Towards Digital Transformation in the Construction Industry: A Selection Framework of Building Information Modeling Lifecycle Service Providers (BLSPs)

Abstract

Purpose: The construction industry is now experiencing a thorough transformation through digital technologies, especially with building information modeling (BIM). Despite significant BIM advantages, most construction projects suffer from low BIM performance due to the fragmented BIM use mode. To facilitate lifecycle-integrated BIM implementation, this study demonstrates that introducing BIM lifecycle service providers (BLSPs) is feasible and offers significant improvements in terms of BIM benefits. Hence, this study proposes a customized framework to select BLSPs. Approach: This study utilized both qualitative and quantitative methods. It first adopted semi-structured interviews as part of the qualitative method to deduce the initial criteria for BLSPs’ selection. 30 interviews were conducted iteratively with managers proficient and experienced in selecting BLSPs, through which 25 initial criteria were identified. Then, as the basis of the applied quantitative method, a questionnaire survey was used to evaluate these criteria by determining the critical ones, identifying the latent factor groupings, and assigning criteria weights. Subsequently, an assessment framework was established. Finally, the study was in favor of eight construction projects, highlighting the practicality and validity of the framework. Findings: The results depicted that project BIM service capability is a primary factor for BLSPs’ selection. Within this factor, several specialized criteria need to be considered, such as “boundary spanning competence of the BIM manager” and “BIM service plans with lifecycle cognition.” Meanwhile, “past innovative BIM service practices” and “BIM research and development (R&D)” that originate in corporate innovation capacity were emphasized when selecting BLSPs. Furthermore, for holistic assessment and recognizing the peculiarities of digital BIM service, the study found that criteria like “Privacy and security” and “Backup system” are required, which demonstrate BIM service reliability. Originality/value: This study expands on the conventional partner selection frameworks in the construction sector and thus defines and validates a tailored one for BLSPs’ selection. Moreover, drawing such a reference solution from the framework, the study enables the selection of appropriate BLSPs for clients.

1. Introduction

The construction industry, composed of real estate, infrastructure, and industrial structures, contributes to approximately 13% of the world’s GDP annually [1] and guarantees over 7% employment of the global working population [2]. Hence, it plays a pillar role in the world economy. Nevertheless, the industry still faces severe performance challenges, including low productivity, cost, and schedule overruns. For example, over the past two decades, the construction industry achieved only 1% annual labor-productivity growth, significantly lower than the global economy’s 2.8% and the manufacturing industry’s 3.6% [2]. Meanwhile, mega infrastructure projects worldwide have experienced nearly 80% cost overruns and 20% schedule delays [1]. These barriers prevent sustainable development, urging a comprehensive transformation [3].

Recently, digital transformation through advanced technologies has emerged as a key pathway for development in traditional industries, including construction [4]. Though the construction industry is historically conservative in adopting innovations, this trend is gradually shifting. A McKinsey report [5] notes that investments in construction digital technologies doubled to $18 billion over the past decade and continue to proliferate. Among these technologies, building information modeling (BIM) is widely recognized as the core driver of construction digital transformation [4,6,7,8]. As a multitasking method that is independent of time and location and accessible to all project participants, BIM supports the entire building lifecycle management, from design optimization and clash detection to operational asset management [9]. BIM offers remarkable benefits, including enhanced collaboration, reduced errors, improved quality, lower costs, and higher profits [10]. Globally, countries such as the UK, Singapore, and China have issued mandatory regulations to promote BIM adoption [11]. In China, as one of the world’s largest construction markets, the government has strategically prioritized BIM to advance industrial upgrading. Since the release of “The Action Plan for Promoting the Development of BIM Technology” in 2015, China has incorporated BIM into national strategies such as the “14th Five-Year Plan for Digital Economy Development” and the “Guiding Opinions on Promoting the Synergistic Development of Intelligent Construction and Building Industrialization.” Mandatory BIM requirements have been enforced in key projects, including urban rail transit, super high-rises, and green building demonstrations. By 2025, newly built public buildings in China are required to implement BIM throughout their lifecycle, reflecting its critical role in national construction modernization.

However, widespread BIM adoption does not guarantee the realization of its benefits. Practical success depends on technical and managerial prerequisites, such as robust infrastructure, skilled personnel, standardized workflows, and shared data environments [7,12]. Crucially, a lifecycle-integrated BIM application model is essential [13,14]. Prior research indicates that fragmented BIM use by individual partners in isolated stages (design, construction, or operation) fails to deliver full benefits [15,16]. This challenge is particularly prominent in China. While BIM adoption has accelerated, most construction participants lack lifecycle awareness and capabilities for integrated application [17]. Traditional stakeholders (e.g., owners, designers, contractors, and subcontractors) focus narrowly on domain-specific BIM functions [18]. Owners prioritize visualization, designers focus on design reviews, and contractors emphasize constructability checks. Moreover, high BIM investment costs discourage stakeholders from developing lifecycle capabilities beyond their scope [19,20], leading to fragmented use, inefficiency, and ineffectiveness [7,21,22].

This study found that engaging BIM lifecycle service providers (BLSPs), defined as “specialized firms providing lifecycle-integrated BIM services to construction project partners,” is a promising strategy to address these challenges. Through three core service categories (detailed in

Table 1. BIM services in construction projects.

Categories	BIM Services Included (Examples)	Service Outcome	Reference
BIM model creation and management	Creating the model chain: design model, the bidding and tendering model, the construction models (e.g., 4D model, 5D model, as-built model), and the operation and maintenance model Model content management: model integration, model checking, synchronization	Lifecycle-integrated BIM models	[24,25]
BIM application	Planning and Design stage: site analysis, design review, design coordination, energy simulation, constructability analysis	Lifecycle comprehensive BIM use cases	[26,27,28]
	Construction stage: clash detection, schedule simulation, quantity take-off, site logistics, off-site fabrication, spatial coordination
	Operation stage: space management, facility management
BIM consulting	Technical consulting: selecting data acquisition software and technologies; determining the data transfer paths and tools; establishing the data recognition and decoding ways; building up a BIM-based platform Managerial consulting: formulating the BIM Execution plan; establishing the BIM manuals; offering BIM guidelines; providing BIM trainings	Lifecycle-integrated and standardized BIM working flows	[13,17,29]

), BLSPs build integrated models, deliver comprehensive functions, and establish unified workflows to support lifecycle-integrated BIM applications. Selecting appropriate BLSPs is the first step toward realizing BIM benefits in any project. However, research on BLSPs, especially their selection, is still in its infancy [19]. Unlike traditional partners focused on stage-specific tasks, the service scope of BIM service providers must span the entire building lifecycle to enable seamless data flow and technology integration across stages from design to operation. Concurrently, digital transformation demands not only standalone BIM tools but also their integration with emerging technologies such as AI, IoT, blockchain, and digital twins [4,8], which significantly expands the complexity of BIM services. Moreover, digital transformation drives dynamic and personalized BIM demands. Clients increasingly seek advanced applications like environmental simulation, smart supply chain management, and digital twin-based operation, making BLSPs’ service content far more flexible and innovation-dependent than that of traditional partners. Existing frameworks for traditional partner selection are thus inadequate [23]. To fill this gap, this study proposes a customized assessment framework for selecting BLSPs, with implications for both China’s rapidly developing construction sector and global digital transformation efforts.

