Let’s talk about the Future of Jamstack — Join us

Jamming into the Mainstream: Jamstack Community Survey 2021

This year we saw Jamstack go mainstream, with big changes in working patterns for Jamstack developers in every region. Our community has grown, adding not only experienced developers but also becoming the first stop for those just breaking into the industry. We also learned that the move to remote work has been significant and lasting among web developers.

Netlify sits at the center of the Jamstack community, and we conduct our annual survey so we can understand our community of developers. This helps us tailor our products and services to our community. In sharing our survey results, we also want to help developers better understand themselves and one another. Working as a developer often means working in a vacuum, without a sense of what’s happening in the broader community. Our survey data can help provide a sense of best practices as well as an idea of what else is happening in the community.

This year, for example, you told us that content management systems Sanity and Strapi are having a moment. You shared that security is a rising priority for all devs. You also offered a lot of insight about new developers, from which specific devs have moved to remote work to the (obvious and less obvious) reasons why.

To get a better picture of the full developer landscape, we surveyed people who build websites in general, regardless of whether they build Jamstack sites, though our analysis focuses on active Jamstack developers.

We ran this year’s survey from June 23 to August 10 and received over 7,000 responses, more than twice as many as last year’s survey. You can review our methodology document for a detailed breakdown of the demographics and accuracy of the survey sample and results.

To the developers who took the time to respond to our survey to help us learn more about the Jamstack and our community: Thank you. We hope you walk away from this report learning something about yourselves and the expanding Jamstack community.

Demographics

Jobs

Jamstack has become the way students learn to deploy.

Between 2020 and 2021, we didn’t see much change in our mix of developer and non-developer respondents. That’s good news—it means we’re sampling the same population both years, and the changes we’re noticing in other parts of the survey are real changes, and not just statistical noise.

Developers and non-developers

Source: Jamstack Community Survey 2020—2021
Show Chart Data
People 2020 2021
Developers 79% 81%
Non-developers 21% 19%

This year, we asked respondents to specifically identify themselves as front-end, back-end, or full-stack developers. It’s not surprising that in a Jamstack survey, the vast majority of devs identified as front-end or full-stack.

Job title

Percentage of survey participants

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Job Title Percentage of Survey Participants
Developer: Front-end 45%
Developer: Full Stack 32%
Management 6%
Other 5%
Developer: Back-end 5%
Designer 4%
Content Producer 2%
DevOps 2%

Here’s where we saw a huge change. When we asked about people’s employment status, we saw a big shift in 2021—the portion of our respondents who are students nearly doubled. We’ll discuss the reasons for this change later on in the report, but needless to say, the effects of the pandemic are one factor responsible for the increase in students.

Employment Status

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Employment Status 2020 2021
Full-time 69% 60%
Student 9% 16%
Contractor 11% 10%
Part-time 4% 5%
Unemployed 4% 5%
Between jobs 4% 3%
Retired 0% 1%

We found more interesting data in job statistics in two other breakouts:

A changing landscape of developer experience

Newer developers are becoming much more geographically diverse.

In both 2020 and 2021, we asked developers how many years of experience relevant to their job they have. We noticed a shift to lower levels of experience this year.

Our category of newest developers, those with less than one year of experience, jumped from just 4% to 13% of respondents, and those with 1-2 years rose from 13% to 19%. All categories of those with more experience fell. Because we saw such a surge of students in our breakdown by job title, this makes sense—students have less experience, after all.

Years of Experience, 2020 vs. 2021

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Years of Experience 2020 2021
< 14%13%
1–213%19%
3–420%18%
5–615%12%
7–89%7%
9–1012%8%
11–125%5%
13–145%3%
15+17%14%

To confirm this, we looked further into this question in breakout 1: experience levels of students vs. non-student.

This year, we also asked developers where they were in the world, split into major geographical regions.

The data was unsurprising—mostly Europe and North America—until we split up the region data by the number of years of experience of the developers. We excluded students when doing the geographical breakout.

