Free tools Windows power users keep installed
One-click scans. No signup required.
Backend development in 2021 did not have one defining technology: developers combined programming languages, web frameworks, databases and cloud platforms to build and run server-side systems. Stack Overflow’s 2021 Developer Survey offers a useful snapshot of technologies respondents said they had worked with extensively over the previous year, but its figures describe developers broadly—not backend-only adoption or the market share of production systems.
How to read the 2021 technology figures
Stack Overflow’s 2021 Developer Survey received responses from more than 80,000 developers. It asked about technologies used extensively over the past year, as well as technologies respondents wanted to use over the next year. Its questions covered programming and scripting languages, databases, cloud platforms, web frameworks and developer tools as separate categories. Stack Overflow’s survey announcement describes the respondent pool and its scope.
These are self-reported survey results, not a census of deployed applications. A framework or cloud-platform percentage should not be read as the share of backend systems using it: respondents were not limited to backend developers, and each question had its own response group. The figures are most useful as signals of what developers reported using or, in a different measure, wanted to use.
Web frameworks: Express and ASP.NET Core
In the survey’s web-framework results, 23.82% of respondents to that question reported using Express, and 18.10% reported ASP.NET Core. These are reported-use figures for the survey’s framework question, not backend-only adoption rates. See the 2021 web-framework results.
#1 Best Overall
The figures show that both frameworks were represented among respondents’ work; they do not establish which is faster, safer, less expensive or better suited to a particular project. Those choices depend on the application’s requirements, the team’s experience and the surrounding stack.
Databases: distinguish interest from use
The database figures often cited from the survey’s interest results are not current-use percentages. They represent respondents who were not already developing with a database but expressed interest in using it. PostgreSQL registered 17.99% interest, MongoDB 17.86%, and Redis 12.58%. The database results show the separate interest and use measures.
That distinction matters when comparing options. These numbers point to technologies developers were interested in, not the proportion of respondents—or applications—already using each database. Nor do they establish that a relational database, document database or in-memory data store is the right fit for any particular workload.
Cloud platforms: AWS and Azure
For the cloud-platform question, which received 46,242 responses, 54.22% reported extensive development work with AWS over the previous year, and 30.77% reported extensive work with Microsoft Azure. These percentages describe respondents to that question; they should not be interpreted as exclusive platform shares or as backend-specific market shares. Review Stack Overflow’s cloud-platform results.
Rank #3
The survey provides a popularity signal, not a comparison of platform performance, security, cost or suitability. Choosing a cloud provider still requires evaluating the project’s technical needs and operational constraints.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Languages and developer tools are separate layers
The survey also covered programming languages and developer tools. Its broad results include languages such as JavaScript, Python, Java, C#, and Go, as well as tools including Docker, Kubernetes, and Git. The cited results do not isolate which of these respondents used specifically for backend development, so they cannot support a backend-language ranking. The survey’s technology results separate these categories.
A backend stack can involve a language for application logic, a framework for structuring web services, a database for persistent data, and infrastructure or tools for deployment and operation. A single language ranking would leave out these other layers and overstate what the survey establishes.
Quick Recap
What the survey can—and cannot—tell you
- It can show: which technologies respondents reported using extensively, and which technologies some respondents not already using them said they were interested in.
- It cannot show: backend-only adoption, the percentage of production systems using a technology, or a technology’s performance, security, cost, scalability or suitability for a specific workload.
- For a project decision: compare options at the same layer—language with language, framework with framework, database with database, and platform with platform—against the project’s actual requirements.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
