How Vercel Turned Shipping Into Its Brand

Developer-tool companies have a weird problem. The products that often earn them the most attention are the ones they give away. A framework, SDK, or open standard can reach millions of people before its creator earns a dollar from it. Adoption depends on developers trusting that the project will remain useful and responsive to the community.

And after they succeed, the company must figure out how to make money.

Vercel and Netlify show two ways to solve this problem. They entered the same market at nearly the same time with very similar products. Netlify was founded in 2014 and launched from private beta in March 2015. Guillermo Rauch started ZEIT, which would become Vercel, in November 2015 and released its first product in April 2016.

I use both for my projects. Mostly Vercel these days, though this blog is hosted on Netlify. Both made it easy to connect a repository, preview changes, and publish globally.

For years, Netlify was bigger. The 2022 Web Almanac still detected Netlify on more Jamstack sites than Vercel. Today, public signals show much stronger financial momentum at Vercel. Its September 2025 funding round valued it at $9.3 billion. TechCrunch reported a $340 million annual revenue run rate by February 2026. By July, the company was handling six million deployments a day, half initiated by coding agents. Meanwhile, Netlify said it had crossed seven million developers in June 2025.

Communications alone rarely explain the difference in outcomes. I doubt it’s the case here. But I can’t stop wondering if Netlify’s approach to creating and promoting the Jamstack concept affected their chances negatively. While Netlify created demand for an ecosystem that many companies could monetize, Vercel focused on one specific project and made it the gateway into their own commercial platform.

I. The two companies

Mathias Biilmann built BitBalloon in 2013 after noticing that Webpop customers wanted to publish simple landing pages with working contact forms. BitBalloon hosted the static files and processed form submissions itself, so customers didn’t have to manage a server.

Biilmann wanted to build the smallest possible tool. A user would drag a folder with a website into the browser and receive a live URL. He opened the beta on August 27, 2013, and posted it to Hacker News as “BitBalloon, insta-deployment of static sites with forms that just work.” The pitch in four words: drag, drop, live website. That Hacker News post reached the front page and collected 24 points before a false positive in the site’s voting-ring detector pushed it down.

Then, together with his friend Christian Bach, Biilmann founded Netlify in 2014. Bach saw a potential to create a new architecture: make the frontend an independent system, built in advance and distributed globally, and abstract away the need to manage a server through APIs. Netlify returned to Hacker News in March 2015 with a post titled “Builds, Deploys and Hosts Your Static Site or App.” This time, it drew 116 points and 41 comments.

The team called this architecture JAMstack (originally JavaScript, APIs, and markup). “Static site” could sound like a plain website, a set of HTML pages with no interactivity. JAMstack described applications that could be highly dynamic while avoiding the need to deal with a traditional web server.

By 2021, Netlify had a working product, a vision of the future, a term widely-used by their ecosystem, and $105 million in new funding at a $2 billion valuation.

Vercel entered the same market through a slightly different door.

II. One command and a URL

Guillermo Rauch had spent years building an audience among JavaScript developers. He contributed to MooTools, created Socket.IO, co-founded LearnBoost, and started Cloudup, a fast file-sharing service. Users dragged in images, videos, documents, or links and immediately received a shareable URL. Automattic bought it to improve WordPress media uploads and collaborative editing.

Rauch later tried to deploy a React application on Kubernetes and encountered the opposite experience: infrastructure, certificates, configuration, networking, and terminology stood between the code and a working product.

So in November 2015, Rauch started ZEIT with engineers Tony Kovanen and Naoyuki Kanezawa. In April 2016 they released now. After installing the command-line tool, a developer could open a project and type one word:

now

ZEIT took the project, built it in the cloud, and returned a secure URL. Rauch’s 2016 retrospective described their mission as making application deployment “simple, instant and global.”

That approach became the cornerstone of Vercel’s comms. Compress a technical process into a clear action, give it a short name, and make the result instantly shareable.

Rauch’s reputation in the open source community brought developers to the new company. ZEIT also released Hyper, an open-source terminal that reached more than 300,000 downloads and 13,000 GitHub stars by the end of 2016. Free tools attracted the people ZEIT wanted as customers.

