Understanding the factors that contribute to project failure is crucial for enhancing project management strategies. This study investigates the relationship between project characteristics such as methodology selection, team composition, and risk management and project outcomes. A structured online questionnaire was created to collect data, offering broad accessibility and facilitating statistical analysis. The survey, currently in progress, has already yielded initial 50 responses, though the dataset continues to evolve. To ensure methodological rigor, the questionnaire adheres to best practices, including clear and concise questions, a mix of closed and open-ended responses, and a logical structure. The ongoing study aims to identify patterns and correlations that can improve project resilience and success rates. This study underscores the importance of a holistic approach to project management, where multiple factors methodology, team dynamics, stakeholder engagement, and proactive risk management must be carefully balanced to improve project outcomes. As data collection continues, further refinements will enhance these insights, providing more precise recommendations for optimizing project management strategies.
Project management has long been a field of extensive research due to its fundamental role in ensuring successful project delivery across industries. Despite the evolution of methodologies and best practices, a significant proportion of projects still fail either by exceeding budgets and timelines or by not reaching completion. Understanding the underlying factors contributing to project failure is crucial for refining management strategies, optimizing resource allocation, and improving overall project success rates.
One of the key elements influencing project outcomes is the choice of project management gain valuable insights into improving project resilience. Statistical analysis of past project data allows researchers and practitioners to identify trends, optimize decision-making, and develop frameworks that enhance project delivery success. As industries continue to evolve, leveraging data-driven approaches to understand failure patterns remains imperative for fostering more robust and efficient project management practices.
Most of the articles [
To systematically analyze the relationship between project failure and project characteristics, a questionnaire-based approach has been employed. Questionnaires are a widely recognized research tool in project management studies, allowing for structured data collection across diverse project environments. There are several types of questionnaires, including structured, semi-structured, and unstructured forms. Structured questionnaires use predefined questions with fixed response options, facilitating quantitative analysis, while semi-structured and unstructured questionnaires allow for open-ended responses that provide deeper insights.
For this study, an online structured questionnaire was chosen as the primary data collection method. This approach offers several advantages, including broader reach due to time saving answers, ease of participation, and real-time data collection. Online questionnaires reduce geographical limitations, enabling respondents from various industries and regions to contribute, which enhances the diversity and reliability of the dataset. Additionally, digital data collection minimizes manual entry errors and facilitates statistical analysis. To ensure the reliability and validity of the collected data, best practices in questionnaire design have been implemented. These include: • • • • • •
Clear and easy understandable questions: Questions are formulated to be direct and easy to understand, reducing the risk of misinterpretation by respondents.
Balanced use of question types: A mix of closed-ended and open-ended questions allows for both statistical analysis and qualitative insights.
Logical structure of the questions: The questionnaire follows a structured format, guiding respondents from general questions to more specific aspects of project management, ensuring logical progression..
Limited response options per question: To avoid overwhelming respondents, multiplechoice questions contain a reasonable number of options while ensuring comprehensive coverage of possible answers.
Pretesting and refinement: The questionnaire was pretested with a sample group before full deployment to identify contradictory and improve clarity.
Not time consuming: Amount of question limited that allow participant to complete them in 5-10 minutes without obligation of providing long written explanation on any of the questions. •
Confidentiality: Respondents' privacy is maintained to encourage honest and accurate responses. Questionnaire does not store or ask for sex, position, country of origin and even doesn’t store mail address.
Main and core auditory for this survey are project managers. The survey is currently active, and initial 50 responses have been gathered. However, data collection remains ongoing, and the dataset is expected to evolve as more responses are recorded. Composed questions from survey [15] can be seen in separated tables below.
All team members where expected seniority
(if previous answer was "Yes") Client; Performance; Person related; Budget related;
Other
What was the final project outcome
start or before start of the past project there were some risks that were converted into issues
1; 0
With information from this survey, after normalization, different parameters can be gathered. Formulas and calculation for parameters such as overall success ratio, percentage of failed project due to methodology, the most remediated and not remediated risk and many more different characteristics are listed below.
Project failure rate has the following formula:
Pf =
N
f × 1 00
N Ps= ( N − N )
f × 1 00
N Si=
N
t × 1 00
N Ra= N p ositive × 1 00
N T eams where Pf is Project failure rate, N f is number of failed and N ❑is number of Projects. By “failed” meant projects that have answer “Delivered but exceeded budget or timeline” and “Not delivered or failed” for question “What was the final project outcome”.
