is therefore an urgent research priority. By situating these priorities within a lifecycle perspective and adopting systems-level approaches, future studies can generate actionable insights. The aim should be to close existing gaps and chart a path toward more resilient, efficient, and innovative construction practices.

The supplementary material for this article can be found online.