We noticed a clear trend: Newer developers are significantly more geographically diverse than more experienced developers. In the most experienced category, only 15% of developers don’t come from Europe or North America. But with the newest developers, that number more than triples to 48%. Our hope is that when people work from more diverse places, this fosters greater diversity of other kinds as well.

Years of Experience, by region

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Years of Experience Africa Asia Pacific Caribbean Central America Eastern Asia Europe Middle East North America South America Southern Asia
< 1 5.32% 17.02% 0.53% 2.13% 2.13% 27.66% 0% 23.94% 8.51% 12.77%
1–2 7.49% 12.75% 0.81% 1.21% 2.43% 32.39% 0.81% 25.91% 7.69% 8.5%
3–4 5.09% 9.86% 1.11% 1.11% 1.75% 38.16% 1.43% 28.78% 5.88% 6.84%
5–6 2.48% 9.5% 0.41% 0.62% 0.41% 44.01% 1.45% 32.85% 4.13% 4.13%
7–8 1.28% 8.97% 0.64% 0.00% 1.28% 42.31% 0.96% 34.62% 7.05% 2.88%
9–10 2.88% 5.48% 0.29% 0.58% 0.86% 46.11% 1.73% 34.87% 5.19% 2.02%
11–12 1.35% 9.01% 0% 1.8% 0.45% 43.24% 0.45% 37.84% 4.95% 0.9%
13–14 0.76% 9.09% 0% 0.76% 0% 40.91% 2.27% 40.91% 3.79% 1.52%
15+ 1.08% 6.92% 0.15% 0.46% 0.77% 45.08% 1.23% 40.15% 2.77% 1.38%

Breakout 1: An uptick in students in 2021 led to a shift in experience levels.

To confirm that our changes to experience levels were caused by an uptick in students, we broke out students from full-timers and then looked at their experience levels.

Sure enough: students were concentrated heavily into the first two years of experience.

Experience levels, full-timers vs. students

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Years of Experience Full-time Students
< 16%41%
1–214%38%
3–419%15%
5–614%3%
7–89%1%
9–1010%1%
11–127%0%
13–144%0%
15+16%1%

This leads to the next question: Where did all these students come from? One potential explanation is that as Jamstack has continued to go mainstream, it’s become the default way of teaching new developers to launch their websites. In turn, that’s been increasing the percentage of students in the Jamstack community. We’ll dig more in future surveys to confirm this effect, but having a huge new population of new developers is a great sign for the community—the more, the merrier!

How the pandemic affected working life

Many developers shifted to remote work, and they don’t plan to return to the office.

In 2020 and 2021, everyone’s life was affected by the COVID-19 pandemic in some way—for the purposes of our survey, we focused on how it changed developers’ working lives. What we found surprised us.

How did the pandemic change your working life?

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Pandemic effect
None of the above38%
Kept job—went fully remote31%
Kept job—went part-time remote14%
Lost a job9%
Took remote job that wouldn't have been feasible in-person9%
Moved to take advantage of remote work8%
Quit job to take care of family3%

The biggest change we noticed was that 31% of developers said their job, previously not remote, went fully remote and will stay remote foreseeably—even after the pandemic ends. This represents a huge shift in working patterns for web developers. We also discovered interesting correlations to developers’ level of experience and their job title, which you can investigate in our breakout.

8% of developers said they moved to a new place to take advantage of remote work.

The other notable change we observed was 9% of developers said they lost a job. In section 1 we noted that the number of people reporting themselves as full-time employed had fallen, and the number of students had risen. It’s possible that some developers who lost their jobs moved into education.

Breakout 2: A third of developers went remote during the pandemic, and don’t plan on coming back.

As mentioned, a whopping 31% of developers said they went remote in the pandemic, and plan to stay remote after the pandemic. We wanted to know more about those people.

First, we split up the respondents by experience. There was a notable effect here: The newest developers were much less likely to have “kept their job and gone remote” because they were more likely to have lost their job in the pandemic.

Twelve percent of those with less than 1 year of experience reported losing a job to the pandemic, and 10% of those with less than 2 years did too.

