Future History
Skip to sources
Time Machine

Web Developers

Scrub through 45years of this role's history, from when it first emerged, through every wave of technology that reshaped it, to the cited projections for where it's heading next.

2026drag to travel through time
2000now
Country
2026
Known today as Web Developers (BLS SOC 15-1254, 2018 SOC)
Latest actual · 2024
86K
BLS OEWS May 2024, sourced from O*NET and the BLS Occupational Outlook Handbook. The 2024 count of approximately 86,000 represents a notable contraction from the 2013 peak of ~112,820. The primary driver is the automation of boilerplate-website creation by AI-assisted site generators (v0 by Vercel, Bolt.new, Lovable, Shopify Magic, 10Web), which compressed multi-day freelance website engagements into tools that non-developers can use themselves. The surviving workforce is concentrated in custom integrations, performance engineering, accessibility compliance, and complex transactional systems that the generators cannot produce reliably. Separately, many practitioners who previously carried the "web developer" title have migrated into the broader Software Developer category (15-1252) as their work expanded into backend systems and DevOps.
Latest actual · 2024
$90,930
BLS OEWS May 2024. The median annual wage of $90,930 ($43.72/hr) represents the highest inflation-adjusted wage level for web developers in the occupational record. Despite falling headcount from the 2013 peak, wages have risen significantly, which is consistent with a bifurcated market: the boilerplate-site segment (lower wages) is contracting while the remaining specialist segment (integrations, performance, accessibility, architecture) sustains and grows compensation. The lowest 10 percent earned less than $48,560; the highest 10 percent earned more than $162,870, reflecting the wide earnings range between a junior CMS theme developer and a senior performance engineer at a major platform.
Each dot is a cited figure over time; the dotted line only links them (values between aren't measured). Hollow dots are estimates.
Tools of the era

The tools that defined the work

Select an era to see how it reshaped the work.

  • WorldWideWeb browser + HTTP/HTML (Berners-Lee, CERN, 1991)

    The first web developers worked with a suite of tools that Berners-Lee had built himself: the WorldWideWeb browser (the only browser in existence, running on NeXT workstations), an HTTP server, and the HTML markup language. There were no text editors designed for HTML, no frameworks, no version control tailored to web work. Developers wrote raw HTML in general-purpose text editors, configured Unix HTTP servers by hand, and tested their work by loading pages in the browser and looking at them. The entire toolchain was about three years old and entirely novel: no patterns existed for how to organise a website, what a web server configuration should look like, or how pages should relate to each other.

    Work toolChanging equipment

  • NCSA Mosaic + early browsers (HTML tables, inline images, 1993)

    NCSA Mosaic, released in January 1993 by Marc Andreessen and Eric Bina at the University of Illinois, was the first browser to display images inline with text. Its April 1993 source-code release made the web accessible on Windows and Mac computers for the first time. For web developers, Mosaic changed everything: suddenly a visual medium existed, and pages needed to look good as well as function correctly. The mid-1990s saw a wave of HTML additions (tables in HTML 3.0, background colors, image maps) that allowed rudimentary visual design on the web. Developers of this era wrote HTML directly in text editors like BBEdit or Emacs and tested pages on multiple browsers whose rendering was wildly inconsistent.

    Effect on the work

    The Mosaic/Netscape era created the webmaster role: a single person responsible for building and maintaining a company's entire web presence. It was the first identifiable job description that would later be called web developer, and it attracted both technical and non-technical workers because the barrier to entry (knowing HTML) was far lower than for traditional programming.

    Work toolChanging equipment

  • JavaScript (Netscape, May 1995) + CSS (W3C, 1996)

    Brendan Eich created JavaScript in ten days in May 1995 at Netscape, originally calling it Mocha, then LiveScript, then JavaScript. It shipped in Netscape Navigator 2.0 in December 1995. For the first time, a web page could respond to user actions without reloading: form validation, dropdown menus, and simple animations became possible. CSS (Cascading Style Sheets) followed from the W3C in December 1996, separating visual presentation from HTML structure. Together, JavaScript and CSS transformed web development from document-authoring into software engineering: pages were now programs, not just text files. The developer who understood both the structural HTML layer and the behavioral JavaScript layer began to emerge as a distinct specialist.

    Effect on the work

    The introduction of JavaScript began a permanent arms race in web developer skill requirements. Webmasters who knew only HTML found their work devalued by employers seeking practitioners who could build interactive experiences. JavaScript also created the first meaningful barrier to entry in web work: writing correct, cross-browser-compatible JavaScript in the mid-1990s was genuinely hard, and the developers who could do it commanded premium salaries during the dot-com boom.

