The 2025 State of Testing in DevOps Report

Understanding that there are nuances in how organizations move through their DevOps transformations, it’s important to understand where they are now in order to give context to the current state of software development and testing. So, while the path for each organization varies, we define it in four stages: aspiring, striding, mostly, and fully DevOps. This allows us to see trends and challenges as teams work towards the goal of continuous integration, deployment, and improvement.

The speed at which teams are adopting a culture of DevOps is staggering. Despite the obvious benefits, only 33% of organizations considered themselves mostly or fully DevOps in 2024. This year, that number has jumped up to 60%, with 34% saying they have fully embraced a DevOps culture, KPIs, and DevOps practices across the organization.

Last year was the first time technology limitations overtook budget as the largest impediment to investing more heavily in quality, and that trend continues in 2025. This is especially true for organizations who have fully embraced DevOps, with 34% reporting it as the largest inhibitor, compared with 20% reporting budget as the main concern. For those who are making strides in their DevOps transformation, however, budget remains the top concern yet again, with 29% reporting it as their most significant inhibitor.

Along a similar vein as DevOps maturity, we continue to see unified testing holds an important place in planning, with 81% of respondents noting that they have a unified or partially unified strategy for end-to-end testing. This is challenged, however, by the data that says 88% of companies are using two or more testing tools, with an eye-watering 23% reporting using five or more tools, up 8% over 2024.

One of the more significant shifts we’ve seen in 2025 is the growing influence of AI and the increasing complexities of testing as teams continue to grow in their DevOps maturity. It underscores the need for organizations to prioritize investment in AI-driven automation and unified testing solutions to help navigate the challenges and maintain a high standard of quality. 

The relentless pursuit of having the “next great thing” is pushing companies to deliver innovation quickly, driving a significant acceleration in release velocities. In fact, 40% of respondents noted that they are deploying code at a 50% or greater speed than they were this time last year. The adoption of DevOps practices is pushing this even further, with 50% of fully DevOps organizations reporting a 50% or greater release speed.

A key driver in this accelerated pace is the adoption of AI in development workflows. On average, 55% of organizations are using AI in their development and testing process. For teams who have fully embraced DevOps, that number is even higher at 65%. This is empowering developers to create code at an unprecedented rate, which in turn shortens release cycles. While this is undoubtedly beneficial in the long-term, in the short-term it is increasing the pressure on testing teams to keep pace.

Adding to the complexity is the increased usage of packaged apps like Salesforce, SAP, Microsoft 365, and Workday. Nearly half (49%) of respondents reported using at least one custom app and one packaged app in the products their test cases span. An additional 39% reported using multiple custom applications. While these approaches offer a number of advantages in terms of speed and cost-effectiveness, they also introduce a whole new set of testing challenges. Maintaining quality in a product that is reliant on the quality of a second product is inherently complex, as it introduces external dependencies, third-party updates, and changes to configurations and client experiences. This means teams are reliant on a variety of testing tools and solutions to maintain high standards of quality.

While innovation continues to move forward, there is a disconnect between that and the quality of what is being released. This is obviously a concern when it comes to the long-term stability of a product, but it also poses a significant risk to customer satisfaction. Our survey revealed a series of critical challenges facing quality teams and contributing to the lag between the two:

• Productivity Drains: Test maintenance consumes 20% of team time
• Incomplete Coverage: Only 14% of teams have 80%+ coverage
• Customer-reported Bugs: 1 in 3 production bugs is found by a customer
• Testing Silos: 82% of teams are using multiple testing tools, hindering unified testing

Where Your Team’s Time is Spent

For the second consecutive year, test maintenance was reported as the most significant use of a team’s time, taking up a full 20%, or one full working day each week. Add to this the amount of time spent on test case management and identifying and fixing defects, and it’s clear that the standards for automation testing are not being met by the plans teams currently have in place.

Missing Test Coverage

Piggybacking off of the productivity drains, we have the related issue of incomplete test coverage. While only 14% of teams have more than 80% end-to-end test coverage, this is a slight improvement over the 9% reported in 2024. If you look at the pace of innovation in development, however, this uptick pales in comparison and shows the need for test coverage to increase if quality is going to keep up with development. One possible correlation here is that too much time is being dedicated to maintenance, leaving test coverage to suffer.  