Excluding those with less than 2 years of experience, the percentage of developers who’d gone remote was even higher—40%! But that number didn’t change with further experience.

Developers who went remote during the pandemic, by years of experience

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Years of Experience Remote
< 117%
1–232%
3–440%
5–640%
7–840%
9–1041%
11–1242%
13–1441%
15+39%

Excluding those first two categories, we proceeded to look at job titles, and we noticed some substantial differences. Nearly 50% of DevOps respondents reported going full-time remote, while only 36% of managers said they did.

Front-end developers were more likely to go remote than full-stack developers, even though the data tells us that full-stack devs report having more experience on average than front-end ones.

We wanted to know why that happened. Initially, we thought it might be that the most experienced people were most likely to choose to work remotely. But, apart from full-stack and front-end not matching that pattern, the data indicated that management has even more experience, but is less likely to go remote.

Our conclusion is that we’re not sure what causes this. Could it be related to the number of meetings you have? Which jobs are harder to hire for? We will have to investigate further.

Who went remote?

Percentage of workers by job title who went full-time remote, excluding those with less than 2 years’ experience

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Years of Experience Remote
DevOps50%
Back-end45%
Front-end43%
Full stack39%
Designer37%
Management36%
Content29%

Jamstack adoption

The purpose behind developers’ Jamstack sites

Jamstack sites are used for real business needs.

We’re always interested in what the websites you’re building are for. Are you putting together a personal site to show off your work? Making an e-commerce play? Or building an enterprise application? Our 2021 categories were a little different from the ones we used in 2020, so the results aren’t precisely comparable. But they’re still fascinating. Since people work on many sites over the course of a year, we allowed respondents to pick more than one option.

What is the purpose of the sites you work on?

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Years of Experience Remote
Personal blog/portfolio37%
B2B software35%
E-commerce34%
Consumer software33%
Informational33%
Lead capture/landing pages29%
Internal tools29%
Enterprise software25%
Documentation16%
Retail12%
News/Entertainment12%
Social media9%
Other7%
Games7%
Streaming media6%
Politics/activism5%

Personal websites dominate, just like last year. Whatever else you build, you’re always building your own website too!

E-commerce is a huge driver, with a surprisingly strong showing by enterprise software at 25%.

Developers are serving audiences of millions

Though developers use Jamstack to build websites for audiences of all sizes, a third build the biggest websites in the world.

We asked developers how many users the sites they build are intended to serve. Unsurprisingly, most devs work on websites for relatively small audiences. But one stat did energize us—32% of developers are working on sites that serve audiences of millions of users.

How many users are your sites intended to serve?

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Number of users Most sites Some sites
Tens28%38%
Hundred31%45%
Thousands42%37%
100 Thousands22%32%
Millions12%19%

Breakout 3: Specialization differentiates developers who work at big websites from smaller ones.

With 32% of developers saying they work on the very biggest websites, which have audiences in the millions of users, we wanted to see if this group differed substantially from the average developer.

Job titles of devs who work on very large websites

Source: Jamstack Community Survey 2020–2021
Show Chart Data
Frequency of large sites Content producer Back-end Front-end Full stack
Never1.48%3.18%31.18%48.75%
Sometimes2.24%4.08%29.25%45.98%
Mostly2.58%4.37%34.92%39.88%

The trend we’ve noticed here is specialization. In other words, the more likely you are to work on a big website as a developer, the more likely you are to be a specialist. Small teams don’t have dedicated content producers and are more likely to have full-stack developers. But those who work on big sites are more likely to specialize in back-end or front-end development.

We found another trend when we asked developers focused on large sites how important mobile devices are as a target. The more likely you are to work on large websites, the more likely you are to target mobile devices, and the more likely you are to consider them “very” important rather than just “somewhat” important.

How important are mobile devices for you as a target?

Source: Jamstack Community Survey 2020—2021
Show Chart Data
How frequently do you build sites for audiences of millions Somewhat Very
Never68%25%
Sometimes76%21%
Mostly83%14%

Industries

Every major industry uses Jamstack—not just tech-forward sectors.