    Work toolChanging equipment

  • Server-side scripting: PHP, ASP, Perl CGI (dynamic web 1999-2008)

    PHP (originally created by Rasmus Lerdorf in 1994 as a set of CGI scripts, released as PHP 3 in 1998) and Microsoft's Active Server Pages (ASP, 1996) made it practical to build database-backed websites without being a systems programmer. A web developer could now write page logic in a scripting language embedded in HTML, query a MySQL or SQL Server database, and return personalised content to every user. The dynamic web era separated web developers into two functional roles that persist today: front-end developers (HTML/CSS/JavaScript visible to the browser) and back-end developers (server-side logic and databases). By 2003, WordPress had launched on PHP and MySQL, giving small-business clients a CMS they could administer themselves and creating a large market for developers who specialised in PHP customisation.

    Effect on the work

    Server-side scripting massively expanded the web developer job market by making e-commerce, content management, and user-account systems buildable by a solo developer or a small agency team, rather than requiring a full software engineering team. The number of web development companies in the US grew from under 1,000 in 1995 to over 30,000 by 2005.

    Work toolChanging equipment

  • AJAX + jQuery (2005-2006), version control with Git and GitHub (2005-2008)

    Jesse James Garrett coined the term AJAX (Asynchronous JavaScript and XML) in February 2005 to describe the technique that Google Maps and Gmail had quietly pioneered: loading data from the server without reloading the whole page. jQuery, released in January 2006 by John Resig, gave every web developer a cross-browser JavaScript library that made DOM manipulation, AJAX calls, and animations practical without the painful cross-browser incompatibilities of raw JavaScript. By 2011, jQuery was used on more than half of all websites on the internet. Separately, Linus Torvalds created Git in April 2005 to manage the Linux kernel; GitHub launched in April 2008 as a hosting platform for Git repositories. Web development teams could now collaborate on code asynchronously, review each other's changes before merging, and roll back to any previous state. Version control ended the era of FTP uploads and "overwriting the live server with your working copy."

    Effect on the work

    AJAX and jQuery raised the median quality of interactive web applications dramatically and created the "front-end developer" as a distinct professional identity from the full-stack webmaster. GitHub transformed web development from a craft practised in isolation into a collaborative, open-source discipline: the best patterns and libraries became public goods that any developer could adopt, accelerating the pace at which practitioner standards rose.

    Work toolChanging equipment

  • Node.js + npm (2009), responsive design (2010), React and single-page apps (2013)

    Node.js, released by Ryan Dahl in May 2009, brought JavaScript to the server for the first time, enabling web developers to write both front-end and back-end code in the same language. npm (the Node Package Manager), which launched alongside Node, created a shared library ecosystem that grew to over one million published packages by 2019, dramatically accelerating how fast a developer could assemble a working application. Ethan Marcotte's 2010 article "Responsive Web Design" in A List Apart established the design pattern for sites that adapt to any screen size, a necessity as mobile devices became the primary web access point for hundreds of millions of users. React, released by Facebook in May 2013, popularised the component-based UI model and the virtual DOM, making large, stateful single-page applications (SPAs) practical. By 2015, the JavaScript ecosystem had become so extensive and fast-moving that "JavaScript fatigue" entered the developer lexicon.

    Effect on the work

    The Node/React era bifurcated the profession into specialists: a React front-end developer working with Redux and component libraries, a Node back-end developer managing REST APIs and databases, and a DevOps engineer handling Docker containers and CI/CD pipelines. Salaries rose significantly for those who mastered the new stack, but the fragmentation of tooling also created a skills-half-life problem: the specific JavaScript framework a developer mastered in 2013 might be obsolete by 2016.

    Work toolChanging equipment

  • Cloud deployment, serverless, and Jamstack (Netlify 2015, Vercel 2015, AWS Lambda 2014)

    The Jamstack architecture (JavaScript + APIs + Markup, coined by Netlify co-founder Mathias Biilmann around 2015-2016) decoupled the build process from the hosting environment and pushed deployment to the edge. Platforms like Netlify and Vercel let a web developer push code to a Git repository and have a production deployment within seconds, with global CDN distribution handled automatically. AWS Lambda (2014) made it possible to run backend logic without provisioning or managing a server. For the web developer, this stack eliminated one of the most friction-laden parts of the job (managing and paying for servers, configuring web hosts, deploying via FTP). It also created a new tier of professionals who understood cloud architecture well enough to wire together managed services (Stripe, Auth0, Supabase, Contentful) rather than building infrastructure from scratch.

    Effect on the work

    Cloud-deployment platforms simultaneously lowered the barrier to shipping a web application (no server management) and raised the value of developers who understood the full cloud services landscape. The "full-stack developer" of 2020 was expected to know not just HTML, CSS, and JavaScript but also CI/CD pipelines, edge functions, CDN configuration, and managed database services. The rise of these platforms also began the long-term substitution of self-hosted CMS installations (WordPress, Drupal) with API-driven headless architectures, reshaping the client base of the typical agency web developer.