With given set of data we have failure rate 66% Project success rate has similar formula:
Expected that success rate will be around 60-70% of overall quantity of the projects. But with given set of the data our success rate is only 34%
In [
Where Si is stakeholder involvement and N i is number of interaction quantities and they should be combined in this way: for high interactions answer “Daily” should be used, for low is “Monthly” and for medium is sum of the rest answers for the “Stakeholder involvement” question.
As result High interaction is 28%, Medium interaction 50% and Low interaction is 22%. Those results falls into common sense that mean that medium iteration should be biggest one.
For resource availability formula will be more complex because it gathers information out of three questions from questionnaire:
Where Ra is resource availability, N positive is number of positive (“Yes”) responses for questions “Team was fully staffed on time and budget”, “There were cases of team member retention” and (1) (2) (3) (4) “All team members where expected seniority” in Teams type. N Teams is quantity of answers for those three questions, typically should be three times more than N .
Result or resource availability with given data is 51,3% and it’s mean that in half case all parameters of Team composition were correct and there were no issues with it.
Next formula is for identifying methodology that was initially used in given (3):
M i=
N i × 1 00
N
Where M i is each methodology type from the question “Which PM methodology was used” and N i is number of answers for each methodology type.
Quite interesting will look metrics about correlation between methodology that was initially used and used after the change, but this not direct scope of this article. Results shown in Table 6. (5) (6) (7) Methodology
Scrum Kanban Waterfall
Hybrid No methodology
Other
Percentage 20 18 24 18 6 14 where N C h ange is number of answers “Methodology not suitable” and “Client insist” in question “Why was the methodology changed” and N Met h odology is overall quantity of answers on this question.
Next and the most important characteristic that needs to be digested is correlation between project failure and wrong methodology. To calculate it Pearson correlation coefficient will be used: where r is Pearson correlation coefficient, P−f¿¿ and M −f¿¿ are means of corresponding items and M f is wrong methodology indicator that calculates with next formula: r =∑ ( Pf −
P−¿)( Mf−M−f¿)
f √∑ ( Pf − P¿f × √∑ ( M f − M ¿f ¿ ¿
¿ ¿ M f =
Result based on the data available is 48% and it means that in most of half of the cases methodology was changed during the project.
In result this Pearson correlation coefficient should show us actual correlation between methodology and project failure and with value more that zero direction will be positive that will mean that have correlation between those characteristics exists.
For methodology correlation coefficient is weak but positive 0,092 and it means that there exists correlation between methodology and project success.
As for the risks there two simple formulas for most remediated and not remediated risk: Where Ri is each remediated risk from question “Most common risks that were mitigated”, Rri is number of answers for each remediated risk and N Rt is number of overall answers in the question. Similar for not remediated:
Ri= NR Rrri × 1 cs 00 Rni=
Rn ri N R r
Interesting enough that Team retention risk sits on first place in both remediated and not remediated risks
Another crucial formula is formula for correlation between risk monitoring frequency and project failure. This formula can show if risk monitoring is important for project failure or not. To calculate this value another variant of Pearson correlation coefficient will be used: r = n ∑ ( Rm f − Po )−∑ Rmf ∑ Po √[ n ∑ Rm f 2−(∑ Rmf ) ] ¿ ¿ ¿ 2 where Rmf is risk monitoring frequency, Po is project outcome and n is number of responses.
Where Rni is each not remediated risk from question “Most common risks that were converted into issues”, Rnri is number of answers for each not remediated risk. Result is shown in Table 7. (9) (10) Risk name
Scope creep Budget overrun
Team retention Unclear requirements
Other 16 18 24 18 24 12 24 26 18 20
Since desired outcome need to have more value and for that reason Fail value have higher numeric value than Success. Results on given data show 0,1475 and it means that there no visible correlation between project success and risk monitoring frequency.
Next value that needs to be assessed is correlation between team composition and project failure. To calculate this value firstly needs to be assessed team composition values and calculate them with given formula:
T s=S + R +∑ C + E where T s is team composition score, S is staffing status from the question in Table 1, R is retention status, ∑ C is sum of retention cases and E is expected seniority value. In that case max value of T s should not exceed 5 for each case.
Since team composition value logic calculated overall Pearson coefficient for correlation between team composition and project failure can be calculated with next formula: where T s team composition score, Po project outcome and n number of responses.
Results here are more expected, correlation is 0,2421 and it means that there correlation between Team composition and project success.