Every year, we ask developers what industry their company is in. Big companies often compete in more than one industry, so respondents can pick several industries.

Similar to last year, the top 10 companies reflect some obvious categories but also some surprises.

Advertising, marketing, media and publishing all seem like tech-forward industries you’d expect to find on the Jamstack. Education, finance, and healthcare, on the other hand, aren’t known for being early adopters.

Top 10 industries for Jamstack development

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Industry Percentage
Advertising & Marketing18.4%
Education15.9%
Finance & Financial Services12%
Media & Publishing11.8%
Business Support & Logistics10.2%
Healthcare & Pharmaceuticals9.5%
Entertainment & Leisure9.4%
Nonprofit8.7%
Telecommunications & Networking8.3%
Retail & Consumer Durables8.3%
N/A (Student/Unemployed/etc.)7.7%
Government6.9%
Food & Beverages5.8%
Real Estate5.3%
Manufacturing5%
Transportation & Delivery4.6%
Utilities Energy and Extraction4.5%
Construction Machinery and Homes4.4%
Insurance4.4%
Automotive4.1%
Consumer electronics3.7%
Agriculture3.3%
Airlines & Aerospace (including Defense)2.9%

How Jamstack developers use server-side technologies

Functions are becoming as ubiquitous as containers.

Jamstack isn’t all client-side, so we asked about the popularity of some major server-side technologies as well. We were particularly interested in how much adoption serverless functions are seeing compared to other server-side technologies.

Awareness was high for most techniques—over 75% in most cases—while usage was below half for each.

Server-side technologies by usage and awareness

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Technique Percentage of respondents who use this Satisfaction score Most aware
1. Containers 49% 3.1 94%
2. Microservices 44% 4.9 89%
3. Orchestration 27% 2.4 89%
4. Functions as a service 46% 8.7 87%
5. Gateways 42% 6.4 80%
6. Event Hubs 14% 2.3 68%

Containers like Kubernetes are popular, and container orchestration is only used by half as many developers.

Functions as a service, like Netlify functions, are now almost as ubiquitous a technology as containers.

We’ll keep an eye on how functions usage has grown when we ask this question next year.

Workflows

Devices you target

The desktop still reigns

Industry thought leaders have been beating the “mobile-first” drum for years now, but is the industry actually listening? We asked how important it is for the sites devs build to work on four main categories of devices.

Most targeted devices: How important are these device types for you?

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Target Very Somewhat
Desktop browsers84%14%
Phones71%24%
Tablets41%50%
Watches / IoT7%18%

Despite the talk of “mobile-first,” desktop remains Jamstack devs’ most-targeted category, although phones are close behind. The gap between the two has narrowed very slightly since 2020.

Another interesting finding is how priorities shift. Desktops are very important for nearly everybody, but when it comes to tablets, more than half of developers who pay attention to them as a target think they’re only somewhat important.

In our breakout on people who build large websites, we found an interesting correlation to the importance of mobile devices as a target.

Priorities

Jamstack developers are taking security more seriously

All developers will agree that it’s important that your website is fast, secure, and has high uptime. But a different question to ask is: Which of these is the most important quality?

What is the most important when building a website?

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Priority Score
Performance4.54
Uptime3.87
Security3.61
Dev Speed3.38
Compliance3.08
Avoid lock-in2.56

Last year, we noted that speed of development was a higher priority than security. This year, in a statistically significant shift, those factors were reversed. Jamstack developers are taking security more seriously than they did a year ago.

Performance and uptime continue to be the top priorities for developers.

Which design tool stands out for Jamstack designers?

There’s no contest. Jamstack designers love Figma.

A substantial population of designers responded to our survey, so in addition to asking about development tools, we also ask about design tools. The story in 2021 is the same as last year: Everybody uses Figma, and everybody loves it.

60% of respondents use Figma. Its satisfaction score is 8.8, meaning 8 times more people want to increase their usage than want to stop using it.

