DEVOPSdigest

Applause has released its fourth annual State of Digital Quality in Functional Testing report, offering a comprehensive look at how organizations test, measure and improve digital experiences. Based on a global survey of more than 2,100 software development and QA professionals, this year's findings reveal a clear shift toward AI-driven testing, broader adoption of human-in-the-loop approaches and an industry-wide move toward continuous quality across every phase of the software development lifecycle (SDLC).

AI Adoption Doubles, but Human Oversight Remains Critical

AI has officially gone mainstream in software testing. 60% of respondents reported that their organization now uses AI somewhere in the QA process, twice the rate reported a year ago. Teams primarily rely on AI to develop test cases (70%), automate scripts (55%) and analyze test outcomes or recommend improvements (48%), while others use it for test prioritization, autonomous execution, gap identification and self-healing automation.

Yet, efficiency doesn't equal maturity. 80% of respondents said their organizations lack in-house AI testing expertise, and 92% struggle to keep up with changing requirements. Additional challenges include unstable environments (87%) and insufficient testing time (85%).

The lesson is clear: AI boosts coverage and speed, but human oversight is still essential to ensure accuracy, safety and usability. As agentic AI systems become more autonomous, embedding human review early and often remains a best practice for preventing quality blind spots.

Shift-Left Testing Becomes Standard Practice

QA is far from an end-of-pipeline checkpoint. The 2025 data shows a strong industry move toward continuous, shift-left testing. In previous years, 42% of organizations limited testing to a single stage of the SDLC; this year, that figure dropped to just 15%.

More than half of respondents now conduct testing during planning (54%), development (59%), design (52%) and maintenance (57%) phases. Meanwhile, 91% of organizations perform multiple types of functional testing, including performance, UX, accessibility and payment testing.

This evolution reflects a more proactive approach to quality, identifying and resolving issues before release rather than reacting to them afterward.

Crowdtesting Bridges the AI Coverage Gap

While AI and automation accelerate many parts of testing, they can't replicate real-world variability. That's why one-third of organizations (33%) now rely on crowdtesting to supplement internal QA.

Crowdtesting introduces diverse, human-in-the-loop validation across devices, locations and accessibility contexts that AI can't fully capture. It's also proving especially valuable as organizations adopt agentic AI systems that operate autonomously and interact across unpredictable environments. Combining machine efficiency with human perspective gives QA teams the balance of scale and context needed to maintain user trust and deliver comprehensive quality.

Digital Quality Metrics Are Becoming More Customer-Centric

This year's data confirms that quality is now defined through the customer's eyes. Customer satisfaction and sentiment remain the top metrics used to evaluate software quality. Among organizations that use multiple metrics, 67% rely on test case reporting and analytics to track trends, and 58% use that data to guide future development priorities.

When it comes to testing types, UX testing (68%) leads the way, followed by usability (59%) and user acceptance testing (54%). These figures signal a decisive shift from validating code performance to validating human experience, ensuring digital products are not only functional but intuitive and engaging.

Documentation and Reproducibility Need Improvement

Even as testing sophistication grows, foundational QA practices still lag. 69% of respondents rated their QA structure and consistency as "Excellent" or "Expanding," yet only 33% said they maintain comprehensive test documentation.

At the same time, 84% struggle to reproduce defects with available test data — a major roadblock to fast, reliable debugging. Addressing documentation and reproducibility gaps remains one of the most impactful ways teams can elevate overall digital quality.

Blending AI Efficiency With Human Insight

The 2025 findings make one thing clear: the future of digital quality is hybrid. Organizations are embracing AI and automation to move faster and test smarter, but they also recognize that human expertise is critical to interpret results, understand context and ensure ethical and functional integrity.

Leading DevOps and QA teams are adopting blended models, integrating AI-powered test generation and analytics with human-driven validation through crowdtesting and expert oversight. They're embedding testing throughout the SDLC, tracking user-centric metrics and focusing on trust as the ultimate measure of success.

As AI continues to reshape development, the organizations that thrive will be those that treat digital quality not as a checkbox, but as a continuous partnership between human insight and intelligent automation.