But now did not give ZEIT a decisive advantage over Netlify. The breakthrough came from a tool the company initially built for itself.

III. Next.js becomes a distribution engine

ZEIT created Next.js, a React framework for building full-stack web applications, for its own website and dashboard, then open-sourced it in October 2016. The company again focused on the aspect of simplification: developers could start building apps immediately instead of spending days building the pipes and connecting various packages.

By the release of Next.js 2.0 five months later, ZEIT said the original announcement had been read 3.1 million times. The project had more than 10,000 GitHub stars, 110 contributors, and a public Slack community. Its Hacker News discussion drew 618 points and 162 comments. InfoQ covered the framework six days later.

ZEIT’s launch post explained the framework’s principles, showed the installation steps, and pointed readers to a README where they could learn the rest. Later, they built an entire ecosystem around Next.js with a dedicated website, courses, documentation, release events, and community spaces. Through this, ZEIT turned this open-source project into a recurring media channel. Each release, RFC, tutorial, benchmark, contributor, and customer project put ZEIT in front of developers.

Next.js became a free distribution engine for their paid platform. While you could have used the framework outside of it, every release, tutorial, and contributor increased the likelihood of converting a developer onto Vercel.

It also gave journalists a stable explanation for the company. Vercel’s own funding posts would later talk about collaboration, workflow, and an end-to-end platform. Outside coverage kept identifying it as the company behind Next.js. VentureBeat framed its 2020 Series B as money to advance the React framework; TechCrunch used Next.js to explain the $102 million Series C. That one free project had become Vercel’s media credential.

But such a concentration also created recurring concerns about lock-in and public claims that Next.js didn’t work equally well on other platforms.

IV. Netlify as the category steward

While Vercel concentrated around the framework it owned, Netlify chose to focus on the ecosystem. It wanted to be the best platform for any framework and sold developers on flexibility and choice and deliberately created a category larger than itself.

Jamstack was essentially a single name for a set of tools and practices. Netlify then promoted the term through their Jamstack Conf, podcast, meetups, open-source projects, and even an investment fund. Companies such as Gatsby, Contentful, Sanity, and Algolia were built in this category and benefited tremendously from it.

Netlify also produced the State of the Jamstack survey that helped them understand the market and the community. When the company proposed Distributed Persistent Rendering, it published an RFC and invited framework authors to help shape an approach that could work across the ecosystem.

That is how a company should communicate an open standard: define the problem, create a shared vocabulary, publish evidence, invite other players, and give them some control over the final result. The company gains authority by running coordination, then can convert that awareness and influence into leads and paid users.

When Netlify raised $2.1 million in 2016, its own announcement said the JAMstack should power most sites and applications. TechCrunch described a service for quickly rolling out static websites.

Netlify found a better hook by focusing on the “enemy”. Its 2018 Series B promised to “replace web servers”. Two years later, its Series C announcement paired the vision with 800,000 developers and enterprise growth. TechCrunch repeated the conflict: Netlify wanted to “kill the web server”. A 2018 WIRED feature translated this into a directly visible result: making the internet load faster.

The $105 million Series D in 2021 relied on three pillars: their OneGraph acquisition, a $10 million fund for ecosystem startups, and $1 million for open source development. Netlify’s communications lead unpacked the announcement on LinkedIn into those three commitments. TechCrunch still led with “Jamstack-focused web development.” Netlify had succeeded in making the category famous, but the category kept defining the company in the press. That was becoming somewhat of a problem.

Now, Jamstack became associated with static sites even as Netlify’s product grew more capable. In 2022, Netlify changed the definition to emphasize a broader architectural approach. In 2023, it repositioned again around the “Composable Web Platform.”

Changing the definition so often didn’t help. “Composable” accurately described the capabilities but sounded like language intended for an enterprise architecture presentation. The New Stack wrote a story titled “Is Jamstack Toast?”, which captured the communications problem.