The paper is arranged as follows. Section 2 presents the research background, followed by an introduction to the research methods in Section 3. Section 4 illustrates the results, while Section 5 and Section 6 discuss the results and implications. Section 7 concludes the findings.

2. State of the Art

In this section, the typical BIM services offered by BLSPs are summarized. Then, based on the service characteristics of BLSPs, challenges for BLSPs’ selection are discussed by comparing BLSPs with traditional partners.

2.1. BIM Services from BLSPs

Against the backdrop of global construction digital transformation and China’s policy-driven BIM promotion, BLSPs have emerged as a rapidly growing segment in the construction technology ecosystem. From a service scope perspective, early BLSPs primarily focused on fragmented BIM model creation, but they have now evolved to offer comprehensive lifecycle services spanning model management, cross-stage application, and standardized consulting. Their development is closely tied to the escalating demand for integrated BIM solutions across project lifecycles, spurred by both top-down policy mandates and bottom-up market needs [19]. Drawing on the project lifecycle BIM use frameworks [13,26], this study summarized services of BLSPs in Table 1.

BIM model creation and management forms the foundation of BIM implementation, as digitalization of a facility’s physical and functional attributes relies on accurate model development. BLSPs build a “BIM model chain” tailored to each project phase, including design, bidding, construction, and operation, to meet stage-specific needs [13]. To support collaborative use, they also provide model content management, such as integrating discipline-specific models into a consolidated repository accessible to all partners [30]. Critical tasks include clash detection during integration to ensure model validity and real-time synchronization of updates, keeping all stakeholders aligned.

BIM application services leverage these models to optimize construction management across stages. Typical applications span site analysis and energy simulation in planning/design; clash detection, schedule simulation, and logistics planning in construction; and space/facility management in operations [27]. These services are highly diverse, adapting to varying user demands [14], and increasingly integrated with advanced technologies like AI, VR, IoT, and blockchain [8,31].

BIM consulting ensures seamless information flow and standardized processes. Technical consulting helps establish data pipelines by selecting acquisition tools, defining transfer paths, and creating data decoding protocols [13], while cloud-based platforms or servers enhance interdisciplinary collaboration efficiency [29]. Managerial consulting includes developing BIM execution plans, drafting manuals, providing guidelines, and delivering training [17,29,32], all of which standardize BIM adoption across the project lifecycle.

2.2. Challenges of BLSP Selection

BLSPs differ significantly from traditional construction partners in two key aspects: broad service scope and flexible service content. These differences make the selection of BLSPs a more complex and challenging process.

As illustrated in Figure 1, which compares the service scopes of BLSPs and other partners, BLSPs operate across the entire building lifecycle, unlike traditional partners who focus on specific project stages. This broad scope introduces unique considerations for BLSP selection. First, BLSPs must possess the ability to coordinate cross-stage partners to enable integrated BIM use, an attribute rarely prioritized when selecting stage-specific partners [33]. Second, their wide-ranging involvement demands a more comprehensive knowledge base. In addition to specialized BIM expertise, BLSPs need familiarity with domain-specific knowledge across all project phases, a requirement not applicable to traditional partners who specialize in individual stages [34]. Third, the long-term nature of lifecycle services means organizational stability is a more critical factor for BLSPs than for short-term stage partners.

In terms of service content, BLSPs differ from traditional partners in the flexibility of their offerings. BLSPs services, particularly BIM applications, are more dynamic due to the evolving demands of users across project stages [14]. Beyond basic BIM applications (outlined in Table 1), users increasingly seek advanced use cases to maximize benefits, such as environmental analysis during design [35], supply chain management during construction [36], and emergency simulation during operations [37]. Additionally, there is growing demand for BIM integration with other digital technologies: BIM combined with blockchain for transparent supply chain management [38], BIM with IoT to create digital twins for enhanced project oversight [39], and BIM with AI to optimize operational decision-making [40]. This flexibility underscores the importance of organizational innovation capability in BLSP selection, an attribute often undervalued when evaluating traditional partners [33].

Given these distinctive service characteristics, existing mature frameworks and criteria for selecting traditional partners are not directly applicable to BLSPs [23]. Thus, this study aims to develop a tailored selection framework specifically for BLSPs.

3. Research Methods

This study consists of two phases with four main steps to develop and validate the assessment framework for BLSPs’ selection (Figure 2). The first step identified the initial BLSPs’ selection criteria through semi-structured interviews. Then, a questionnaire survey was adopted to collect data for evaluating these identified criteria in the second step. The third step used three analytic approaches to determine the critical criteria, identify the latent factor groupings from the criteria, and assign weights to the criteria. Then, the criteria were finalized through the above three steps, and the assessment framework for BLSPs’ selection was established. In the fourth step, the framework of BLSPs’ selection was validated through eight construction projects.

By integrating qualitative and quantitative methods, this study addresses the research gap in BLSPs’ selection, where existing literature lacks systematic criteria and validated frameworks, by deriving context-specific insights from practice and rigorously testing them through empirical data. The mixed-method approach ensures both depth of practical understanding and breadth of statistical validation, enhancing the framework’s reliability and applicability. Particularly, the mixed-method design addresses the limitations of single-method approaches:

Qualitative interviews (Section 3.1) captured context-rich, practice-driven criteria (e.g., “Boundary spanning competence”) that quantitative surveys alone might miss, ensuring ecological validity.

Quantitative surveys and analyses (Section 3.2 and Section 3.3) validated these criteria across a large sample, quantifying their importance and reducing subjectivity, which qualitative methods alone cannot achieve.

Case studies (Section 3.4) integrated both by testing the framework in real-world settings, ensuring the final tool is both theoretically sound and practically applicable.

3.1. Identify the Initial Criteria for BLSP Selection

Semi-structured interviews were employed to identify BLSPs’ selection criteria, as existing literature remains scarce in research on BLSPs [19,20]. This method is particularly suited to exploring nuanced perceptions of practical phenomena like BLSPs’ selection [41], encouraging interviewees to share detailed accounts of their past selection experiences. This study is limited to users of BIM in China for two key reasons. First, prior research highlighted that institutional factors significantly shape BIM adoption practices [32,42]. Thus, this study recognizes that cross-national cultural differences among BIM users could compromise the consistency of responses, undermining efforts to develop a reliable selection framework. Notably, most existing BIM evaluation frameworks are designed within specific national contexts [43]; Second, China’s top-down BIM implementation policies, such as the 2015 “Action Plan for Promoting the Development of BIM Technology”, the inclusion of BIM in national strategies like the “14th Five-Year Plan for Digital Economy Development,” and mandatory BIM requirements for key projects (e.g., urban rail transit, super high-rises, and green building demonstrations), have accelerated BIM adoption [11,44]. This policy-driven ecosystem has equipped many project and BIM managers with rich hands-on experience in selecting BLSPs, making them valuable sources of professional insights into critical selection criteria.

To ensure both reliability and validity of findings, interviewees were selected from project and BIM managers who had overseen award-winning BIM-enabled projects (national or provincial level) in China. A total of 30 iterative interviews were conducted until no new criteria emerged. The interviewee profile (detailed in

Table A1. Profile of the interviewees.