    Work toolChanging equipment

  • AI code generation: GitHub Copilot (2021), Cursor (2023), v0/Bolt.new/Lovable (2023-2024)

    GitHub Copilot, released as a technical preview in June 2021 and generally available in June 2022, was the first AI coding tool adopted at professional scale: by the Stack Overflow Developer Survey 2025, 84% of developers reported using AI coding tools, with Copilot among the most common. For web developers, the more disruptive tools arrived in 2023-2024: v0 by Vercel (October 2023), Bolt.new (October 2024), and Lovable generated complete, deployable web applications from a plain-English description in minutes. Lovable reached $400 million ARR by February 2026 and raised $330 million at a $6.6 billion valuation. v0 by Vercel had 6 million developers on the platform with 9.6 million projects created in 2025 alone. Bolt.new hit $40 million ARR within five months of launch. These tools automated the exact work that had employed the largest segment of the freelance and agency web developer workforce: small-business marketing sites, landing pages, and simple e-commerce storefronts. The AI-generation era is the first technology in the web development timeline to directly reduce demand for the occupation rather than augmenting practitioner productivity.

    Effect on the work

    BLS OEWS data shows web developer employment declining from approximately 112,820 (2013) to 86,000 (2024), a 24% contraction. The strongest drivers are AI site generators automating the boilerplate-website segment plus the reclassification of many practitioners into the Software Developers category (15-1252) as their work expanded. Despite falling headcount, median wages rose to $90,930 in 2024, indicating that the remaining workforce is concentrated in higher-value specialist work.

    AI audit toolsPattern detection
Projection cone · present → 2034

What credible sources project

Scrub the slider past now to anchor each scenario on the scrubber. The spread is the range of futures credible sources project for this role.

Employment outlook
Projected change in the number of people doing this work.
BLS National Employment Matrix 2024-34
→ 2034
+7.5%
BLS National Employment Matrix projects 15-1254 Web Developers from 86,000 (2024) to approximately 92,500 (2034), a 7.5% increase. The matrix-level projection is slightly more precise than the OOH headline figure and reflects the same demand drivers: e-commerce expansion, mobile web growth, and demand for accessible, performant web applications. The projection does not separately model displacement by AI site generators, which the BLS classifies as a productivity-enhancing tool for remaining practitioners rather than a displacement event for the occupation as a whole. This distinction may prove analytically important if the generators continue to reduce the practitioner count needed per project delivered.
BLS Occupational Outlook Handbook 2024-34
→ 2034
+7%
BLS projects overall employment of web developers and digital designers to grow 7 percent from 2024 to 2034, much faster than the average for all occupations (3%). About 5,400 annual openings are projected for web developers (15-1254) specifically; the broader combined category (15-1254 plus 15-1255 digital designers) projects 14,500 annual openings. The BLS methodology models continued e-commerce growth and mobile web demand as the primary tailwinds. The projection baseline is 86,000 web developers employed in 2024, implying approximately 92,000 by 2034. This is a notably optimistic projection given the concurrent decline in employment from 2018 to 2024; the BLS model does not fully capture the AI site-generator displacement risk within the 2024-34 window. The projection is most credible for the specialist segment of the occupation (accessibility engineering, complex integrations, performance optimization) and least credible for the boilerplate-website-build segment that AI generators have most directly automated.
AI task exposure
Share of the role’s tasks that researchers estimate AI can do. This is a measure of task exposure, not a forecast of jobs lost.
Eloundou et al. — "GPTs are GPTs" (2023)
→ 2028
80%
of tasks
GPT-4 task-by-task LLM exposure labeling across O*NET tasks for Computer and Mathematical Occupations. Web developers score among the highest-exposure occupations in the Eloundou framework: writing code, reviewing code, generating markup, and composing technical documentation are all tasks LLMs execute fluently. The 80% exposure figure reflects the fraction of web developer tasks for which LLM-powered tools (code completers, code generators, documentation writers) can provide substantial assistance at equivalent quality. Critically, Eloundou measures LLM-specific exposure, not headcount displacement. The correct interpretation is that 80% of a web developer's task surface is addressable by LLM tools as assistants or partial substitutes, not that 80% of web developer jobs will disappear. The actual employment effect depends on whether the task-level productivity gain translates into hiring fewer developers per project (substitution) or building more projects per team (expansion), and in the 2021-2026 record, both effects are visible simultaneously in different market segments.
Goldman Sachs Research — "Generative AI could raise global GDP by 7%" (2023)
→ 2030
30%
of tasks
Goldman Sachs estimates that approximately 30% of software development work is addressable by generative AI as of 2023-2026, with the highest near-term exposure in junior-role tasks: writing boilerplate code from specifications, generating standard UI components, and producing documentation. This figure is consistent with the Eloundou task-exposure framework for the specific sub-segment of coding tasks. Goldman distinguishes between tasks exposed (automatable in principle) and work eliminated (actually displaced), noting that developer demand has been expanding faster than AI has contracted it at the aggregate level, though this aggregate masks the sharp contraction in the boilerplate-website and junior-developer segments.
Today, in this role

What's shifting in the work right now

The historical view above shows how this role has moved. This is the present-day detail: which AI tools are picking up which tasks, where the edge still is, and the natural directions this work can grow.