Next parameter is correlation between project complexity and project failure.: r = n ∑ ( T s− Po )−∑ T s ∑ Po √[ n ∑ T s2−(∑ T s) ] ¿ ¿ ¿
2 r = n ∑ ( Pc− Po )−∑ Pc ∑ Po √[ n ∑ Pc2−(∑ Pc ) ] ¿ ¿ ¿ 2 (11) (12) (13) (14) where Pc project complexity, Po project outcome and n number of responses.
Result is same as in risk monitoring parameter, correlation coefficient is negative, -0,006 and it meant that there not visible correlation between project complexity and project success rate.
This formula is correlation between project success and project industry:
I si= I i
P f i where I si is success rate in specific industry and I i quantity of specific industry in questionnaire answers. Results are shown in Table 8.
Transportation Food and beverages
Logistics E-Commerce
Edtech Government
Oil&Gas Manufacturing
Other 40 0 14 60 20 50 66 0
As the foundational dataset for this research is still undergoing refinement, it must be acknowledged that the current responses collected from various industries may not fully reflect the final or most accurate data. This limitation arises from the preliminary nature of the dataset and the fact that certain industry sectors lack sufficient. In particular, some industry types reported zero or near-zero success percentages, which may not indicate an actual absence of successful or failed projects but instead highlight limitations in sample size at this stage of the study.
This observation presents a methodological challenge from a statistical standpoint. The presence of sectors with missing or non-numeric success data can distort the overall interpretation of industry-level performance trends. As a result, caution must be exercised when drawing generalized conclusions across industry categories, as some of the observed inconsistencies may be attributable to incomplete data rather than to underlying differences in project execution or methodology selection.
Despite these constraints, the article presents a comprehensive analysis of project characteristics and their potential correlations with outcomes. These analyses were performed using structured data collected through questionnaires. The summarized findings of this analysis, including identified correlations and their respective significance levels, are presented in Table 9 below. This table encapsulates the core empirical insights gained through the current phase of research and serves as a foundation for further refinement and validation as the dataset evolves.
The analysis of project failure rates and their correlation with various project characteristics has yielded several key insights. The study found a project failure rate of 66%, with only 34% of projects classified as successful. Among the factors investigated, methodology selection, team composition, and risk monitoring practices emerged as critical influences on project outcomes.
The data revealed a weak but still existing correlation (0.092) between project failure and incorrect methodology selection, with 48% of respondents indicating that the chosen methodology was later deemed unsuitable. This suggests that while methodology plays a role in project success, other factors may have a stronger impact. Similarly, team composition showed a weak correlation (0.2421) with project failure, reinforcing the idea that staffing issues contribute to project challenges but may not be the sole determining factor.
On the other hand, risk monitoring frequency showed no significant correlation (-0.1475) with project failure, implying that merely tracking risks does not necessarily prevent failure unless appropriate mitigation actions are taken. Additionally, project complexity also showed no correlation (-0.006) with failure, suggesting that project outcomes are not solely dictated by complexity but rather by how well they are managed.
Stakeholder involvement also played a notable role, with high interaction levels (28%) correlating with better project success rates, while low interaction (22%) tended to be associated with poorer outcomes. The study also identified key risks that were either mitigated or not, with budget overruns (24%) and team retention issues (26%) being the most common risks that led to project difficulties.
Overall, these findings highlight the need for a holistic approach to project management, where methodology selection, risk mitigation, stakeholder engagement, and team composition are all carefully considered to improve success rates. As data collection continues, further analysis may refine these insights and provide more definitive conclusions on how best to enhance project management practices for better outcomes.
The results of this study have the potential to be integrated into a specialized mobile application tailored for use within specific industries. This application could serve as a valuable tool not only for internal departments, such as a Project Management Office (PMO), but also for external stakeholders including consulting firms, partner organizations, and companies that are involved in auditing processes. By embedding the study's findings into the app, it becomes a centralized platform for informed decision-making and project planning.
In addition to suggesting appropriate methodologies, the application would function as a proactive advisory system for project managers. It would offer practical guidance and predictive alerts about potential challenges or obstacles that may arise during different project phases. This feature aims to empower project leaders with foresight and preparedness, ultimately contributing to higher project success rates and more efficient resource allocation.
Ihor Berezutskyi would like to express his deepest gratitude to his wife, Hanna Zaviriukha, whose support, advices around questionnaire composition helped to craft this work. Also worth mention that Hanna was one of the first participants in survey and helped to identify several different mistakes in wording and in logic.
Also Ihor Berezutskyi would like to thank her daughter Liza for bringing joy and necessary distraction that allow to systematically structure article and adjust formatting.
In addition, Ihor Berezutsky would like to thank his supervisor, professor Tetyana Honcharenko, for her technical expertise and guidance around work structure.