Respondents	Positions	Working Experiences in Construction Industry (Years)	BIM Experiences (Years)	BIM Awards (Ever Won)	Number of Projects Participated in Selecting BLSPs	Types of Projects Participated in Selecting BLSPs
Owner1	PM&BM	25	8	N&P	50~60	H
Owner2	PM&BM	17	12	P	70~80	H&I
Owner3	PM	22	9	N&P	40~50	H
Owner4	PM&BM	22	12	N&P	50~60	H
Owner5	PM	18	11	N&P	60~70	IS
Owner6	PM	13	12	P	60~70	H&I
Owner7	PM	15	10	P	40~50	H&I
Owner8	PM	20	11	P	60~70	H
Contractor1	PM&BM	23	8	P	60~70	H
Contractor2	PM&BM	22	9	N&P	50~60	H&I
Contractor3	BM	28	11	N&P	50~60	H&I
Contractor4	PM	30	8	N&P	50~60	H&I
Contractor5	PM&BM	22	8	N&P	40~50	H&I&IS
Contractor6	PM&BM	18	11	N&P	30~40	H&I&IS
Contractor7	PM	23	12	N&P	40~50	H&I&IS
Contractor8	PM&BM	28	11	N&P	50~60	H&I
Contractor9	PM&BM	20	10	N&P	70~80	H&I
Contractor10	PM&BM	25	11	N&P	50~60	H&I&IS
Contractor11	PM	20	9	P	50~60	H&I&IS
Contractor12	PM&BM	22	8	N&P	80~90	H&I&IS
Contractor13	PM&BM	26	12	N&P	50~60	H&I&IS
Designer1	PM&BM	22	9	P	60~70	H&I
Designer2	BM	18	8	N&P	50~60	H&I
Designer3	BM	14	10	N&P	40~50	H&I&IS
Designer4	BM	14	11	N&P	30~40	H&I&IS
Designer5	PM&BM	20	8	N&P	50~60	H&I&IS
Designer6	PM&BM	22	12	N&P	50~60	H&I
Designer7	PM&BM	20	11	N&P	50~60	H&I
Designer8	PM&BM	22	9	N&P	60~70	H&I
Designer9	PM&BM	20	13	N&P	50~60	H&I&IS

Note: PM = Project manager; BM = BIM Manager; N = National level; P = Provincial level; H = Housing; I = Infrastructure; IS = Industrial.

) included 8 from the owner’s side, 13 from contractors, and 9 from design firms. Most had over 15 years of construction industry experience and over 8 years of BIM practice. Moreover, they had participated in selecting BLSPs for over 50 projects across housing, infrastructure, and industrial sectors, ensuring their competence in identifying relevant criteria. Each interview began with an explanation of the research objectives, followed by open-ended questions prompting interviewees to list criteria they had used in BLSPs’ selection, with follow-up questions for each criterion mentioned to explore its meaning and relevance for clarity. A saturation test was conducted to determine when no new criteria emerged. New criteria continued to surface until the 24th interview, after which responses repeated existing criteria, and the subsequent 6 interviews confirmed saturation with no new additions. This process ultimately yielded 25 distinct criteria, each with detailed descriptions (

Table 2. Criteria for BLSPs’ selection.

Codes	Criteria	Criteria Description
C01	BIM certificates of the BIM manager	BIM certificates awarded by the construction authority or regime to the BIM manager.
C02	Available BIM infrastructure	The adequacy of BIM hardware (e.g., the computing power, the storage volume) and software (e.g., the diversity of software) for the project BIM service.
C03	BIM certificates of the BIM engineer	BIM certificates awarded by the construction authority or regime to the BIM engineer.
C04	BIM service experience of BIM manager	The participation of BIM managers in previous BIM service projects, including the size of projects, the number of projects, similar projects experience, etc.
C05	Market share and reputation	The dominance and influence of the provider in the BIM service market.
C06	Past BIM service experience	Completed BIM service projects by the provider, including the size of projects, the number of projects, similar projects experience, etc.
C07	Exclusive BIM staff for the project	The number of exclusive BIM staff supporting the project BIM service.
C08	Integration with other project digital systems	BIM service can be easily integrated with other project digital systems (e.g., Project management systems) to promote project performance.
C09	Organization’s BIM staff	The number of long-term BIM staff employed by the service provider.
C10	BIM service plan with lifecycle cognition	BIM service plan with core lifecycle cognition considering the digital integration (e.g., model integration) across different project stages.
C11	Service compatibility with partners’ operating systems	BIM service has less impact on partners’ normal operating systems and can be integrated to operating systems.
C12	Boundary spanning competence of the BIM manager	BIM managers can work as boundary spanners among partners to facilitate BIM-based digital integration (e.g., digital model integration, and workflow integration).
C13	BIM efficiency and effectiveness	The whole project performance improvement from BIM service.
C14	BIM awards	The recognition of the excellent BIM service awarded by the construction authority or regime.
C15	BIM service claims	Conflicts for disabled BIM service (i.e., BIM service failures).
C16	BIM strategy and vision	The company-level strategic target, route, and vision around the provider’s BIM service.
C17	Finance stability	The soundness of the cash flow of the BIM service provider.
C18	BIM research and development (R&D)	The resources devoted to BIM research and development (R&D) for BIM service innovation.
C19	Experience in the construction market	Years that the service provider worked in the construction market.
C20	Past partnership and trust relationship with customers	Established partnership and trust relationships with served customers for successful BIM services
C21	Past innovative BIM service practices	Exploitative (e.g., BIM advanced use cases during building lifecycle) and explorative (e.g., BIM integration with other digital technologies) BIM services offered by the provider.
C22	Privacy and security	Data protection measures during the BIM service.
C23	Backup system	Data recovery measures for abnormal situations (e.g., cyber-attacks) during the BIM service.
C24	Complete service bidding document	BIM service bidding document satisfying the basic project and customer requirements.
C25	Cost	Setup cost (e.g., building up BIM models, conducting specific BIM applications) and continuous BIM cost (e.g., model maintenance).

).

3.2. Collect Data for Evaluating Criteria for BLSP Selection

To compare the relative significance of the identified criteria and determine the critical ones for BLSPs’ selection, this study employed a questionnaire survey. As a systematic data collection method, questionnaire surveys are well-suited for soliciting professional opinions in BIM research, facilitating the generation of widely accepted conclusions [16,42]. Notably, researchers commonly relied on questionnaire surveys for data collection when developing BIM-related evaluation frameworks [45].

The questionnaire was structured into three main sections. The first section provided a detailed explanation of the research objectives, enhancing respondents’ understanding of the study’s purpose. It also allows those uninterested or unfamiliar with BLSPs’ selection to opt out, thereby saving time for both parties. The second section gathered respondents’ background information, including their professional roles, years of experience, and qualifications, to support subsequent demographic analysis. The third and core section asked respondents to rate the importance of the 25 identified criteria using a five-point Likert scale (1 = not important, 2 = less important, 3 = neutral, 4 = important, 5 = very important). Prior to formal distribution, the questionnaire’s reliability was pre-tested by the 30 interviewees from the qualitative phase. Cronbach’s alpha ($\alpha$) was then used to assess internal consistency, with values ranging from 0 to 1. The higher scores indicate greater reliability. A threshold of alpha ($\alpha$) ≥ 0.7 is generally accepted for adequate reliability [46].