What's changing in your day

Three parts of your work where AI is already doing real lifting, and what stays yours.

AI is taking this onBuild and deliver small-business marketing sites, landing pages, and portfolio sites — work that Bolt.new, Lovable, and v0 now generate end-to-end from a plain-English prompt in 10-30 minutes, compressing what was once a multi-day freelance engagement into an automated scaffold the client can self-serve.

Build and deliver small-business marketing sites, landing pages, and portfolio sites — work that Bolt.new, Lovable, and v0 now generate end-to-end from a plain-English prompt in 10-30 minutes, compressing what was once a multi-day freelance engagement into an automated scaffold the client can self-serve.[6],[5],[4]

Where your edge is

Stop competing on site creation speed; position on what the generators cannot deliver — bespoke integrations (payment flows, CRM sync, inventory APIs), accessibility compliance, and performance audits. Charge for outcomes, not hours of HTML.

AI is sitting alongside you herePrototype and iterate on new UI features by prompting v0 or Bolt.new for an initial scaffold, then extending the AI-generated component with custom logic, state management, and API wiring inside Cursor or Windsurf — compressing what was a multi-day spike into a same-day working demo.

Prototype and iterate on new UI features by prompting v0 or Bolt.new for an initial scaffold, then extending the AI-generated component with custom logic, state management, and API wiring inside Cursor or Windsurf — compressing what was a multi-day spike into a same-day working demo.[4],[11]

Where your edge is

Treat AI-generated prototypes as throw-away scaffolding: own the architectural decision of whether to refactor the AI output or rewrite with proper component structure; the judgment call on when generated code is fit for production is permanently human.

AI is sitting alongside you hereWrite, review, and maintain React / Vue / Svelte component code using AI pair-programming editors (Cursor, Windsurf, GitHub Copilot), accepting or rejecting AI-generated completions based on correctness, performance, and alignment with the project's design system.

Write, review, and maintain React / Vue / Svelte component code using AI pair-programming editors (Cursor, Windsurf, GitHub Copilot), accepting or rejecting AI-generated completions based on correctness, performance, and alignment with the project's design system.[2],[12]

Where your edge is

Develop a systematic review practice for AI-generated frontend code: run Lighthouse and axe-core on every AI-authored component, enforce design-token compliance, and build the judgment to reject visually correct but semantically broken output.

Where this role is heading

Natural next steps for someone with your foundation: not exits, evolutions.

A direction you could grow

Software Developers

Web Developers already write production JavaScript and manage deployment pipelines; the pivot adds backend systems, algorithms, and computer science depth. The 2025 job market shows companies that needed 10 Web Devs now hire 4 Software Developers with AI tools — but those 4 command significantly higher compensation. McKinsey finds the strongest demand for engineers who can architect systems and review AI output, which requires the broader SWE skill set.

What you'd add
· Data structures and algorithms (LeetCode-style proficiency for interviews)
· Backend language depth: Node.js/TypeScript server-side, Go, or Python
· Cloud platforms: AWS or GCP beyond basic hosting (IAM, VPC, managed databases)
What it takesSome new skills to pick up
Share this year
Drops anyone you send it to straight into 2026.
Preview card
Part of Tech & Data · see all 28roles →
Different role?

See the same long-arc view for your own profession.

Browse the directory by industry, or search by title or SOC code. New roles ship every few weeks. Every profile cites every claim.

Browse all roles

The data behind this timeline

On record since1991
Latest tracked employment86,000 (US, 2024)
Latest median pay$90,930 (2024)
Outlook+7% by 2034 (BLS Occupational Outlook Handbook 2024-34)
View all 7 cited data points
YearUS employmentMedian annual paySource
19955,000n/aESTIMATE
199975,000$40,800ESTIMATE
2013112,820$67,540BLS-OEWS
202184,820$77,030BLS-OEWS
202288,620$78,580BLS-OEWS
202385,350$84,960BLS-OEWS
202486,000$90,930BLS-OEWS
Embed this timeline on your site

Free for any site. Paste this where the timeline should appear; it stays interactive, every datapoint stays cited, and it sets no cookies on your page. How embedding works

<iframe src="https://futurehistory.earth/embed/15-1254"
  width="100%" height="430" style="border:0"
  title="Web Developers, a Future History timeline"
  loading="lazy"></iframe>

See all roles in Tech & Data