Comparison of the Cost-Effectiveness of Manual and Automated Testing in IT Projects of Small and Medium-Sized Businesses

Наукові записки Львівського університету бізнесу та права. Серія економічна. Серія юридична. Випуск 42/2024 505 https://doi.org/10.5281/zenodo.17356041 Deviatko А. QA Engineer, PGA TOUR, https://orcid.org/0009-0008-8424-7188 COMPARISON OF THE COST-EFFECTIVENESS OF MANUAL AND AUTOMATED TESTING IN IT PROJECTS OF SMALL AND MEDIUMSIZED BUSINESSES JEL Classification: D24, L86, O33, M21 SECTION “ECONOMICS”: Економіка Анотація. Дослідження присвячене порівняльному аналізу економічної ефективності ручного та автоматизованого тестування в ІТ-проєктах малого й середнього бізнесу. Методологічну основу становить аналіз офіційної статистики цифрової зрілості МСБ – рівнів впровадження робототехніки, штучного інтелекту (ШІ), ERP та CRMсистем – які розглядаються як проксі-показники готовності до інтеграції автотестів і швидкої окупності інвестицій у цифровізацію. Отримані дані систематизовано у табличному форматі з подальшим застосуванням порівняльного контент-аналізу наукових і практичних джерел. На основі синтезу результатів сформовано профіль витрат і вигід, який включає порівняння початкових і поточних витрат, часових характеристик, масштабованості, ризиків та організаційних обмежень для двох підходів. Установлено, що початкові витрати ручного тестування є відносно низькими й здебільшого формуються за рахунок заробітної плати тестувальників, тоді як автоматизація вимагає початкових інвестицій у ліцензії, інфраструктуру, навчання персоналу та підтримку середовища. Для прикладу, орієнтовна вартість ліцензії UFT One становить близько $3000, що відображає вагому частку у стартовому бюджеті автоматизованого підходу. Водночас результати показують, що автоматизація забезпечує прискорення тестування у 10–100 разів, зниження трудомісткості процесів на 70–75 %, а також підвищення стабільності результатів при повторних перевірках. Однак її ефективність знижується у випадку коротких життєвих циклів проєктів або нестандартних завдань, оскільки щорічна підтримка та оновлення скриптів може сягати до 50 % початкового бюджету. Проведений аналіз свідчить, що економічна доцільність автоматизованого тестування зростає пропорційно частоті релізів і стабільності функціональних сценаріїв, тоді як у короткострокових або одиничних проєктах доцільніше використовувати ручне тестування як більш гнучке та менш капіталомістке рішення. Узагальнені результати дозволяють визначити оптимальні умови переходу від ручних до автоматизованих практик у МСБ та формують основу для подальших економічних розрахунків рентабельності інвестицій у цифрові інструменти контролю якості. Ключові слова: програмне забезпечення, термін окупності інвестицій, тестувальник, ліцензія, штучний інтелект, капітальні витрати, інженер. Annotation. The study analyses official statistics of digital maturity of small and mediumsized enterprises (SMEs) (robotics, AI, ERP/CRM) as proxy indicators of readiness for the implementation of automated tests and rapid payback. The data are further arranged and Наукові записки Львівського університету бізнесу та права. Серія економічна. Серія юридична. Випуск 42/2024 506 tabulated. Comparative content analysis of literature and practical data synthesized in a cost/benefit profile is carried out, which compares initial and ongoing costs, time characteristics, scalability, risks and limitations for manual and automated approaches. It was determined that the initial costs of manual testing are minimal (mainly the salary of a specialist tester), while automation includes licenses, infrastructure, and training. The profile provides an example of UFT One from $ 3,000 as a guideline for the license component. It was found that automation provides 10-100x faster testing and reduces labour costs by 75%, but requires annual script support up to 50% of the budget and is tool-dependent, while manual testing requires additional working hours and personnel. It was established that automation is more cost-effective in repetitive scenarios with regular releases, while short projects are more expedient to keep manual. Key words: software, payback period, tester, license, artificial intelligence, capital expenditures, engineer. Introduction Small and medium-sized businesses often face a dilemma when developing and supporting software products: to maintain the manual testing practice or to invest in the automation of this process. With limited financial resources and time, each additional hour of work, as well as each monetary unit spent, acquire a certain weight. Available analytical assessments indicate that testing automation can significantly reduce time consumption, reduce quality assurance costs, and increase the stability and reliability of software releases [1; 2]. The issue of economic feasibility arises in this context: are SMEs’ investment in automated testing systems able to provide a quick payback and a positive profitability index or is it more expedient to maintain the preference for traditional manual approaches given the specifics of