Consistent with the semi-structured interviews, the questionnaire survey was restricted to users of BIM in China (see Section 3.1 for rationale). A snowball sampling approach was adopted, leveraging the network of the initial 30 interviewees, who were asked to share the questionnaire with other qualified professionals in China. A total of 257 questionnaires were distributed via email, and 178 valid responses were ultimately collected, with validity defined by respondents having participated in BLSPs’ selection for at least one project. This yielded a response rate of 69.2%, which is substantially higher than those reported in prior BIM-related studies in China (e.g., [47]).

Table 3. Profile of the survey respondents.

	Frequency	Proportion
Respondents	178 (total)	100% (total)
Working role
Contractor	104	58.4%
Designer	43	24.2%
Owner	31	17.4%
Company size
Large	32	18.0%
Medium	82	46.0%
Small	64	36.0%
Working positions
Project manager/BIM manager	118	66.3%
Chief engineer	34	19.1%
Engineer	26	14.6%
Construction experience
1~5 years	6	3.4%
5–10 years	75	42.1%
10–15 years	48	27.0%
Above 15 years	49	27.5%
BIM experience
1~5 years	32	18.0%
5–10 years	128	71.9%
10–15 years	18	10.1%
Number of projects participated in selecting BLSPs
1~10	21	11.8%
10~30	33	18.5%
30~50	85	47.8%
Above 50	39	21.9%

Note: Large firm = revenue of 800 million or more RMB, Medium firm = revenue of 60 million to under 800 million RMB, Small firm = revenue of less than 60 million RMB [48].

presents the demographic profile of respondents. They represented three key professional roles: contractors (58.4%), designers (24.2%), and owners (17.4%). This distribution is justified by the practical concentration of BIM applications in the construction stage [20]. Hence, more contractors were invited to participate in the survey. In terms of company size, 18% of respondents were from large firms, 46% from medium-sized firms, and 36% from small firms, ensuring representation across different organizational scales and reducing potential bias. Additionally, 85.4% of respondents held executive positions (e.g., project manager, BIM manager, or chief engineer), and over 50% had more than ten years of experience in the construction industry, indicating strong industry expertise. Regarding BIM experience, 71.9% of respondents had 5–10 years of practice, aligning with China’s BIM development trajectory, where widespread adoption has occurred over the past decade. Finally, over 88% of respondents had participated in BLSPs’ selection for more than ten projects, confirming their competence to provide reliable evaluations of the criteria. Therefore, this study considers the selected respondents as best fit to analyze and evaluate BLSPs’ selection based on their rich experience.

3.3. Assess and Determine Critical Criteria and Potential Criteria Groupings for BLSP Selection

(1)

Ranking analysis

To rank the importance of the identified criteria, this study employed mean score (MS) ranking and standard deviation (SD) analysis, two methods widely adopted in BIM research [42,49]. MS was used to determine the relative importance and rankings of the criteria. For criteria with identical mean scores, those with lower SD values were ranked higher, indicating greater consensus among respondents. To identify critical criteria from the 25 initial ones, normalized mean scores were calculated using Equation (1) [32]. The normalized mean of a criterion is defined as the ratio of the difference between its mean score and the minimum mean score of all criteria to the difference between the maximum and minimum mean scores in the dataset. Criteria with a normalized mean of ≥0.5 were classified as critical, while others were excluded.

$$\mathrm{N}\mathrm{o}\mathrm{r}\mathrm{m}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{z}\mathrm{e}\mathrm{d} \mathrm{m}\mathrm{e}\mathrm{a}\mathrm{n}=\left(mean-minimummean\right)/\left(maximummean-minimummean\right)$$

(1)

Given the diversity of respondent groups (owners, contractors, and designers), a ranking agreement analysis was further conducted. Kendall’s coefficient of concordance (W) was used to assess the ranking agreement within each respondent group [50]. Ranging from 0 to 1, a higher W value indicates stronger agreement within the group. The null hypothesis for Kendall’s W assumes “no agreement in rankings within the group.” A significance level <0.05 leads to the rejection of this hypothesis, indicating significant intra-group consensus. Conversely, an analysis of variance (ANOVA) test was applied to examine inter-group ranking agreement [51]. Using a significance level < 0.05, ANOVA determines whether statistically significant differences exist in mean score rankings across respondent groups.

(2)

Factor analysis

To identify underlying groupings of criteria, factor analysis was employed, a statistical method that explores inter-variable relationships to extract a smaller set of latent factors [52]. Principal component analysis (PCA) was used for factor extraction, consistent with its common application in BIM research for variable grouping [32]. Prior to factor analysis, the Kaiser-Meyer-Olkin (KMO) test of sampling adequacy and Bartlett’s test of sphericity were conducted to validate its suitability [52]. A KMO value ≥ 0.5 (ranging from 0 to 1) indicates that distinct factors can be extracted, while a significant Bartlett’s test of sphericity (large statistic value with p < 0.05) confirms that the correlation matrix is non-identity, justifying the use of factor analysis [52].

(3)

Weighting analysis

Based on the factor analysis results, the weights of the latent factors (factor groupings) can be concluded through the following Equation (2).

$$W_{Fj}={\displaystyle \frac{V_{Fi}}{\sum V_{Fj}}}$$

(2)

where $W_{Fj}$ is the weight of latent factor j; $V_{Fj}$ is the extracted variance of latent factor j; $\sum V_{Fj}$ is the summation of extracted variance of all the latent factors.

Meanwhile, the local weight of each criterion to its belonging latent factor can be calculated through Equation (3).

$$W_{Cij}={\displaystyle \frac{L_{Ci}}{\sum _j{L}_{Ci}}}$$

(3)

where $W_{Cij}$ is the local weight of criterion i to its corresponding latent factor j. $L_{Ci}$ is the highest loading of criterion i to its corresponding latent factor j; $\sum _j{L}_{Ci}$ is the summation of loadings of all the criteria belonging to the latent factor j.

Finally, the global weight of each criterion can be determined by Equation (4).

$$W_{Ci}=W_{Fj}\times W_{Cij}$$

(4)

where $W_{Ci}$ is the global weight of criterion i; $W_{Fj}$ is the weight of latent factor j; $W_{Cij}$ is the local weight of criterion i to the latent factor j.

3.4. Validate the Assessment Framework for BLSP Selection

The proposed assessment framework was validated through practical case studies. Eight construction projects were selected, all of which had previously selected appropriate BLSPs (primarily based on experience) and achieved significant BIM benefits, such as reduced errors/rework, cost savings, and improved customer satisfaction. Framework reliability was evaluated by comparing its predicted selection results with actual project outcomes [53].

Table A2. Profile of the projects for validation.