It’s getting kind of embarrassing for everyone clustered over there in the bottom left…

Design tools: usage vs. satisfaction

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Design tool Percentage of respondents who use this Satisfaction score Awareness
1. Figma 62.3% 8.8 90.49%
2. Framer 10.82% 2.2 52.67%
3. Adobe XD 36.27% 1.4 88.07%
4. Balsamiq 14.46% 1.0 58.27%
5. Sketch 32.59% 1.0 81.69%
6. Zeplin 20.11% 0.9 55.92%
7. InVision 24.90% 0.8 67.19%

Developers’ favorite code editors

An ambiguous question unintentionally gave us valuable insights. Jamstack developers love CMS editors.

This is the first year we’ve asked about the kinds of editors developers use. We were curious about the growth in popularity of web-based code editors, including products like Glitch and GitHub’s Codespace.

However, we mistakenly presented an ambiguous question. A “web-based editor” can also mean a CMS editor, like WordPress. And since CMSes are very popular, developers answering that question dominated the results: The more you use a CMS, the more likely you are to answer that you used a web-based editor.

But this in itself is a fun result! The popularity of CMSes means more people use web-based editors than traditional editors like Vim.

However, integrated development environments (IDEs) are by far the most popular option and have the highest levels of satisfaction by a long shot.

Types of editors: usage vs. satisfaction

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Editor Percentage of respondents who use this Satisfaction score Awareness
1. IDEs 95.89% 15.6 99.3%
2. Web-based 47.47% 2.6 97.28%
3. Vim etc. 42.32% 1.4 96.98%
4. Plain text 25.66% 0.7 94.12%

Technology choices

Which content management systems were devs’ favorites in 2021?

2021 was a breakout year for Sanity and Strapi.

CMS usage vs. satisfaction

Source: Jamstack Community Survey 2020—2021
Show Chart Data
CMS Usage Satisfaction Awareness
1. WordPress 39.43% 0.3 93.32%
2. Contentful 22.13% 2.2 70.13%
3. WordPress (Headless) 21.57% 1.1 86.67%
4. Strapi 17.89% 2.9 64.28%
5. Drupal 13.36% 0.4 83.95%
6. Sanity 12.72% 3.2 54.13%
7. Prismic 12.31% 2.0 53.79%
8. Wix 11.4% 0.5 86.47%
9. Ghost 10.44% 1.3 63.67%
10. Webflow 10.38% 1.1 62.92%
11. Squarespace 10.22% 0.5 74.67%
12. Forestry 9.32% 1.5 48.39%
13. Weebly 5.77% 0.6 61.35%
14. Contentstack 5.45% 1.0 35.28%
15. Agility CMS 5.33% 1.6 38.84%
16. Builder 5.13% 1.2 31.03%

Content management systems: Can’t live with them, can’t live without them. We asked developers to tell us a lot about CMSes: Whether they were aware of them, whether they used them, and whether they wanted to use them more or less.

We turned the ratio of “people who want to use it more” vs. “people who want to use it less” into something we call our “satisfaction score,” and you’ll see us use it a few times.

A score of less than one in “satisfaction” means more people want to stop using a technology than want to continue using it.

If you compare usage against satisfaction, you begin to see some clear takeaways:

  • WordPress is a clear leader as a content management system, but it is not well-loved, with a score of 0.3, significantly less than 1. However, using WordPress as a headless content manager to a hosting provider like Netlify, while less common, is an arrangement with a much higher satisfaction score.
  • Several well-known content management systems have scores well below one.

CMS changes in usage and satisfaction

Source: Jamstack Community Survey 2020—2021
Show Chart Data
CMS Usage change (%) Satisfaction change Usage
1. WordPress -10.08% 0.2 39.43%
2. Contentful 5.58% 0.3 22.13%
3. WordPress (Headless) 7.26% 0.2 21.57%
4. Strapi 102.15% -1.2 17.89%
5. Drupal -10.09% 0.2 13.36%
6. Sanity 48.08% -3.5 12.72%
7. Prismic 43.98% -0.3 12.31%
8. Wix 95.54% 0.2 11.4%
9. Ghost -2.06% -0.3 10.44%
10. Webflow 77.44% 0.2 10.38%
11. Squarespace -2.06% 0.2 10.22%
12. Forestry -3.82% -0.5 9.32%
13. Weebly 130.80% 0.3 5.77%

We asked the same question last year about many of the same systems, allowing us this year to show a new visualization: How these things are changing over time.