business processes. Software testing is one of the fastest growing technology industries, with a market size of $40 milliard in 2021 [2] and an expected compound annual growth rate of 6% from 2022 to 2030. The importance of quality assurance in the software industry is undeniable, as repeatedly demonstrated by examples of promising solutions that ultimately fail because of the lack of proper testing. Recent studies confirm that the cost-effectiveness of testing automation in small and medium-sized IT projects directly depends on the repeatability of scenarios, the stability of requirements, and the maturity of management processes. The implementation of modern project management tools such as Jira, Trello, and Monday creates the infrastructure prerequisites for a faster payback of automated tests, reducing transaction costs for coordination, and improving discipline in task performance [3-5]. Comparative analyses across domains from industrial software to web applications shows that automation significantly outperforms manual testing in terms of regression speed, coverage, and defect risk reduction, but requires high initial investment and script maintenance costs [6-8]. The regression, smoke, and API tests showed the greatest returns, while exploratory and UX testing remain largely manual. The choice of automation tools is determined not so much by the license cost, but by the total cost of ownership of the ecosystem: the costs of integration with CI/CD (CI – continuous integration, and CD – continuous delivery), the stability of the environments, the presence of a community, as well as the team qualifications [9]. An additional factor is the emergence of generative artificial intelligence (GenAI) models capable of automatically generating tests and accelerating code coverage, but their output requires mandatory human validation to prevent false confidence in quality [10]. The general conclusion of the academic literature is that it is optimal for SMEs to implement automation in stages, starting with important and repetitive scenarios, integrating it into managed processes and combining it with manual testing where human flexibility and creativity remain key to ensuring quality. Despite the significant number of academic papers on test automation and IT project management, there still scarce direct systematic comparisons of the cost-effectiveness of manual and automated approaches in the academic community. Most publications focus either on the technical aspects of the tools or on individual cases, while there are few generalized quantitative and qualitative assessments for the context of SMEs. That is why our study aims to close this gap by combining cost-benefit analysis with practical recommendations for choosing the optimal testing strategy depending on the project parameters. Наукові записки Львівського університету бізнесу та права. Серія економічна. Серія юридична. Випуск 42/2024 507 The aim of the study іs to determine the cost-effectiveness of automated testing compared to manual in the context of SMEs’ IT projects, considering the technological maturity of enterprises, cost structure, and expected benefits. Research objectives: 1. Identify the current level of technological maturity of SMEs and identify areas where automated testing can provide the greatest economic impact. 2. Compare the costs and benefits of manual and automated testing, determining the conditions under which each approach is economically feasible. Results Assessing the economic feasibility of implementing automated testing in IT projects of SMEs requires considering not only direct costs and benefits, but also the level of technological readiness of enterprises for automation. In this context, statistical indicators of the use of robotics, artificial intelligence (AI), and corporate information systems (ERP, CRM) are relevant, as they reflect the actual state of digital transformation of business, as well as the degree of integration of automation tools into production and management processes. Enterprises that have already implemented such technologies have a lower barrier to integration of automated test environments. Therefore, the probability of a quick return on investment in these solutions increases. The current level of technological maturity of SMEs and areas where automated testing can provide the greatest economic effect were identified by using official statistics on the share of enterprises that use robotics, AI, and business systems for various purposes and types. Table 1 presents the summarized results. Table 1 Statistics on the use of modern digital technologies in SMEs in Ukraine in the context of the implementation of automated testing Indicators and reasons for using technology in enterprises Small enterprises (from 10 to 49 people) Medium enterprises (from 50 to 249 people) 1. Share of enterprises using robotics in the total number of enterprises by reasons that influenced the decision to use robots Share of enterprises using robotics in the total number of enterprises, %: 3.7 5.0 due to high labour costs 0.6 1.1 due to