Project Number	Project Type	Project Scale (Million Yuan)	Project Nature	Project Location	Number of Qualified BLSPs Bided	BLSP Selected
1	School	35	Public	East	5	A2
2	Real Estate	48	Private	Southeast	5	B3
3	Hospital	89	Private	East	7	C2
4	Semi-conductor factory	220	Private	Southeast	5	D4
5	Power plant	454	Public	Southwest	5	E1
6	Rail station	115	Public	North	7	F5
7	Bridge	821	Public	Southeast	5	G1
8	Underground tunnel	642	Public	East	5	H3

Note: BLSPs bid for project 1 are A (1~5), and project 2 are B (1~5), and project 3 are C (1~7), and project 4 are D (1~5), and project 5 are E (1~5), and project 6 are F (1~7), and project 7 are G (1~5), and project 8 are H (1~5).

details the profiles of these eight projects, covering housing (first three), industrial (middle two), and infrastructure (last three) categories. They varied in scale, nature, location, and number of bidding BLSPs, ensuring the framework’s applicability across diverse scenarios. For validation, each project’s bidding team first rated the performance of bidding BLSPs against each criterion on a 0–10 scale. These ratings were then multiplied by the criteria’s global weights and summed to generate a global score for each BLSP, with higher scores indicating better performance. The BLSP with the highest global score was identified as the optimal choice. For any projects with “N” number of bided BLSPs (N is an odd number less than 10), the followings were the calculation equations for selecting the BLSPs.

$$G_j=\sum _{i=1}^{25}{S}_{ij}\times W_{ci};G_j^{*}=Max\left\{G_j\right\}\left(j=1toN\right)$$

(5)

where $S_{ij}$ depicts the score of criterion i to the BLSP j, $W_{ci}$ is the global weight of criterion i, $G_j$ reflects the global score of BLSP j, $G_j^{*}$ represents the final BLSP j that achieves the highest global score.

4. Results

4.1. Ranking Results

The overall Cronbach’s alpha ($\alpha$) for all criteria was 0.816, exceeding the recommended threshold of 0.7 [46], indicating high internal consistency at the 5% significance level. This finding confirms the reliability of the measurement items, validating the subsequent ranking and factor analysis results.

Table 4. Ranking of the criteria for BLSP selection.

Code	All Respondents (N = 178)				Designer (N = 43)			Owner (N = 31)			Contractor (N = 104)			ANOVA
	Mean	SD	Rank	Normalized Mean	Mean	SD	Rank	Mean	SD	Rank	Mean	SD	Rank	F	Sig
C14	4.44	0.744	1	1.00 a	4.42	0.763	1	4.23	0.921	3	4.52	0.668	1	1.910	0.151
C13	4.40	0.820	2	0.98 a	4.37	0.874	2	4.10	1.106	4	4.51	0.668	2	3.149	0.045 b
C20	4.39	0.698	3	0.97 a	4.35	0.573	4	4.29	0.864	2	4.43	0.693	3	0.581	0.560
C15	4.35	0.739	4	0.95 a	4.37	0.618	3	4.10	1.044	5	4.42	0.664	4	2.381	0.095
C04	4.30	0.696	5	0.91 a	4.21	0.773	6	4.45	0.723	1	4.30	0.652	5	1.102	0.335
C21	4.16	0.987	6	0.83 a	4.30	0.887	5	3.74	1.237	18	4.23	0.916	6	3.600	0.029 b
C06	4.14	0.695	7	0.81 a	4.19	0.824	7	3.94	0.629	12	4.18	0.650	7	1.645	0.196
C12	4.11	0.736	8	0.79 a	4.02	0.801	14	4.03	0.706	8	4.17	0.717	8	0.852	0.428
C25	4.08	0.571	9	0.78 a	4.05	0.688	13	4.00	0.447	9	4.13	0.552	9	0.693	0.501
C22	4.08	0.809	10	0.78 a	4.12	0.851	10	3.94	0.892	10	4.12	0.767	11	0.633	0.532
C11	4.06	0.730	11	0.76 a	3.93	0.856	20	4.06	0.680	6	4.12	0.687	10	0.979	0.378
C24	4.04	0.667	12	0.75 a	4.14	0.710	9	3.81	0.601	15	4.07	0.658	12	2.509	0.084
C23	4.00	0.844	13	0.72 a	4.07	0.828	12	4.03	0.795	7	3.96	0.869	15	0.275	0.760
C10	3.99	0.705	14	0.72 a	4.00	0.690	16	3.81	0.654	16	4.04	0.723	13	1.305	0.274
C05	3.96	0.801	15	0.70 a	4.16	0.814	8	3.81	0.792	14	3.91	0.790	17	2.146	0.120
C07	3.94	0.757	16	0.69 a	4.12	0.851	11	3.87	0.763	13	3.89	0.709	18	1.491	0.228
C08	3.94	0.711	17	0.69 a	3.91	0.750	21	3.94	0.772	11	3.96	0.682	16	0.091	0.913
C19	3.94	0.745	18	0.68 a	3.98	0.831	18	3.77	0.669	17	3.97	0.730	14	0.909	0.405
C02	3.83	0.727	19	0.61 a	4.00	0.787	17	3.71	0.693	20	3.79	0.706	20	1.782	0.171
C16	3.81	0.784	20	0.61 a	3.81	0.795	22	3.68	0.702	21	3.86	0.806	19	0.615	0.542
C17	3.79	0.773	21	0.59 a	4.00	0.787	15	3.71	0.783	19	3.72	0.756	22	2.191	0.115
C18	3.77	0.772	22	0.58 a	3.93	0.768	19	3.58	0.807	22	3.76	0.757	21	1.886	0.155
C09	3.63	0.694	23	0.49	3.81	0.732	23	3.48	0.626	23	3.61	0.689	23	2.289	0.104
C03	3.45	0.890	24	0.38	3.19	0.958	24	3.29	0.824	24	3.61	0.852	24	4.128	0.018 b
C01	2.85	0.752	25	0.00	3.00	0.873	25	2.90	0.651	25	2.78	0.723	25	1.401	0.249
					W = 0.153 (Sig = 0.000)			W = 0.169 (Sig = 0.000)			W = 0.203 (Sig = 0.000)

Note: SD = Standard deviation; Sig = Significance; W = Kendall’s coefficient of concordance; ^a The normalized mean indicates that the criterion is critical ($\ge$0.5); ^b The ANOVA result is significant at 0.05 level (significance < 0.05).

presents the results of the ranking analysis and agreement assessment. The mean scores of the criteria ranged from 2.85 to 4.44, with the top three criteria being “BIM awards” (C14), “BIM efficiency and effectiveness” (C13), and “Past partnership and trust relationship with customers” (C20). These top-ranked criteria inherently reflect BLSPs’ historical service performance and quality, aligning with prior research that highlights the importance of historical indicators in construction service provider selection [23].

Based on the normalized mean threshold (≥0.5), 22 out of the initial 25 criteria were classified as critical. Notably, personal BIM certificates (C01: BIM certificates of the BIM manager; C03: BIM certificates of the BIM engineer) were excluded, as they were deemed less necessary for BLSPs’ selection. While such certificates may partially reflect individual BIM competence, their limited recognition (due to the absence of a unified national BIM training and certification system in China) will undermines their practical value [11]. In contrast, personal BIM experience (e.g., C04: BIM service experience of BIM manager, ranked 5th) was prioritized over formal certifications. Additionally, “Organization’s BIM staff” (C09) received low importance, as clients appeared to value project-specific resources (“Exclusive BIM staff for the project,” C07, ranked 16th) more than the total number of BIM personnel employed by the BLSPs.

Kendall’s coefficient of concordance (W) revealed intra-group consensus, with values of 0.153 (designers), 0.169 (owners), and 0.203 (contractors) all significant (p = 0.000), indicating strong agreement within each professional group. ANOVA results further showed that 22 of the 25 criteria exhibited no statistically significant differences across groups (p > 0.05). The exceptions were “BIM efficiency and effectiveness (C13),” “Past innovative BIM service practices (C21),” and “BIM certificates of the BIM engineer (C03).” A striking divergence emerged for “Past innovative BIM service practices (C21)”: owners ranked it 18th, significantly lower than designers (5th) and contractors (6th). This discrepancy reflects differing priorities: owners typically focus on basic BIM applications (e.g., visualization and virtual tours), while contractors and designers actively seek advanced BIM use cases to enhance design and construction productivity [35,36,54], thus placing greater value on BLSPs’ innovative track records.