It Matters Who is Testing

When implemented correctly, the shift-left approach to testing enhances quality efforts by integrating testing earlier in the software development life cycle and giving everyone a sense of responsibility for the ultimate outcome of a product. Following the mantra of “test early, test often,” teams can identify issues and deploy fixes well before they hit production. In the absence of proper training, whole-team buy-in, and adequate tools, however, it can lead to inconsistencies and overlooked defects. 

• 2 in 3 respondents to the survey report that developers are involved in end-to-end testing
• 1 in 3 respondents to the survey report that customers are the ones to report production-level bugs

The above statistics expose a significant gap in internal testing processes, and highlight how teams are effectively relying on their customers to act as end-user testers. When you consider that 29% of respondents express concerns about introducing packaged apps to their product based on the implications they have on the customer experience, one can’t help but wonder why the same concern isn’t present in properly training developers to execute effective testing plans. 

Excess Tools Create Testing Silos

A staggering 82% of teams are using more than one testing tool as a part of their quality plan, something that aligns with the proliferation of custom and packaged app integrations. This is in stark contrast to the number of teams who recognize the importance of a unified testing policy, sitting at 81%; the more tools you integrate into your test process, the less likely you are to have a truly unified plan. Oftentimes, a single testing tool isn’t able to accommodate the rapidly-evolving use cases for each individual packaged or custom app, so teams wind up with one specific tool to test one specific app. The result is that there is an incomplete and siloed picture of the testing landscape due to the limitations of each tool. This usage also varies based on the size of the business, with large Enterprises being far more likely to rely on multiple apps as well as multiple testing tools.

It’s an indisputable fact that AI is changing the landscape of development, but the room for opportunity (and where it will become critical to an organization’s success) really lies in testing. Our survey revealed that a majority are leveraging AI tools in both their development and testing processes. This widespread adoption underscores the importance of AI adoption in the coming years.

• 55% of organizations are using AI for development and testing
• 44% of organizations are seeing a reduction in testing needs based on the increased quality of code produced by AI
• 38% of organizations who are not currently using AI are planning to in the next year

Company Size Determines the Detriments

While a majority of organizations are using AI with increasing frequency, its adoption is not uniform across the board. Large enterprises are more likely to use AI for development – 29% versus a 21% average – which tracks with the ability to move faster and at scale. It is interesting to note that teams who have fully embraced DevOps have significantly higher adoption rates across the board, with 65% using AI in development and 70% using it in testing. Teams who have automated all of their workflows are reporting even higher rates of AI usage, with 74% using it in development and 71% using it in testing.

AI in Testing: How Teams are Using It

The most prominent use of AI in testing is to optimize test execution by prioritizing which tests to run. This strategy is employed by 55% of organizations and, interestingly enough, is even more popular in the UK, where 63% of organizations use it for this purpose. Company size also affects how likely a team is to use AI to optimize test runs, with adoption by large enterprises sitting at 65%.

The stage a company is in its DevOps transformation is also an important indicator of the way they use AI in testing, with fully DevOps teams being more than twice as likely to use it for managing test cases and test suites than those who are still aspiring to make that transition.

Overall, the move towards adoption of AI for testing is increasing, especially when it comes to summarizing test results – 45% in 2024 vs 49% in 2025 – and generating test cases – 38% in 2024 vs 47% in 2025. These increases, while incremental, show a trend that we expect to continue into the coming years.

Barriers Still Need to Be Broken Down

Regardless of how prevalent AI is in our everyday lives, there are still a number of barriers to adoption that need to be addressed before organizations are comfortable with taking the plunge. The top concerns are security (34%), quality (31%), and the lack of AI expertise (30%). It is interesting to note, however, that the specific concerns vary based on the size of the organization. In addition to being more concerned about the ROI of AI tools, large enterprises are also more likely to have corporate policies that restrict the usage of them.

While this data underscores the potential for AI to transform the development and testing landscape, it also highlights the need for organizations to address security, quality, and expertise concerns to fully capitalize on the numerous benefits of the tools. By being strategic with integrations, properly vetting tools for specific needs and guidelines, and dedicating budget to tools that save money in the long run, organizations can overcome these barriers quite easily.