Vercel took a narrower route. It spent less time defining how the whole web should be assembled and focused on what a customer wanted to accomplish: ship a product, quickly.

V. Turning Vercel into a brand

In April 2020, ZEIT became Vercel. The company paired the rebrand with a $21 million Series A and a new focus on the workflow of teams rather than individual developers.

The name came from Lexicon Branding, the naming company behind brands such as PowerBook, BlackBerry, and Pentium. Its brief was to create something short, globally resonant, and easy for a developer to type as a command. Lexicon constructed “Vercel” from associations with versatile, accelerate, and excel. Lexicon founder David Placek later described an extensive process where they generated 60 promising candidates and repeated the exercise three times over two weeks before ending up on Vercel.

The company kept its simplified triangle logo, folded the now product into the Vercel name, and began migrating the identity. It even published a Rebrand Field Book (that’s no longer available).

While ZEIT had started by helping an individual developer publish an application, Vercel now promised the “Develop. Preview. Ship.” workflow for entire teams. Business Insider provided the largest launch piece, interviewing Rauch and publishing the pitch deck behind the $21 million round (getting a BI piece by sharing a pitch deck with them has always been this unusual option). The Register, t3n, and SD Times also covered the rebrand, while a Hacker News discussion drew 190 comments. Much of the attention went to the new name rather than the new team workflow though.

The most valuable part of the rebrand was probably the word “ship.” It described the desired outcome without tying Vercel to a specific technology. Vercel could use it for a website, an application, an AI feature, or an agent. Ship became the company conference. Launches, customer stories, founder interviews, and product pages all kept repeating the same verb.

Vercel used funding rounds as state-of-the-company launches (as most companies do). The $40 million Series B told stories about how Airbnb, TripAdvisor, and HashiCorp used Vercel, showing that it had truly moved from a solo-developer utility to a team workflow. The $102 million Series C led with increases in Next.js downloads, deployments, enterprise customers, and network traffic. The $150 million Series D expanded the claim to an end-to-end platform and connected the money to open-source support, edge infrastructure, and collaboration.

Each announcement gave reporters all possible hooks: financials, product evidence, customer names, and an updated vision. On social media, Vercel often compressed these into a single line. For their Series C, they just posted “we announce $102M in new funding to continue building the next Web, together.”

VI. Founders as distribution

I don’t think it is entirely on me that I knew Rauch and could not name the person leading Netlify.

Netlify’s CEO Biilmann has said that the company intentionally made Jamstack inclusive and larger than Netlify, and warned against putting one company at the center of a new category. But that approach weakened the association between it and the company.

What it meant is Biilmann spoke to people already interested in the category. On the other hand, Rauch had a large open-source audience, then used Next.js and Vercel to increase his reach through announcements, social posts, Ship keynotes, demos, and long podcast appearances.

Relying on the founder’s own image can also be a negative. In September 2025, Rauch posted a photograph from a meeting with Israeli PM Benjamin Netanyahu. Some developers and companies called for a boycott or said they were moving away from Vercel. Days later, Rauch had to address the episode in a Bloomberg interview. He said the backlash had cost Vercel no customers or employees but apologized to anyone hurt by the situation.

VII. Lessons for devtools and open-source companies

• Creating a category cuts both ways. Name one (Jamstack) and you become its authority through conferences, surveys, RFCs, etc. But you don’t own the word. It can outgrow you, stick to what you’ve left behind (static sites), and force endless redefinition.
• Build a free product with unlimited demand, then publish through it. Next.js was free, useful, and its own thing, so demand had no ceiling. Every release, RFC, tutorial, and benchmark put Vercel in front of developers and reporters. Stars and downloads showed momentum before revenue.
• Build proof before naming the trend. Vercel shipped the AI SDK and v0 first, named AI Cloud only after the demand was there. Netlify ran it backwards, named Agent Experience before the evidence.
• Make the founder a distribution channel but manage the exposure. Rauch’s audience carried every launch while Netlify’s CEO stayed largely unknown outside the category. Personal reach compounds distribution but also concentrates risk when something goes wrong.
• Turn milestones into full-kit launches. Vercel ran funding rounds as company-wide moments packed with financials, product features, named customers, and their vision. Oldie but good.