difficulties in recruiting personnel 0.6 1.0 due to increased safety at work 1.2 1.7 due to ensuring high accuracy or standardized quality of processes and/or manufactured goods and services 0.2 0.3 due to expanding the range of goods produced or services provided 1.6 2.8 due to tax or other government benefits 2.4 2.8 2. Share of the number of enterprises using AI technologies in the total number of enterprises by the purpose of using AI Share of the enterprises using AI technologies in the total number of enterprises, % in 2022: 5.4 5.2 for marketing or sales 2.7 2.9 for production processes for business administration processes 2.1 1.9 for enterprise management 2.3 1.5 for logistics 1.9 1.6 for ICT security 1.2 1.2 Наукові записки Львівського університету бізнесу та права. Серія економічна. Серія юридична. Випуск 42/2024 508 for personnel management or recruitment 0.9 0.9 3. Share of the enterprises using AI technologies in the total number of enterprises by type of AI technologies in 2022: technologies for analysing written language 3.1 2.9 technologies for converting spoken language into a machine-readable format 1.2 0.6 technologies for generating written or spoken language 0.6 0.3 technologies for identifying objects or people based on images 1.1 0.9 machine learning for data analysis 1.2 1.5 technologies that automate various workflows or assist in decisionmaking 2.2 1.5 technologies that provide physical movement of machines using autonomous solutions based on environmental observation 0.7 0.4 4. Share of the enterprises using ERP and CRM software in the total number of enterprises in 2022: Share of enterprises using enterprise resource planning (ERP) software in the total number of enterprises, % 4.7 8.1 Share of enterprises using customer relationship management (CRM) software in the total number of enterprises, % 2.3 4.4 managing the collection, storage, and provision of customer information to various business functions 2.0 3.7 managing the analysis of customer information for marketing purposes 1.9 3.6 Source: developed by the author based on [11] Analysis of the above statistics shows that the level of implementation of high-tech solutions in the segment of SMEs in Ukraine remains fragmented and uneven. In particular, the use of robotics is noted by only 3.7% of small and 5.0% of medium-sized enterprises. The main drivers of its implementation are the expansion of the product range (1.6% and 2.8%, respectively) and obtaining tax or other state benefits (2.4% and 2.8%). The low share of enterprises implementing robotic solutions due to the need to increase accuracy and standardize processes (0.2–0.3%) indirectly indicates a limited level of automation of control and test operations, which is important for the issue under research. In the segment of AI technologies, the share of users among SMEs is about 5%, with the application for marketing tasks (2.7–2.9%) and business process management (1.5–2.3%) prevailing. The use of AI in ICT security (1.2%) and for personnel selection (0.9%) demonstrates potential, but so far minor attempts to automate routine processes, which can include testing. The structure of AI technologies application by type shows that the most widespread tools are text data analysis tools (3.1% and 2.9%) and workflow automation tools (2.2% and 1.5%). This directly correlates with the trend towards the use of AI-based platforms for automated software testing, which operate with large volumes of logs, reports, and specifications. Regarding corporate IT systems, ERP solutions are implemented in 4.7% of small and 8.1% of medium-sized enterprises, CRM systems – in 2.3% and 4.4%, respectively. Although this indicates limited penetration of integrated management systems, such platforms create a basis for integrating automated testing tools directly into the software development and implementation lifecycle. So, the presented data give grounds to draw a preliminary conclusion: in most SMEs, the level of technological readiness for the transition from manual to automated testing is low, but the existing local implementations of AI and ERP/CRM solutions create potential growth points. This means that comparing the cost-effectiveness of the two approaches in our study should consider not only direct costs and benefits, but also the infrastructural readiness of enterprises for testing automation. The next step in our study will be to compare the costs and benefits of manual and automated testing. Manual testing is a process in which a QA engineer (tester) performs tests manually without the use of automation tools. It is a human-cantered approach that offers a perspective that no machine can fully replicate [12]. This type of software testing has existed since the early days of software development, long Наукові записки Львівського університету бізнесу та права. Серія економічна. Серія юридична. Випуск 42/2024 509 before automation tools were available. Initially, it was all about testing basic functionality – making sure that the software worked as expected. QA engineers manually went through each