  • In the top right quadrant are systems that are both getting more popular and more loved at the same time. Good going, Weebly, on more than doubling usage!
  • In the bottom right, we see systems that became more popular but less well-loved. Sanity and Strapi, as we saw above, still have exceptionally high satisfaction scores, but as they’ve gained in popularity, we’ve seen a trend we’ve observed previously: Their user base no longer exclusively consists of super-fan early adopters, so their stratospheric scores from last year have dropped. However, their users are still some of the happiest around and they’ve seen huge jumps in usage.
  • In the top left are systems that are getting less popular but more loved. How do we explain this trend? Well, it’s the inverse of what’s happening to Sanity and Strapi. As people migrate away from older systems like Drupal and WordPress, their remaining users tend to be super-fans who are on average happier with them.
  • Finally in the bottom left is a quadrant of systems that are both becoming less popular and less loved at the same time.

And Jamstack’s top programming languages are…

TypeScript is gaining at JavaScript’s expense.

JavaScript is unsurprisingly the primary language for the majority of developers at 55%. However, this is down since last year, from 63%.

Where are those points going? To TypeScript, mostly, which 15% of developers now report as their primary language, not just one they use. Python also went up, from 5% to 7%.

Separate from devs’ primary programming language, we asked about all languages. First, let’s look at their absolute usage and satisfaction:

Programming languages by usage and satisfaction

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Language Usage Satisfaction Awareness
1. JavaScript 96.48% 4.2 99.76%
2. SQL 67.1% 1.6 98.18%
3. TypeScript 60.37% 7.6 97.85%
4. Shell (Bash) 52.08% 1.5 95.98%
5. PHP 46.42% 0.5 97.41%
6. Python 42.41% 2.8 98.03%
7. Java 25% 0.6 96.9%
8. C# 21.5% 1.4 97%
9. Ruby 16.78% 1 95.22%
10. C/C++ 16.17% 0.9 96.92%
11. Go 15.23% 3.4 94.37%
12. Swift 9.42% 2.9 92.06%
13. Rust 8.91% 4.0 91.89%
14. Visual Basic 7.84% 0.5 91.37%
15. Objective-C 6.38% 0.5 91.11%
16. Perl 5.22% 0.5 91.73%
17. Elixir 4.83% 2.4 77.97%

In absolute terms, JavaScript continues to dominate. Workhorse languages like SQL, Bash, and Python remain very heavily used, with TypeScript positioned as a newcomer to the big languages for web developers, with the happiest user base by some distance.

Smaller, newer languages like Rust, Go, Swift, and Elixir are in a cluster of small but happy user communities.

The cut-off line for satisfaction is 1.0, so below that line you see languages with unhappy users: PHP, Java, Perl, Objective-C, and all the C variants, except C#.

But now let’s look at annual changes:

Programming languages by 1-year change in usage and satisfaction

Source: Jamstack Community Survey 2020—2021
Show Chart Data
CMS Usage change (%) Satisfaction change Usage
1. JavaScript -0.7% -1.6 96.48%
2. SQL 4.2% 0.4 67.1%
3. TypeScript 14.1% 1.9 60.37%
4. Shell (Bash) -2.7% 0.1 52.08%
5. PHP -0.2% 0.2 46.42%
6. Python 7.4% 0.5 42.41%
7. Java 2.2% 0.3 25%
8. C# 4.6% 0 21.5%
9. Ruby -1.4% -0.1 16.78%
10. C/C++ 6.3% 0.2 16.17%
11. Go 1.2% 0.9 15.23%
12. Swift 2.3% -0.5 9.42%
13. Rust 4.0% -3.5 8.91%
14. Visual Basic 3.6% 0.3 7.84%
15. Objective-C 1.4% 0.2 6.38%

TypeScript has seen a huge leap in usage (bigger than the change in people who call it their primary language). Its usage increased from 46% to 60%. Even more surprisingly, it experienced a jump in satisfaction—usually as something gets more popular, satisfaction falls, since the user base has expanded beyond the super-satisfied early adopters.