VIII. Is Vercel free and does it matter?

“Every vibe coder uses Vercel” is an exaggeration but hardly a falsehood. Developers use it, along with designers, founders, product managers, and people trying an idea over a weekend. Coding agents now initiate a large share of its deployments. Vercel has become a default destination for a much wider group of software creators.

Vercel’s Hobby plan lets an individual deploy personal, non-commercial projects for free. The company talks about millions of developers creating on the platform; in 2025, it said it had more than 100,000 paying teams. Most people in Vercel’s orbit do not pay it directly. Revenue comes from a far smaller set of teams and high-usage projects.

Netlify also has a free plan, but Vercel attached free deployment to several powerful distribution loops: Next.js, templates, v0, AI SDK, and coding agents.

Meanwhile, Vercel earns all their money from the projects and teams that need more. Pro subscriptions, enterprise contracts, bandwidth, requests, and compute convert a minority of users into revenue. A project becomes valuable to Vercel when it attracts a team or consumes serious resources.

But that success immediately gives its owner a reason to compare infrastructure costs or move elsewhere. Hobby accounts pause when they reach the included limits. After a user moves to Pro, which starts at $20 and allows metered usage, traffic and mistakes can raise the bill.

In 2023, a user reported that a small project had accumulated more than $3,000 in Vercel charges after an unexpected process kept running. Rauch joined the Reddit thread, apologized, got on a Zoom call with the user, refunded the overages, published the cause, and said Vercel would deploy a fix across frameworks. The company later added spend controls that pause production deployments. This is an effective crisis response: Rauch accepted responsibility, explained the problem, refunded the customer, and initiated a product change. But he can’t do that for every customer because racking up the bill is the business model.

They also have the long-term problem of Cloudflare. The company began by putting its network between websites and their visitors to make sites faster and block attacks. It now sells CDN, security, DNS, storage, and computing services to everyone from hobbyists to large enterprises. Unlike Vercel and Netlify, Cloudflare is a public company with a much larger financial base. It generated $2.17 billion in revenue in 2025. After making $696 million in the second quarter of 2026, it forecast nearly $2.9 billion for the year and held $4.16 billion in cash and investments.

This scale gives Cloudflare the money and several ways to monetize a user who deploys an application for free. Cloudflare Workers now covers much of the same job as Vercel: building and hosting full-stack applications, generating previews, attaching storage, and deploying through coding agents. Its free plan includes 100,000 requests a day while the paid plan starts at $5 a month and does not charge for data transfer or bandwidth. Successful free projects on Cloudflare have far fewer reasons to leave.

Cloudflare also has its own distribution: 25% of all websites use its tools, such as the CDN. Cloudflare told investors that it ended 2025 with more than 4.5 million active human developers. It has developer mindshare but also the security and infrastructure teams that can manage everything.

Cloudflare has already attacked Vercel’s strongest distribution channels. It developed an OpenNext adapter that lets Next.js applications run on Workers. In 2026, Next.js shipped a stable adapter API, built with Cloudflare, Netlify, OpenNext, and other cloud providers, which makes it easier for them to match Vercel’s support. Cloudflare also released vinext, an open-source attempt to rebuild Next.js on Vite, and brought the companies behind Astro and Vite into Cloudflare while promising to keep their projects portable.

And their comms have caught up to the product. Cloudflare now tells developers to connect a coding agent and deploy directly from an editor or terminal. It publishes a complete reference architecture for vibe-coding platforms, presenting its network as the place where AI-generated applications can be created and hosted. It even provides a specific guide for moving projects from Vercel to Workers.

Netlify challenged Vercel for roughly the same users and workflow. Cloudflare approaches from the network layer, where it already has extraordinary reach, and can earn on providing security, traffic, storage, and compute. As agents make more deployment decisions, developers might want to prioritize price predictability, compatibility, and agent-readable documentation over the human experience.

Free and easy is no longer a moat.