function, pressing every button and checking for errors. However, manual testing has evolved but remains an important part of the quality assurance process. Now, it covers complex areas such as usability, exploratory testing, and real-world user behaviour, tasks that automation still struggles with. At the start, manual testing looks attractive due to the minimal initial costs. It does not require expensive tools or complex infrastructure; it is enough to hire a specialist tester or use an existing specialist. Automated testing, on the contrary, requires significant capital investments at the beginning: paying for test platform licenses, setting up environments, training personnel or hiring an Automation QA engineer. According to analysts, commercial tools can be extremely expensive. The costs of specialists are added: the average salary of a test automation engineer in Ukraine is about $ 2,200 per month, excluding the cost of training [13]. So, the initial threshold for entry into automation is noticeably higher, which can become a barrier for small businesses [2]. In contrast, manual testing does not require special tools – the costs are mainly reduced to the testers’ working time. This factor explains why, manual testing is often more cost-effective in the short term for small projects with modest budgets [14]. Automated testing is the use of scripts and tools to automatically run tests. Instead of testers manually executing tests, testers write scripts that ask a machine to run tests for them. Commercially available tools such as Selenium, Appium, and JMeter are quite popular in automated testing [14]. However, ongoing costs and scalability should be considered. Manual testing scales linearly: more features mean more hours of work, and therefore more costs. As the product evolves, each regression testing cycle will become more expensive over time, as manual tests need to be repeated. For small businesses, this means that expanding functionality directly increases testing costs by hiring more testers or paying overtime. Automated tests, on the other hand, scale much more efficiently once implemented: a single engineer can run thousands of scripts, and adding new scripts is often cheaper than the equivalent number of hours of manual work. However, there are some obstacles here too – the cost of supporting automated tests. Scripts need to be updated when the application code changes, and this can consume up to 50% of the automation budget each year. The World Quality Report 2022–2023 notes that 30–50% of a testing team’s resources can be spent on supporting test scripts [15]. So, for long-term projects, it is necessary to consider not only the initial investment, but also the ongoing costs of updating automated tests (engineer time, fixing erroneous tests, adapting to new functionality, etc.). There are no support costs as such in the case of manual testing. However, the human factor (vacations, errors due to overwork) can lead to hidden “costs” in the form of missed defects. Despite the mentioned costs, automation, when used skilfully, can provide significant benefits that more than pay for the investment. The main one is saving time and labour. Automated tests work much faster than people: where a manual tester clicks through dozens of scenarios all day, an automated script will run them in a matter of minutes. In other words, if a company spent a conditional $100 thousand per year on manual testing, the implementation of automated tests can save about half of this amount (approximately $50 thousand). Some modern approaches automate so much routine that the amount of manual labour is reduced by 80% or more [16]. For example, Testvox analysts noted in 2024: the implementation of regression automated tests at the client reduced labour costs by 75%, and the release of updates was accelerated by an average of two weeks [16]. Table 2 shows the comparative economic profile of manual and automated testing for SMEs. Table 2 Comparative economic profile of manual and automated testing for SMEs Item No. Criterion Manual testing Automated testing 1 Initial costs Minimum – most importantly – tester salary: Junior QA: $475–$510/month (Ukraine); Middle QA: $1,800–$2,500/month; Senior High – purchase of licenses, infrastructure setup, training: UFT One license – from $3,000 [17] Наукові записки Львівського університету бізнесу та права. Серія економічна. Серія юридична. Випуск 42/2024 510 QA: ~$3,400–$4,100/month [13] 2 Ongoing costs Directly proportional to the amount of work: more tests = more hours, additional rates/overtime Autotest support: up to 50% of the automation budget each year (script updates, adaptation to new functionality) 3 Speed of execution Slower: one regression cycle takes days or weeks depending on the amount of test cases 10–100 times faster: autoscripts execute thousands of tests per minute; regression can be nightly and automatic 4 Accuracy There is a risk of human errors (inattention, fatigue) High stability of script execution; risk of false results due to “broken” scripts 5 Scalability Linear: to