In the top right, you can see languages that experienced an increase in usage and satisfaction. Python and Go are represented in this area, which is great news for them, as they are also big and popular.

The languages with unhappy users (shown in the previous graph) are also mostly in this top right quadrant: Visual Basic, Java, SQL, and the C variants. Their users aren’t happy, but they’re happier than last year.

In the top left are languages with increased satisfaction scores but decreased usage. You’d expect to see legacy languages with fans who are hanging on here, and so here we find PHP and Bash.

In the bottom right are languages that experienced more usage, but are less popular. As expected, up-and-coming languages, like Swift and Rust, are represented here. Go is not here, which is a surprise. It managed to make its users happier even as it grew.

Finally, the bottom left includes languages that lost both usage and popularity: Ruby is here, and so is JavaScript. Keep in mind that JavaScript is still far and away the most-used language, but its dominance is not so total as it was.

Jamstack devs’ favorite frameworks revealed

React is most popular, and Next.js is a big deal.

We asked about an enormous number of frameworks—over 30 of them. Visualizing so many frameworks at once is a challenge, so we split them into two groups: Major frameworks, where the cut-off is at least 10% usage, and then minor frameworks, where usage is less than 10% (keep in mind, developers use multiple frameworks and could mark all the frameworks they used, so these percentages add up to well over 100%).

The major frameworks have a lot of surprises. React is the most frequent choice, as it has been for a long time, and its satisfaction score remains high. Vue has higher satisfaction but roughly half the usage.

Next.js has stellar satisfaction and is really big these days, and Nuxt.js is a little smaller. If you were looking for a safe pick for a new kitchen-sink framework based on this data, Next or Nuxt are where to go.

The big legacy frameworks, jQuery and Express, aren’t going anywhere. But with a satisfaction score below 1.0, jQuery users seem to wish it would.

Relative newcomers Svelte and 11ty are doing very well, with 11ty continuing a strong showing despite relatively low awareness. Early-ish adopters, check these out.

Both flavors of Angular have scores under 1.0, despite maintaining growth.

Major frameworks by usage and satisfaction

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Framework Percentage of respondents who use this Satisfaction score Awareness
1. React 68.1% 4.3 98.4%
2. Express 51.3% 1.9 91.5%
3. jQuery 50.8% 0.2 97.5%
4. Next.js 43.2% 7.0 93.9%
5. Vue 39.4% 5.2 94.8%
6. Gatsby 36.9% 1.9 91.7%
7. Nuxt.js 24.8% 5.5 84.6%
8. Angular 2+ 19.9% 0.8 95.3%
9. 11ty 17.4% 5.9 60.9%
10. Jekyll 16.5% 0.5 80.8%
11. Angular 1.x 15.3% 0.2 94.4%
12. Hugo 14.8% 1.3 73.6%
13. Svelte 14.4% 5.5 82%
14. Vite 14.2% 9.6 59.3%
15. Preact 10.5% 2.7 74.4%

Let’s look at the next group with under 10% usage. SvelteKit, with huge satisfaction and riding the coattails of Svelte over in the major frameworks group, looks likely to break out by next year.

Bad news for those under 1.0: Dojo, Hexo, Ember, Meteor and Backbone, with small and unhappy user bases, might be frameworks you consider moving away from.