4.2. Factor Analysis Results

KMO test yielded a value of 0.832, which is above the threshold of 0.5 and indicates a great degree of common variance [52]. This result confirms that the dataset is suitable for factor analysis, as distinct clusters can be derived from the criteria. Concurrently, Bartlett’s test of sphericity produced a large statistic value of 1588.717 with a significance level of p = 0.000, demonstrating that the correlation matrix of the criteria is non-identity. Together, the KMO and Bartlett’s test results validate the appropriateness of conducting factor analysis to identify latent groupings within the 22 critical criteria.

Principal component analysis (PCA) with varimax rotation was applied to the 22 critical criteria using data from 178 responses.

Table 5. Factor analysis of the criteria for BLSP selection.

Code	Criteria For BLSP Selection	Factor Loadings					Eigenvalue	Variance (%)	Cumulative Variance (%)
		1	2	3	4	5
Factor 1: Project BIM service capability							3.945	17.932	17.932
C04	BIM service experience of BIM manager	0.677	−0.012	−0.026	−0.051	0.255
C07	Exclusive BIM staff for the project	0.741	0.159	−0.098	0.074	0.077
C12	Boundary spanning competence of the BIM manager	0.749	0.285	0.036	−0.010	−0.101
C10	BIM service plan with lifecycle cognition	0.636	0.418	−0.027	0.049	−0.043
C11	Service compatibility with partners’ operating systems	0.691	0.143	−0.055	−0.166	0.251
C02	Available BIM infrastructure	0.727	0.167	−0.059	−0.030	0.223
C08	Integration with other project digital systems	0.690	0.324	0.057	−0.091	0.107
Factor 2: Organization BIM service capability							3.402	15.464	33.395
C17	Finance stability	0.104	0.590	−0.089	0.288	0.365
C05	Market share and reputation	0.136	0.585	0.053	−0.042	0.328
C19	Experience in the construction market	0.325	0.722	0.049	−0.161	0.099
C06	Past BIM service experience	0.346	0.646	0.147	−0.012	−0.126
C16	BIM strategy and vision	0.174	0.791	−0.028	0.023	0.035
C18	BIM research and development (R&D)	0.159	0.759	0.013	−0.042	0.138
Factor 3: Past BIM service performance and quality							2.867	13.034	46.429
C15	BIM service claims	−0.070	0.078	0.737	0.198	−0.155
C21	Past innovative BIM service practices	−0.125	0.123	0.651	0.338	−0.089
C20	Past partnership and trust relationship with customers	0.034	−0.058	0.773	0.122	0.099
C14	BIM awards	−0.019	0.024	0.768	0.036	−0.070
C13	BIM efficiency and effectiveness	0.012	−0.016	0.760	−0.128	0.052
Factor 4: BIM service reliability							1.765	8.022	54.451
C23	Backup system	−0.120	0.009	0.187	0.828	0.050
C22	Privacy and security	0.011	−0.079	0.145	0.855	−0.045
Factor 5: BIM service document and cost							1.582	7.192	61.643
C24	Complete service bidding document	0.165	0.312	0.035	0.020	0.756
C25	cost	0.340	0.083	−0.147	−0.031	0.680

Note: The threshold of the factor loading is set as 0.5 and depicted in bold.

presents the factor analysis outcomes, revealing five latent factors with eigenvalues greater than 1. These factors collectively explain 61.643% of the total variance, exceeding the minimum acceptable threshold of 60%, indicating that the extracted factors adequately represent the dataset [51]. Each criterion’s highest loading on its corresponding latent factor is highlighted (with a threshold of 0.5). Higher loadings signify stronger association between the criterion and its factor [55]. Overall, the factor loadings demonstrated good reliability. Based on the common characteristics of the criteria within each factor, the five latent factors were labeled as “Project BIM service capability” (factor 1), “Organization BIM service capability” (factor 2), “Past BIM service performance and quality” (factor 3), “BIM service reliability” (factor 4), and “BIM service document and cost” (factor 5). The naming of the five factors was mainly based on the common features of the included criteria.

4.3. Weighting Results and the Selection Assessment Framework

Through Formulas (2)–(4), the weights of the latent factors, the local weight of each criterion to its belonging latent factor, and the global weight of each criterion can be calculated and summarized in

Table 6. Factor and criteria weights.

Factor	Weights	Criteria	Local Weights	Global Weights
Project BIM service capability	0.291	C04	0.138	0.040
		C07	0.150	0.044
		C12	0.152	0.045
		C10	0.130	0.038
		C11	0.141	0.041
		C02	0.148	0.042
		C08	0.141	0.041
Organization BIM service capability	0.251	C17	0.144	0.036
		C05	0.143	0.036
		C19	0.177	0.044
		C06	0.158	0.041
		C16	0.193	0.048
		C18	0.185	0.046
Past BIM service performance and quality	0.211	C15	0.201	0.043
		C21	0.176	0.038
		C20	0.209	0.043
		C14	0.208	0.044
		C13	0.206	0.043
BIM service reliability	0.130	C23	0.491	0.064
		C22	0.509	0.066
BIM service document and cost	0.117	C24	0.491	0.061
		C25	0.509	0.056

. Accordingly, an assessment framework for BLSPs’ selection was established in Figure 3.

4.4. Validation of BLSP Selection Assessment Framework

Using Equation (5), the performance of BLSPs candidates for each construction project was evaluated, with results summarized in

Table 7. BLSPs selected from the proposed framework for the eight projects.

Project Number	BLSPs Candidates and Their Scores	Consistent or Inconsistent with the Practically Selected BLSPs
1	A1 (6.385); A2 (7.446); A3 (6.059); A4 (6.678); A5 (5.719)	Consistent
2	B1 (6.041); B2 (6.175); B3 (7.384); B4 (6.543); B5 (6.808)	Consistent
3	C1 (5.266); C2 (7.123); C3 (6.285); C4 (6.709); C5 (5.646); C6 (6.458); C7 (6.191)	Consistent
4	D1 (5.965); D2 (6.087); D3 (5.994); D4 (7.004); D5 (5.867)	Consistent
5	E1 (6.886); E2 (7.373); E3 (6.795); E4 (6.023); E5 (6.273)	Inconsistent
6	F1 (5.63); F2 (5.923); F3 (6.603); F4 (6.063); F5 (7.43); F6 (6.025); F7 (6.032)	Consistent
7	G1 (7.46); G2 (6.514); G3 (6.375); G4 (6.362); G5 (6.567)	Consistent
8	H1 (6.289); H2 (6.127); H3 (6.687); H4 (7.111); H5 (6.336)	Inconsistent

Note: Bolded dark numbers are those best BLSPs deduced from the proposed framework. Bolded red numbers are those practically selected BLSPs inconsistent with the deduced ones.

. This table also compares the framework-deduced BLSP selections with the actual choices made in practice. Six out of the eight projects (Projects 1, 2, 3, 4, 6, and 7) showed consistency between the framework’s recommendations and real-world selections, achieving a 75% consistency rate. This exceeds the 50% threshold typically required for model validation [53], indicating that the assessment framework is applicable and reliable for BLSPs’ selection across diverse project contexts.