Lastly, Ihor Berezutskyi wishes to express his sincere gratitude to his project manager colleagues, whose participation in the survey and willingness to share their experiences provided invaluable data for this research. Their support and engagement have been essential in making this study possible.
During the preparation of this work, the authors used ChatGPT in order to: Paraphrase and reword. After using this tool/service, the authors reviewed and edited the content as needed and takes full responsibility for the publication’s content. IEEE 4th International Conference on Smart Information Systems and Technologies (SIST), May 2024, URL: DOI: 10.1109/SIST61555.2024.10629587 [3] I. Berezutskyi, T. Honcharenko, Potential implications of artificial intelligence for project management information systems, 2024. URL: https://ceur-ws.org/Vol-3709/paper3.pdf [4] D. Chernyshev, G. Ryzhakova, H. Petrenko, I. Chupryna, N. Reznik, Digital Administration of the Project Based on the Concept of Smart Construction, Lecture Notes in Networks and Systems, vol.495 LNNS, pp. 1316–1331, 2023, International Conference on Business and Technology, ICBT 2021, DOI: 10.1007/978-3-031-08954-1_114 [5] Mihaylenko, V., Borodavka, Y., Dolya, E., Savenko, V. Information tools for project management of the building territory at the stage of urban planning, CEUR Workshop Proceedings, 2021, 2851, рр. 22–33, http://ceur-ws.org/Vol-2851/paper3.pdf [6] S. Bushuyev, K. Pilyugina,I. Achkasov, A. Zaprivoda, Clip Thinking in the Digital Age: Complimentary or Contradictory, 2023 IEEE 18th International Conference on Computer Science and Information Technologies (CSIT), 2023. DOI: 10.1016/j.procs.2023.12.210 [7] Khan, A., Hasana, M.K., Ghazal, T.M., Islam, S., Alzoubi, H.M., Mokhtar, U.A., Alam, R., Ahmad, M., 2022. Collaborative Learning Assessment via Information and Communication Technology, in: Proceedings, RIVF 2022., pp. 311-316 [8] S. Bushuyev, N. Bushuyeva, D. Bushuiev, V. Bushuieva, Integrated Intelligence Model for Assessment Digital Transformation Project, 2023 IEEE International Conference on Smart Information Systems and Technologies (SIST), 2023. DOI:10.1109/SIST58284.2023.10223578 [9] M. El Khatib, A. Al Falasi, Effects of Artificial Intelligence on Decision Making in Project
Management, American Journal of Industrial and Business Management 11, 2021, pp. 251-260. [10] Harvard Business Review Analytic Services, The Evolution of Decision Making: How Leading Organizations Are Adopting a Data-Driven Culture. URL: https://hbr.org/resources/pdfs/tools/17568_HBR_SAS%20Report_webview.pdf [11] R. Nicoara, Project Management Methods, 2023, DOI: 10.1007/978-1-4842-9663-9_12 [12] R. L. Moura, T. C. J. Carneiro, T. L. Dias, VUCA environment on project success: The effect of project management methods, Brazilian Business Review 20(3), 2023, DOI: 10.15728/bbr.2023.20.3.1.en [13] M. Nekrasova, Using methods of probability theory and mathematical statistics in project management, Bulletin of the National Technical University «KhPI» Series Dynamics and Strength of Machines, 2023, DOI: 10.20998/2078-9130.2023.2.293040 [14] S. Turney Pearson Correlation Coefficient (r). Guide & Examples, 2024 https://www.scribbr.com/statistics/pearson-correlation-coefficient/ [15] I. Berezutskyi, Method for analyzing project characteristics, Management of Development of
Complex Systems, Kyiv, 2025. DOI: 10.32347/2412-9933.2025.61.145-150 [16] Aliyev, Ali. The AI and Quantum Era: Transforming Project Management Practices. Journal of
Future Artificial Intelligence and Technologies 1.4 (2025): 348-370 [17] G. Ryzhakova, T. Honcharenko, K. Predun, N. Petrukha, O. Malykhina, and O. Khomenko.
Using of Fuzzy Logic for Risk Assessment of Construction Enterprise Management System. 2023 IEEE International Conference on Smart Information Systems and Technologies (SIST), Astana, Kazakhstan, 2023, pp. 208-213, doi:10.1109/SIST58284.2023.10223560 [18] O. Ilin, S. Bushuyev, V. Gots, T. Liashchenko, Management of IT projects of business analytics in the BANI environment, Management of Development of Complex Systems, 2024, DOI: 10.32347/2412-9933.2024.59.45-52.