As organizations navigate the complexities of accelerated releases, increased usage of package apps, and the integration of AI, strategic investments are paramount. To that end, our survey reveals a clear trend: relief for problem areas is on the way in the form of quality assurance budget increases. The investment isn’t limited to hiring more people; it’s about strategically deploying these resources to enhance automation, tooling, and upskilling initiatives.

That being said, 51% of organizations report that part of their increased spend is in fact going towards adding new QA hires to their teams. This indicates that they see the need for more capacity to execute on the additional spends they are planning, like the 62% who will be increasing spend on automation tooling. These investments in human resources and technological capabilities are indicative of a commitment to moving quality forward at the same pace as development.

The push for automation isn’t just about improving efficiency, though that is a significant benefit; it’s about leveraging AI in a way that addresses productivity drains like test maintenance and coverage while also streamlining the process. 38% of organizations who are not currently using AI for development or QA are planning on investing in it in the next year, signaling a growing recognition of its importance to the development lifecycle. 

Organizations who report having all of their workflows automated are even more likely than their counterparts to indicate that they will continue investing in automation in the future, highlighting the fact that as automation has benefitted them already, there are still benefits left to be realized by investing into it even more. 

The 2025 State of Testing in DevOps Report highlights the fact that we are at a critical juncture in the future of software development and testing. The relentless drive for speed, fueled by DevOps practices and AI-driven development, has transformed release cycles and underscores the industry’s commitment to rapid innovation and faster time-to-market.

This pursuit of speed has also revealed a growing tension with quality, with 59% of QA teams noting that the increased development speed has put significant strain on their testing programs. Productivity drains, largely stemming from test maintenance (consuming 20% of team time), and persistent gaps in coverage (with less than 15% of teams getting over 80% test coverage) are impeding even the best efforts to build robust and reliable software. The fact that 1 in 3 production-level bugs are still being discovered by end users, despite 80% of respondents reporting a high level of customer satisfaction, points to a dangerous disconnect that organizations can’t afford to ignore. 

Navigating this complex landscape will take a strategic focus on automation and unified testing. The current siloing of testing tools and practices, with 82% of organizations using multiple testing tools, despite virtually the same number – 81% – recognizing the importance of a unified testing strategy, highlights the growing and urgent need for more integrated solutions. These solutions should not only streamline workflows, enhance collaboration, and boost productivity for the organization but also leverage the transformative power of AI.

It’s clear that AI is more than a passing trend; it’s the new face of quality in DevOps. With 55% of organizations already adopting AI for development and testing, and mature DevOps teams leading the way with 70% adoptions, it has never been more obvious that AI-powered testing will be an essential piece of a successful DevOps strategy. This aligns with findings from Gartner in a 2023 study that shows “63% of organizations are currently piloting, deploying, or have already deployed AI code assistants,” and that, “by 2028, 75% of enterprise software engineers will use AI code assistants, up from less than 10% in early 2023.” (Gartner Press Release, “Gartner Says 75% of Enterprise Software Engineers Will Use AI Code Assistants by 2028,” April 11, 2024, GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved.) 

The ability to find a solution that optimizes test execution, automates test maintenance, and expands coverage is what offers teams a pathway to achieving both speed and quality for their end product. This will become increasingly important as agentic AI continues gaining steam and makes itself integral to the customer experience.

In conclusion, the future of DevOps hinges on a strategic balance:

• Equity in speed, quality, and customer satisfaction: Organizations must continue to innovate, while prioritizing quality and recognizing that customer satisfaction is the ultimate measure of success.
• Embracing automation and unified testing: Investing in automation and platforms that harness unified testing is crucial in addressing productivity drains and testing siloes.
• AI-powered testing is the way of the future: Using AI for testing programs is no longer a luxury; it is a necessity. Achieving high standards for velocity and quality requires modern solutions, and they are built on and powered by AI.

By embracing these principles, organizations can effectively navigate the challenges and opportunities of the AI-driven DevOps era and deliver high-quality software at a pace that drives innovation forward, optimizes QA teams as a resource, and delights customers.