Minor frameworks by usage and satisfaction

Source: Jamstack Community Survey 2020—2021
Show Chart Data
Framework Percentage of respondents who use this Satisfaction score Awareness
1. Nest 8.8% 2.6 59.9%
2. VuePress 8.8% 2.4 62.3%
3. Gridsome 8.5% 1.6 54.2%
4. SvelteKit 8.1% 6.0 58.1%
5. Backbone 7.8% 0.3 76.5%
6. Meteor 6.5% 0.5 69.5%
7. Docusaurus 6.2% 1.9 50.5%
8. Sapper 6.1% 1.2 45.1%
9. Hapi 5.6% 1.3 52.7%
10. Ember 5.3% 0.7 78.2%
11. Stencil 4.3% 1.8 43.8%
12. RedwoodJS 4.3% 1.8 48.9%
13. Clojure 3.6% 1.5 68.1%
14. Hexo 3.4% 0.8 43.2%
15. Blitz.js 3.3% 2.0 50.3%
16. Dojo 2.9% 0.9 51.5%
17. Remix 2.4% 1.4 43.0%

We didn’t ask about every framework this year last year as well, but we did for a lot of them, allowing us to show an annual changes chart.

Frameworks by 1-year change in usage and satisfaction

Source: Jamstack Community Survey 2020—2021
Show Chart Data
CMS Usage change (%) Satisfaction change Usage
1. React 5.7% 0.2 68.1%
2. Express 5.3% -0.4 51.3%
3. jQuery -1.2% 0.1 50.8%
4. Next.js 22.2% 0.9 43.2%
5. Vue 1.9% -0.6 39.4%
6. Gatsby -2.9% -2.5 36.9%
7. Nuxt.js 5.2% 24.8%
8. Angular 2+ 1.8% 0.2 19.9%
9. 11ty 7.0% -7.2 17.4%
10. Jekyll -2.4% -0.1 16.5%
11. Angular 1.x -0.7% 0.1 15.3%
12. Hugo 0.1% -0.4 14.8%
13. Svelte 3.4% -1.4 14.4%
14. Preact 2.9% -4.0 10.5%
15. Nest 4.5% -5.3 8.8%
16. VuePress 1.1% -2.5 8.8%

Here you can see a clear trend mentioned several times. The more your usage grows, the more your satisfaction falls, as your user base expands beyond early adopters. As we saw, 11ty broke into major territory for the first time this year. It has a great satisfaction score, but it’s much lower than last year. This is true, to some degree, for all the growing frameworks. VuePress, Preact, Nest and 11ty form a straight line going down and to the right: the more users they gained, the less happy they were.

There are a few obvious exceptions. Next.js, despite growing enormously, also improved its satisfaction score.

React managed to keep its score steady. No small feat when you’re the biggest framework by a long shot.

Gatsby lost in both usage and satisfaction, an unfortunate outcome for that team.

The top third-party services Jamstack developers use

Jamstack devs are using third-party services and CMSes to make their lives easier.

Jamstack is all about APIs, and many third-party APIs exist to make life easier for developers.

The most popular third-party service is authentication, which is unsurprising—using an existing login reduces signup friction and the security challenges associated with maintaining your own authentication.

CMS services are likewise very popular. Why build a rich text editor when one already exists?

Third party services by usage

Source: Jamstack Community Survey 2020—2021
Show Chart Data
API Most used
User authentication 55%
Headless CMS 53%
Asset/Image management 45%
Headless database 36%
Headless e-commerce 24%

Jamstack has become the standard architecture for the web.

The world is changing, and the Jamstack—and the developers who use it—is changing with it. Based on what you’ve told us this year, we’ve observed shifts within our developer community. What these changes tell us paints a picture of a world in which more people are coming to our community as new developers, often as students; some of the largest web sites and apps are built on the Jamstack; and more and more industries are represented in Jamstack development.

In other words, the Jamstack is touching all industries. We’ve gone mainstream as more developers learn about Jamstack. It’s the new way to build applications and websites, and it’s where the industry is going. Even more, it’s a thriving community that is growing fast as a wave of mainstream adoption continues, driven by fantastic scaling, high performance, and workflows and tooling that developers love.

Thank you for being part of the Jamstack community again this year. Join us and thousands of developers virtually at Jamstack Conf 2021 to learn how Jamstack started, where it’s going, and what’s next for our community.

Learn more:

To review the in-depth methodology, review this document.