Find out how code quality, craft software and IA integration impact developer productivity and software development costs.
In the complex universe of software development, the impact of code quality is often underestimated. However, this is a determining factor for developers’ productivity and development costs. Long remained abstract, the influence of a healthy and well -maintained code can now be evaluated concretely thanks to the first calculator to measure the impact of the Craftware and the AI (IA & Craft Impact Calculator).
Find out how Software Craftware, and now the integration of artificial intelligence (AI) can transform your approach and generate tangible benefits.
Learn more about the AI & Craft Impact Calculator thanks to AROLLA.
Hidden costs due to a poor quality code
A lack of visibility on the quality of the source code frequently leads to favoring short -term gains, to the detriment of long -term maintainability and scalability. Neglecting technical alert signals can cause growing technical debt, where former rapid tasks become laborious, slowing innovation. A poor quality code slows down development and undermines progression (1.).
Research carried out by codescene (2.) indicate that the improvement in the quality of the code could lead to a reduction of 15 times the number of bugs, a doubling of the development speed and an uncertainty nine times less as for the realizations of realization. The commercial advantage of a healthy code is therefore proven. In addition, the study reveals that 23 to 42% of developers’ time is wasted due to technical debt and bad code.
The new AI challenges
The growing adoption of AI in software development brings new dynamics. If the AI can boost individual productivity, the Dora 2024 report (3.) reveals that its overall impact on the performance of software delivery is negative. In particular, there is a decrease in the stability of delivery. In addition, despite a perceived improvement in the quality of the code thanks to AI, this does not necessarily result in an improvement in the overall delivery performance. The report suggests that the increase in the amount of code produced by AI could cause greater changes, which traditionally makes deliveries slower and more prone to instability.
A data-driver approach
Software Craft advocates technical excellence and attention to retail in software development (2.). By adopting its principles and good practices, it is possible to make the quality of the code a key performance indicator (KPI) for the company (1.) Knowledge of the state of your code (Red, Yellow, Green Code; see the Codescene study (1., 2.)) allows an informed decision -making, based on data (1.).
Our approach to an “calculator of the impact of the AI and the Software Craft” (IA & Craft Impact Calculator at Arolla) is based on several pillars.
We start by measuring the “code health” (health code). Automated code analysis tools can assess the quality and complexity of the code. Identify the “RED Code” areas (poor quality code) is crucial, because they are associated with 15 times more flaws than the quality code.
We then assess the lost time caused by the technical debt by estimating the time that the developers spend each week to manage the consequences of the technical debt. Studies show that this considerably harms the productivity of developers (4.).
We also calculate the return on investment (king) of improving the quality of the code: quantifying the time lost and the causes linked to the bad code makes it possible to estimate the potential king of improvement efforts (2.). The management of technical debt can increase the efficiency of the delivery of features by at least 25%, equivalent to a gain in significant development capacity (4.).
Finally, we also use unexpected working time as an indicator. The time dedicated to the correction of bugs and the “urgent” refactoring reveals the cost of the technical debt. Analyzing the percentage of unforeseen work can indicate an unexploited potential linked to the quality of the code. If your organization spends more than 15% of its unforeseen work time, it is an alert signal (1.).
A strategic integration of the AI
Regarding the AI, the Dora 2024 report (3.) highlights contrasting impacts. If the developers bring back a positive impact on individual productivity, flow and work satisfaction, there is a decrease in time devoted to actually valuable work. The AI also seems to improve the quality of the code, the documentation and the speed of code journals, while reducing the complexity of the code.
However, these improvements do not translate, for the moment, by a better performance of the delivery of software; On the contrary, the adoption of AI even seems to deteriorate it. The report suggests that this could be linked to an increase in the size of code changes induced by the speed of generation of AI code, going against the principle of small lots of changes beneficial to stability.
Software craft and AI: towards more productivity and less costs
The adoption of Software Craft and its good development practices has a direct impact on productivity and costs (2.) by allowing in particular: improving the quality of code, reduction of technical debt, acceleration of development cycles, automation of QA tests, the use of high -performance tools, culture of psychological security and continuous learning, integration of AI or the appearance of internal development platforms.
Practices such as the Test-Driven Development (TDD), the Pair Programming and the Code Review contribute to a cleaner code, easier to maintain and less subject to errors. A quality code makes it possible to develop new features twice as quickly and considerably reduces the number of defects. AI can also contribute to improving the quality of the code and reducing its complexity (3.).
