Launched by Microsoft in 2021, GitHub’s AI coding assistant Copilot came with a promise: to help developers “rediscover the joy of coding” while speeding up the development cycle. 

How would it do that? Artificial intelligence and automation. One of the first AI code-assist platforms of its kind, Copilot is designed to automate the grunt work of coding: generating boilerplate code, writing basic functions, and creating test cases.

By suggesting code snippets and handling tedious tasks, Copilot frees developers up for more creative and complex work, boosting job fulfillment and developer productivity.

So how has that promise held up? Recent research —including a GitHub survey and our own findings at Opsera— provides insights into Copilot adoption and usage across various industries.

By the numbers: Copilot adoption trends

So, what do the numbers say about Copilot adoption? According to a GitHub survey of Accenture developers, many programmers are eager to adopt Copilot; 81.4% of respondents installed the GitHub Copilot IDE extension on the day they received their license and 67% say they use Copilot at least five days per week. The survey results indicate strong enthusiasm for  Copilot; among Accenture developers, 43% find Copilot “extremely easy to use,” and 51% rate it as “extremely useful.”

While GitHub’s data provides valuable insights, we wanted to take a closer look at how Copilot adoption plays out across various organizations. Using real-world data from companies leveraging Opsera Unified Insights, we analyzed Copilot adoption and usage trends. Our research found that the majority of developers adopted Copilot as soon as a license was available to them, with 80% of licenses in use.

Developer productivity

Does Copilot make a difference when it comes to productivity? Github’s data (from the Accenture study) says it does:

• 8.69% increase in pull requests (PR)
• 15% increase in pull request merge rate

What does this mean? Measuring pull requests is key to measuring development speed; more PRs mean that work is being broken into smaller, manageable chunks, leading to faster iteration and quicker releases. 

While measuring PR volume gives a sense of development speed, it doesn’t always capture the full picture, such as code quality or team collaboration. Combine that with Opsera’s research, which shows that Copilot users not only reduced time to PR by days (from 9.6 to 2.4 days) but also maintained or improved work quality, and you get a more complete view of productivity.

Together, these findings show that Copilot helps developers become more efficient, whether by increasing the volume of PRs or reducing the time to complete them.‍

Code quality

When it comes to the quality of the AI-created code, Github’s research found that:

• Developers accepted around 30% of Copilot’s suggestions
• 91% of the developers reported that their teams had merged pull requests containing AI-suggested code. 
• Accenture developers retained 88% of GitHub Copilot-generated characters in their editor. 
• Accenture saw an 84% increase in successful builds, suggesting that the AI-assisted pull requests were often of higher quality when reviewed.

Opsera’s data confirms this; 

• Developers accepted 31% of Copilot suggestions
• 17% of Copilot-generated suggestions remained in the final codebase after edits and reviews.

Together, this data reinforces that while developers are selective about AI-generated code, the portions they keep contribute to high-quality, production-ready software.

Adoption by industry

GitHub Copilot adoption varies across industries. Below are key adoption metrics for several sectors: 

Different industries adopt AI coding tools at different rates, and the data shows clear trends in how Copilot is being used across sectors. While some industries fully embrace AI-assisted coding, others take a more cautious approach.

• Tech and startups: Highest Copilot acceptance rate, suggestion rate, and productivity gains. These industries are early adopters, relying on agility and new technologies.
• Banking and finance: See similar productivity increases as tech but accept fewer suggestions due to strict security and quality standards.
• Healthcare: Shows caution with lower Copilot acceptance rates, driven by the need for rigorous testing and validation.
• Insurance: Lowest acceptance rate, likely due to complex industry regulations. However, anything above 35% is still considered good.
• Industrial sector: Lower acceptance rate and productivity gains, though slightly ahead of insurance. Likely impacted by reliance on legacy systems.

Overall, industries with fewer regulatory hurdles and a culture of rapid iteration are adopting Copilot at a faster pace. More risk-averse sectors are integrating AI more cautiously, but the data shows that even in highly regulated industries, Copilot is making an impact.

Opsera: driving Copilot adoption

To successfully drive GitHub Copilot adoption and measure its impact, leaders need real-time visibility into how developers are using it. Clear baseline data, along with dashboards tracking adoption, usage, and other KPIs are essential. 

Opsera Unified Insights simplifies DevOps measurement, delivering insights into Copilot’s impact on DevEx, productivity, ROI trends, and overall business value. It’s the only platform that provides a comprehensive view of Copilot’s effect on developers, teams, and projects.

By connecting all your DevOps tools, Opsera gives you a complete picture of the software delivery process, enabling continuous improvement. With built-in historical tracking, Opsera makes it easy to compare baseline performance over time—helping your organization measure the true ROI of GitHub Copilot.

Ready to see Copilot data for your own team? Unlock real-time insights, optimize workflows, and drive better business outcomes with Opsera’s Unified Insights. Try it today and get the data you need to drive your organization’s Github usage.