To explore the reasons behind the two inconsistent cases (Projects 5 and 8), follow-up interviews were conducted with project managers. They confirmed that the framework’s deduced results aligned with their expectations of the optimal BLSPs but explained that sub-optimal choices were ultimately made due to strategic security considerations. Notably, these security concerns differ from the operational security criteria (C22: Privacy and security; C23: Backup system) included in the framework. As critical national infrastructure projects, senior leaders of Projects 5 and 8 prioritized strategic security, expressing a preference for state-owned enterprises to provide BIM services. Consequently, the bidding teams incorporated these strategic considerations into their decision-making, leading to selections that diverged from the framework’s recommendations. This finding offers valuable insights for refining the BLSP selection framework from a strategic perspective. It also highlights that while the framework is robust for technical and operational evaluations, it is not a one-size-fits-all solution. Instead, it should serve as a reference model, with flexibility to accommodate context-specific strategic factors in critical projects.

5. Discussion

5.1. The Necessity of the Five Factors for BLSP Selection

Project BIM service capability, accounting for the largest variance (17.932%) and 29.1% of the assessment framework, is identified as the dominant factor. To enable lifecycle BIM implementation, project managers primarily focus on whether BLSPs possess adequate project-specific BIM resources and capabilities. First, projects with varying complexity and diversified client demands often require customized BIM solutions [14], making it essential for BLSPs to provide flexible and sufficient technical and human resources (i.e., “available BIM infrastructure” and “exclusive BIM staff for the project.”). These resources are widely recognized as prerequisites for effective BIM application [7,12]. Notably, as core personnel, BIM managers from BLSPs must have relevant experience. Experienced managers can enhance service efficiency and quality through optimal solutions. Furthermore, to support lifecycle-integrated BIM use, BLSPs managers must act as boundary spanners, developing lifecycle BIM plans to ensure consistent information flows and workflows. This aligns with prior findings that transforming fragmented BIM use into an integrated mode depends on the boundary-spanning capabilities of key partners [15,21]. Additionally, service compatibility and integration capability are critical. Incompatibilities with partners’ operating systems can cause data integration issues and errors, undermining service effectiveness, while integration with existing project systems (e.g., project management systems) amplifies BIM benefits. Collectively, these seven criteria establish project BIM service capability as the primary factor in BLSPs’ selection.

Organization BIM service capability explains 15.464% of total variance and constitutes 25.1% of the framework. Given that construction stakeholders prefer collaborating with more established enterprises (e.g., state-owned entities) [33], assessing organizational capability is vital for BLSPs’ selection. “Past BIM service experience” and “market share and reputation” directly reflect organizational capability. On one hand, accumulated project experience enhances organizational competence [56]. On the other hand, strong capability enables BLSPs to secure more contracts, gaining greater market share and reputation. Due to the long-term nature of BLSPs services, “experience in the construction market” and “financial stability” indicate stable organizational capability. BLSPs with rich industry experience and sound cash flows are more resilient in competitive markets [19]. Moreover, “BIM vision and strategy” and “BIM research and development (R&D)” reflect growth potential. Clear strategies guide sustainable development, while R&D drives innovation, fostering competitive advantages. These six criteria underscore the necessity of organizational BIM service capability in BLSPs’ selection.

Past BIM service performance and quality accounts for 13.034% of variance and 21.1% of the framework. After evaluating project and organizational capabilities, this factor examines whether BLSPs have delivered high-quality services historically. First, “BIM service efficiency and effectiveness” must be assessed, as they measure improvements in efficiency (e.g., faster construction cycles, higher productivity) and effectiveness (e.g., fewer errors, reduced rework, better coordination, enhanced quality, and sustainability) enabled by BIM services in past projects [21]. Second, “BIM service claims” and “past partnership and trust relationship with customers” indirectly reflect performance. More claims indicate lower satisfaction and poorer quality, while strong partnerships signal high satisfaction and accolades. Additionally, “past innovative BIM service practices” are critical. Exploitative innovations (e.g., environmental analysis, supply chain management) and explorative innovations (e.g., BIM integrated with blockchain or IoT) generate significant project benefits [54], elevating service quality. Finally, “BIM awards” serve as a comprehensive indicator, as they recognize projects with exceptional BIM performance and quality [19]. These five criteria highlight the importance of historical performance in BLSPs’ selection.

BIM service reliability explains 8.022% of variance and 13% of the framework, with “privacy and security” and “backup system” as core criteria. Privacy and security are major concerns in IT service selection [57]. Amid national initiatives for information security in China, clients increasingly prioritize construction data protection. As BLSPs are responsible for generating and hosting BIM model data (Table 1), their ability to safeguard privacy, through measures like confidentiality agreements and local data servers, is critical to mitigating risks. Backup systems are equally essential, enabling recovery of BIM data lost to cyber-attacks or other anomalies. These two criteria confirm the necessity of BIM service reliability in BLSPs’ selection.

BIM service document and cost accounts for 7.192% of variance and 11.7% of the framework. “Complete service bidding documents” and “cost” are indispensable in selecting traditional construction partners [23] and remain critical for BLSPs. While bid documents vary by BLSPs, their completeness in meeting basic project and client BIM requirements must be evaluated. Additionally, despite regulated BIM service pricing in many Chinese provinces (focused on setup costs), cost, especially for maintaining BIM deliverables (e.g., models), remains a competitive factor. BIM deliverables require long-term maintenance even post-building demolition, and BLSPs may offer varied, cost-competitive maintenance strategies. Thus, document completeness and cost are critical for BLSPs’ selection.

5.2. Comparison Between BLSPs and Traditional Partners Selection

The results confirm significant differences in selection criteria between BLSPs and traditional construction partners (e.g., designers, contractors, subcontractors). The primary distinction lies in the emphasis on project service capability. Compared to traditional partners, BLSPs require more advanced and comprehensive managerial capabilities. According to the findings, BLSPs managers must possess lifecycle awareness to develop integrated BIM use plans and boundary-spanning competence to ensure consistent information flows and workflows across project stages. These senior managerial capabilities are not prioritized in traditional partner selection, as traditional roles are limited to individual project stages and only demand domain-specific expertise (e.g., subcontractors specializing in construction-stage activities) [34]. Additionally, due to the linear nature of traditional project management, partners operate in silos, making compatibility and data integration non-critical selection factors [23]. In contrast, “service compatibility with partners’ operating systems” and “integration with other project digital systems” are essential for BLSPs to support cross-stage integrated BIM use.

A second key difference relates to innovation capability requirements. The findings indicate that clients prioritize “past BIM service innovations” when selecting BLSPs, driven by the need for flexible exploitative and explorative BIM use cases (Section 2.2). This demand also elevates “BIM research and development (R&D)” as a critical criterion for BLSPs. In contrast, while recent studies advocate for greater focus on innovation in traditional partner selection [58], such capabilities remain non-essential in practice. Construction owners tend to be conservative, exhibiting resistance to change and innovation [7].

Finally, due to the digital nature of BIM services, “Privacy and security” measures and the presence of a “Backup system” are required for BLSPs’ selection, factors irrelevant to traditional partner selection [33].