increase the volume of tests, a proportional increase in staff is required High: one automation engineer can support thousands of test cases 6 Time to market Longer: increases with the expansion of functionality Time-to-market reduction by an average of 30% [2] 7 Return on Investment (ROI) Does not require payback – payment only for the time of work ROI is often positive in 6–12 months: example: +150% in half a year when replacing regression with automated tests [18] 8 Payback Period No capital costs, therefore, no payback period Short-term projects: payback may not come Long-term with regular releases: payback in 6–18 months [18] 9 Suitable Scenarios 1. One-off projects 2. Tests that require creativity or constant changes 1. Repeated regression tests 2. Large and long-term projects 10 Additional Benefits Flexibility, the ability to improvise during testing 1. Frees testers from routine 2. Increases test coverage by approximately 25% 3. Reduces manual work by 50–85% [16] 11 Limitations Limited speed, high cost of scaling, fatigue High entry cost, requires ongoing support, not suitable for frequently changing tests 12 Risks Human factor, missing defects False positive/false negative results due to script failures, tool dependency Source: arranged by the author based on [2; 13; 16-18] To illustrate the economic feasibility of automation in SMEs, let us consider a hypothetical IT project with the following parameters: • Manual testing scenario: • 2 manual QA engineers at $1,800/month each → $43,200 per year. • Regression cycle takes ~5 days per release. With 12 releases/year, cumulative testing time ≈ 60 days. • Automated testing scenario: • Initial costs: • Tool license (UFT One) − $3,000 • Infrastructure setup − $5,000 • Training / onboarding − $4,000 • Total initial investment: $12,000 • Ongoing support costs: ~30% of automation budget = $12,000/year. Наукові записки Львівського університету бізнесу та права. Серія економічна. Серія юридична. Випуск 42/2024 511 • 1 Automation QA engineer at $2,200/month → $26,400 per year. • Regression cycle reduced from 5 days to 1 day, saving 4 days × 12 releases = 48 days of engineer time annually (~$18,000 in saved labor). Comparison: • Manual: $43,200 annual cost • Automated: $12,000 (initial, one-time) + $26,400 (engineer) + $12,000 (support) − $18,000 (savings) = $32,400 first year • From Year 2 onward, with no initial setup: $20,400 less than manual each year. ROI formula: ROI = \frac{(Savings - Costs)}{Costs} \times 100\% For this case: • First-year ROI = \frac{(43,200 - 32,400)}{32,400} \times 100\% = +33\% • Payback period = ~9 months. • From Year 2 onward, cumulative ROI increases to +95% annually. This simplified model confirms that automation becomes economically more advantageous starting in the first year of use, with accelerated benefits in subsequent years. Analysis of open parameters showed a clear pattern: manual testing is more affordable at the start, especially when it comes to SMEs, but loses efficiency with increasing volume and frequency of releases. The initial costs of automation are significant (licenses, infrastructure, training), but with a stable release cycle, the payback can occur within 6–12 months. The ROI can reach +150% in half a year in the case of replacing regressions with automated tests. Speed and scalability favour automation for long-term projects: time-to-market is reduced by an average of 30%, test coverage increases by approximately 25%, and the share of manual labour is reduced by 50–85%. On the other hand, the manual approach retains its value for short or unique tasks, where the tester’s flexibility and creative improvisation are more important than speed. There are hidden operational costs behind the official costs of licenses (from $3,000 for UFT One) and infrastructure. Supporting automated tests can consume up to 50% of the annual automation budget, especially with frequent changes to the code base. This means that real payback is possible only with a stable product or a well-planned update process. For manual testing, a hidden cost factor is the risk of missing defects due to the human factor, which can cost the company more than maintaining an automated system in critical cases. It is also not always considered that automation requires not only engineers, but also integration with CI/CD, reporting tools and test environments. This is an additional budget that can become a barrier for SMEs. On the other hand, in manual testing, the cost of losing time to scale the team as the volume of tasks increases, which also reduces cost efficiency. Conclusions Thus, the economic feasibility of automation for SMEs depends not only on direct cost and benefit figures, but also on the management of hidden components – support, process changes, integrations, quality risks. Manual testing is optimal for short, one-time and non-standard tasks, and automation – for stable, repeatable and long-term processes, where each release is significant for the business. The most balanced for SMEs is a hybrid approach, when important regression tests are automated, and