In addition, the reduction in technical debt involves favoring quality from design and regularly refactor the code. The teams thus reduce the accumulation of technical debt, avoiding future slowdowns of delivery of new features and high maintenance costs (1.). Priorifying the health of the code is essential when planning projects (1.).
Development cycles can also be accelerated through practices such as continuous integration (CI) and continuous delivery (CD), which automate construction, test and deployment processes, reducing manual deadlines and errors (5.). Efficient organizations using CI/CD have deployment deadlines 440 times faster (5.). AI can be integrated into CI/CD pipelines to improve integration, tests, deployment and surveillance (6.).
Automation of tests also makes it possible to quickly identify defects, reduce test costs and improve the overall quality of the software (7.). The automation of repetitive tasks also releases time for activities with higher added value (5.). The AI can develop test cases and prioritize the code sections to be tested (6.).
Planning, development, collaboration and CD/CD tools can contribute significantly to developer speed and innovation (6.). Companies with high -performance tools are 65% more innovative (10.). AI is increasingly integrated into these tools to offer improved features (6.).
An environment where experimentation and learning are encouraged, and where developers feel safe to take risks, promotes better performance and reduces burnout (8.). Investing in continuing education and providing clear career paths also help to attract and retain talents (8.).
The use of artificial intelligence tools for code writing, documentation and optimization can also improve the productivity and quality of the code. AI seems to improve the quality of the code and reduce its complexity. However, uncontrolled integration will affect the stability of delivery associated with serious safety problems (12.). It is crucial to create a culture of continuous learning and experimentation with AI, while setting up a measurement framework focused on significant impacts rather than simple adoption (3.).
Finally, a well-designed platform can increase the independence and productivity of developers by offering them self-service tools and services (9.). AI can improve internal development platforms by offering prediction and automation capacities (6.).
Invest in Software Craft and an enlightened IA strategy for a sustainable return on investment
According to the Stack Overflow’s 2024 Developer Survey, 63% of professional developers said they are currently using AI in their development process. Additional 14% plan to do so soon4. The developers seem to consider AI as a means of writing more code, more quickly (11.).
The “calculator” of the impact of the quality of the code, associated with the principles of Software Craft, good development practices, and a strategic and measured integration of AI, offers a concrete approach to assess and improve the productivity of developers and reduce development costs (2.) by understanding the costs of technical debt and by adopting a culture of technical excellence, while sailing with prudence AI, companies can transform the quality of the code and artificial intelligence into real competitive advantages (1.).
It is essential to measure the real impact of the AI on the entire development cycle, and not only at the individual level, in order to maximize its benefits and to mitigate its potential negative effects on the stability and performance of delivery (3.). Investing in software craft and a thoughtful AI integration strategy is not only a technical approach, but a judicious business strategy for lasting success (2.)
Discover the “Calculet of the impact of AI and Software Craft” (IA & Craft Impact Calculator) at Arolla.
Sources:
1. Codescene report-https://codescene.com/blog/evaluate-code-quality-at-scale/7
2. Codescene report-Business-iMPACT-OF-CODE-QUALITY.PDF
3. Dora report – https://services.google.com/fh/files/misc/2024_final_dora_report.pdf
4. Source Sciencedirect – Software Developer Productivity Loss due to Technical Debt – Replication and extension Study Examining Developers’ Development Work
5. Source Enhancing Developer Productivity Through Automated CI_CD Pipelines_ A CONCECTHENSIVE Analysis.pdf
6.. Intelligent source-devops-harnessing-artificial-intelligence-to-revolutionize-cd-pipelines-and-optimize-software-delivery-lifecycles.pdf
7.Source Impact-of-Automation-On-Quality-Assurance-TESTING-A-COMPARATIVE-ANAVALYSIS-OF-MANUAL-VS-ATOMATED-QA-PROCESSES.PDF
8. Agilealliance article https://www.agilealliance.org/wp-content/uploads/2024/07/michele-brissoni-behavioral-engineering-at-the-oftware-craftsmanship-dojo-a-14-yaar-adventure-empowing-over-15000
9. The Dora 2024 report “Accelerate State of Devops”
10. McKinsey article entitled “Developer Velocity: How Software Excellence Fuels Business Performance”
11. Gitclear report 2025 https://gitclear-public.s3.us-west-2.amazonaws.com/gitclear-ai-copilot-code-quality-2025.pdf
12. Snyk report https://snyk.io/fr/blog/copilot-amplifies-instécure-codebases-by-replicatting-vulnerabilities/