Despite these differences, BLSPs and traditional partners share some selection commonalities. For example, organizational capability criteria such as “Market share and reputation,” “Construction market experience,” and “Past service experience” are standard for both. Additionally, “Complete service bidding documents” and “Cost”, long-standing considerations in traditional partner selection, are equally applicable to BLSPs. Nevertheless, these commonalities do not overshadow the distinct requirements that make BLSPs’ selection a unique process compared to conventional partner selection.

6. Implications

The study yields implications for both theory and practice in the field of construction digital transformation. For theory, this study advances traditional partner selection frameworks by developing a customized assessment framework tailored to BLSPs. This framework addresses the unique demands of lifecycle-spanning service scope and flexible service content of BLSPs, distinct from criteria for traditional partners. Specifically, the framework identifies critical additional criteria for BLSP selection, including project service capability indicators such as “Boundary spanning competence of the BIM manager,” “BIM service plan with lifecycle cognition,” “Service compatibility with partners’ operating systems,” and “Integration with other project digital systems.” It also emphasizes innovation-related criteria like “BIM research and development (R&D)” and “Past innovative BIM service practices,” as well as digital characteristic criteria such as “Privacy and security” and “Backup system.” While this study’s framework is developed based on China’s construction context, its core logic and structural design hold potential for international adaptation, with context-specific adjustments. The framework’s emphasis on lifecycle integration capability, innovation capacity, and digital reliability aligns with global trends in construction digital transformation, where BIM is increasingly adopted as a cross-phase collaborative tool. These universal principles, such as prioritizing boundary-spanning competence for cross-stage coordination and valuing past innovative practices, address common challenges in lifecycle BIM implementation, making them relevant beyond China. Therefore, this framework establishes a foundational reference for future research, enabling further refinement of criteria to enhance reliability and facilitating investigations into specialized BLSP selection frameworks adapted to other national contexts.

For practice, the findings offer valuable guidance to clients, particularly project owners, who often lack specialized knowledge and rely on generic criteria (e.g., cost, organizational capability) for BLSPs’ selection. Different owners can leverage the framework to align BLSP selection with project-specific priorities. For example, critical infrastructure owners (e.g., urban rail transit, bridges, and power plants) should prioritize “Privacy and security” and “Backup system” due to the sensitivity of construction data and national security requirements. Additionally, “Boundary spanning competence of the BIM manager” is critical for coordinating multi-stakeholder collaborations across long-term construction and operation phases. Residential and commercial project owners (e.g., housing complexes and office buildings) may focus more on “Cost” and “Service compatibility with partners’ operating systems”. For private real estate projects, optimizing lifecycle BIM costs, including model maintenance fees, directly impacts investment returns. Ensuring BLSPs’ services integrate seamlessly with existing property management systems also enhances post-construction operational efficiency, a key concern for residential developers. By applying the proposed framework, owners can conduct comprehensive assessments of BLSPs capabilities, ensuring the selection of optimal partners and enhancing BIM performance in construction projects. Moreover, as owners witness transformative BIM benefits, they are more likely to promote widespread BIM adoption, accelerating the industry’s digital transformation [12].

BLSPs also stand to benefit from the insights. To gain competitive advantage in the BIM service market, they should prioritize the critical factors identified in the framework. First, strengthening project BIM service capability is essential, particularly by training managers in advanced competencies like boundary spanning and lifecycle cognition, as this factor dominates client selection decisions. Second, sustained innovation in service practices, supported by investments in BIM R&D, is crucial to meeting dynamic client demands. Third, elevating service performance and quality builds client confidence in their ability to deliver excellence. Additionally, implementing robust data protection solutions alleviates client concerns about privacy, enhancing trust in service reliability. Finally, offering cost-effective services by optimizing maintenance costs for BIM deliverables can create a competitive edge in pricing.

7. Conclusions

Digital transformation in the construction industry hinges on the effective implementation of BIM, yet fragmented BIM use remains a critical barrier to unlocking its full potential. This study addresses the key gap by proposing a tailored assessment framework for selecting BIM Lifecycle Service Providers (BLSPs), a strategic solution to promote integrated BIM application across project lifecycles. Unlike traditional partner selection frameworks, which fail to account for BLSPs’ unique characteristics of broad service scope and flexible content, our framework fills this void through rigorous mixed-method research.

The core originality of this study lies in three dimensions. First, it identifies 22 critical criteria specific to BLSPs’ selection, distinguishing it from conventional partner evaluation by emphasizing lifecycle-oriented capabilities (e.g., “Boundary spanning competence of the BIM manager” and “BIM service plan with lifecycle cognition”), innovation capacity (e.g., “Past innovative BIM service practices” and “BIM R&D”), and digital reliability (e.g., “Privacy and security” and “Backup system”). These criteria reflect the distinctive demands of coordinating cross-stage BIM integration and adapting to dynamic technological advancements. Second, through factor analysis, the study consolidates these criteria into five latent factors (Section 5), providing a structured and parsimonious framework that balances practical applicability with theoretical rigor. Third, validation across eight diverse construction projects confirms the framework’s reliability (75% consistency with real-world selections), demonstrating its value in guiding evidence-based BLSP selection. From the findings, we derive that successful BLSP selection must prioritize not only technical capabilities but also organizational stability, historical performance, and adaptability to digital innovation. This shift from fragmented, stage-specific evaluation to holistic lifecycle assessment marks a critical advancement in aligning BIM service provider selection with the industry’s digital transformation goals. Notably, the framework’s ability to capture both operational needs (e.g., service compatibility) and strategic imperatives (e.g., innovation) addresses the limitations of existing tools, which often overlook BLSPs’ unique role as integrators of BIM across project phases.

Despite these contributions, the study has limitations. First, the framework is rooted in China’s institutional context, where top-down BIM policies and specific market dynamics (e.g., state-owned enterprise preferences in critical infrastructure) shape selection practices. This setting both represents an inspiration for other countries (regions with different regulatory or cultural contexts, highlighting the need for context-aware adaptations) and can be considered as Chinese-specific analysis eventually reflecting similarities. A notable constraint is the current scarcity of research on BLSPs across global contexts, as existing studies on BIM service providers remain fragmented and geographically limited. This lack of comparative data on BLSPs’ roles, service models, and selection criteria in different countries (e.g., variations in regulatory support, market maturity, or industry needs) prevents direct cross-national comparisons of BIM practices related to lifecycle service provision. Second, while eight validation projects cover diverse types and scales, further testing in larger samples or cross-national settings is required to enhance robustness. Third, the framework currently excludes strategic factors like geopolitical considerations, which emerged in validation as influential in critical infrastructure projects, suggesting opportunities for refinement.

In summary, this study advances the field by providing the first validated, criterion-based framework for BLSPs’ selection, bridging theory and practice in construction digital transformation. By emphasizing lifecycle integration, innovation, and digital reliability, it equips stakeholders with a tool to select BLSPs that can truly drive BIM’s transformative potential, ultimately fostering more efficient, collaborative, and technologically advanced construction practices. Future research should focus on three directions: (1) cross-national comparative studies to adapt the framework to diverse institutional contexts, exploring how factors like regulatory regimes or market maturity influence BLSPs’ selection criteria; (2) integration of dynamic assessment mechanisms to account for evolving technologies (e.g., AI-BIM integration) and long-term service performance, moving beyond static pre-selection evaluation; and (3) exploration of strategic factors (e.g., government-led digital transformation initiatives, geopolitical security, sustainability alignment) to enhance the framework’s comprehensiveness in complex project scenarios.