unique scenarios remain under human control, which enables controlling costs and maximizing the benefits of both methods. The study showed that the cost-effectiveness of automated tests for SMEs is not a universal constant: it is realized in the presence of repeatability, rhythmic delivery, and basic process maturity. The manual approach, however, remains rational for short or non-standard initiatives. The key is the phased implementation and management of hidden support costs, which turns the speed and scalability of automated tests into real financial benefits. The optimal solution in most cases remains a balanced, hybrid strategy that combines the strengths of both approaches and meets the resource constraints of SMEs. References 1. Mishra, S., Ewing, M. T., & Cooper, H. B. (2022). Artificial intelligence focus and firm performance. Journal of the Academy of Marketing Science, 50(6), 1176-1197. https://doi.org/10.1007/s11747-022-00876-5 Наукові записки Львівського університету бізнесу та права. Серія економічна. Серія юридична. Випуск 42/2024 512 2. Heinemann, G. (2022). Business models in B2B eCommerce. B2B eCommerce: basics, business models and best practices in business-to-business online trade Wiesbaden: Springer Fachmedien Wiesbaden. 91-162. https://doi.org/10.1007/978-3-658-38528-6_3 3. Trushkina, N. (2023). Upravlinnia proiektnymy ryzykamy u systemi ryzyk-menedzhmentu IT-kompanii: teoretychni aspekty [in Ukrainian]. Ekonomika ta suspilstvo, (55). https://doi.org/10.32782/2524-0072/2023-55-88 4. Khrapkin, O., Kindrat, O., & Chopei, R. (2023). Upravlinnia proiektamy v IT-haluzі: metodyky, instrumenty ta keruvannia ryzykamy [in Ukrainian]. Ekonomika ta suspilstvo, (55). https://doi.org/10.32782/2524-0072/2023-55-110 5. Shashkova, N. I., Fadieieva, I. H., & Kazakova, T. S. (2021). Upravlinnia proiektamy v IT sferi: zastosuvannia hnuchkykh metodolohii [in Ukrainian]. Scientific Notes of Lviv University of Business and Law, 28, 166-172. http://dx.doi.org/10.5281/zenodo.5346802 6. Enoiu, E., Sundmark, D., Čaušević, A., & Pettersson, P. (2017, March 13–17). A comparative study of manual and automated testing for industrial control software. 2017 IEEE International Conference on Software Testing, Verification and Validation (ICST), Tokyo, Japan. IEEE. https://doi.org/10.1109/ICST.2017.44 7. Dobles, I., Martínez, A., & Quesada-López, C. (2019, June 19–22). Comparing the effort and effectiveness of automated and manual tests. 2019 14th Iberian Conference on Information Systems and Technologies (CISTI), Coimbra, Portugal. IEEE. https://doi.org/10.23919/CISTI.2019.8760848 8. Bhanushali, A. (2023). Impact of automation on quality assurance testing: A comparative analysis of manual vs. automated QA processes. International Journal of Advances in Engineering Research, 26(IV), 39-47. https://www.researchgate.net/publication/375342615_Impact_of_Automation_on_Quality_Assurance_Test ing_A_Comparative_Analysis_of_Manual_vs_Automated_QA_Processes 9. Umar, M. A., & Zhanfang, C. (2019). A study of automated software testing: Automation tools and frameworks. International Journal of Computer Science Engineering, 8(6), 199-206. https://doi.org/10.5281/ZENODO.3924794 10. Yuan, Z., Lou, Y., Liu, M., Ding, S., Wang, K., Chen, Y., & Peng, X. (2023). No more manual tests? evaluating and improving chatgpt for unit test generation. arXiv preprint, arXiv:2305.04207. https://doi.org/10.48550/arXiv.2305.04207 11. Bilan, Y., Oliinyk, O., Mishchuk, H., & Skare, M. (2023). Impact of information and communications technology on the development and use of knowledge. Technological Forecasting and Social Change, 191, 122519. https://doi.org/10.1016/j.techfore.2023.122519 12. Testfort. (2024). Automation testing vs. manual testing: Rivals or strategic partners? https://testfort.com/blog/automation-testing-vs-manual-testing-will-automation-replace-manualqa#:~:text=environments%2C%20automation%20can%20become%20outdated,friendly%20option 13. DOU. (2023). Зарплати українських тестувальників – зима 2023 https://dou.ua/lenta/articles/salary-report-qa-winter-2023 14. PixelQA. (2024). Automation testing vs. manual testing: Which approach delivers better ROI? https://www.pixelqa.com/blog/post/automation-testing-vs-manual-testing-betterroi#:~:text=,need%20maintenance%20in%20this%20way 15. Sogeti. (2023). World quality report 2022-23. https://www.sogeti.com/research-andinsight/world-quality-report-2022-23/ 16. Testvox. (2024). The ROI of software testing: A cost-benefit analysis. https://testvox.com/the-roi-of-software-testing-a-cost-benefit-analysis/#:~:text=1 17. Ahmad, I. (2024, May 31). UFT One: A functional testing tool that supports scriptless test creation for web, mobile, API, and desktop applications. CheckOps. https://www.checkops.com/uftone/?utm_source 18. Reine De Reanzi, S., & Ranjit Jeba Thangaiah, P. (2021). A survey on software test automation return on investment, in organizations predominantly from Bengaluru, India. International Journal of Engineering Business Management, 13, 18479790211062044. https://doi.org/10.1177/18479790211062044