TL;DR: Bereket Engida, a self-taught Ethiopian developer, built Better Auth from his bedroom in Addis Ababa and sold it to Vercel roughly thirteen months after raising a $5 million seed round. The library's weekly npm downloads grew from about 150,000 in mid-2025 to more than 4.7 million by the acquisition, with over 850 open-source contributors. No public revenue figure drove the deal. Adoption did. The lesson for solo founders and owner-operators building software: distribution metrics, not revenue projections, are what a strategic acquirer pays for when the asset is infrastructure.

I read the Better Auth story twice before I believed the numbers. Not because the acquisition seemed implausible. Because the speed did. Thirteen months from seed to exit is not a normal cadence in software, and understanding why it happened this fast tells you more about how modern acquisitions get priced than almost any deal I have studied this year.

I once advised a founder who had the opposite problem. Strong resume, an MBA from a name-brand school, two prior stints at recognizable companies on his LinkedIn profile, and a product that maybe 400 people used actively after eighteen months in market. He kept asking why investors were not returning his calls despite his background. I told him the truth nobody wanted to say in his pitch prep sessions: nobody funds a resume. They fund a curve. His curve was flat. Engida's curve, with zero name-brand credentials behind it, went from 150,000 to 4.7 million in thirteen months, and that curve is the entire reason his calls got returned before he even finished dialing.

The Timeline, Compressed

Engida taught himself to code at 18 in Addis Ababa. According to The New Stack's reporting on the acquisition, he started Better Auth while building an unrelated open-source analytics project and hitting a wall: no existing authentication library handled multi-tenant team accounts and permission levels the way he needed. He spent roughly seven months building a framework-agnostic alternative from scratch and published the first version to GitHub in September 2024.

The project moved through Y Combinator and raised a $5 million seed round in mid-2025 from Peak XV Partners, Y Combinator, P1 Ventures, and Chapter One, according to African-Startups' coverage of the deal. At the time of the raise, weekly npm downloads sat around 150,000, with roughly 15,000 GitHub stars and a Discord community above 6,000 members. Those are strong open-source metrics for a young project, but not yet the numbers of a category-defining piece of infrastructure.

Thirteen months later, in July 2026, Vercel announced the acquisition outright. Weekly downloads had grown to more than 4.7 million. Contributors had grown past 850. Terms were not disclosed, but the growth curve between seed and exit is the real story: roughly a thirty-one-fold increase in weekly downloads in just over a year.

Why Vercel Bought Adoption, Not Revenue

Better Auth is open source. Its business model was never built around a large recurring revenue line the way a typical SaaS acquisition target's would be. So what did Vercel actually pay for?

The New Stack and The Stack's coverage both converge on the same answer: Vercel bought a foothold in agent identity, the emerging problem of giving AI agents their own scoped, revocable credentials distinct from the humans who deploy them. Better Auth had already begun building Agent Auth, a protocol addressing exactly this problem, before the acquisition closed. Vercel CEO Guillermo Rauch framed the deal publicly as an extension of Vercel's "open by default, loosely coupled, portable" philosophy, the same approach applied previously to Next.js and the Vercel AI SDK.

That framing matters, but the download curve is what made the deal happen on this timeline. A protocol with an interesting idea and 150,000 weekly downloads is a research bet. A protocol with the same idea and 4.7 million weekly downloads, integrated into hundreds of thousands of production codebases, is infrastructure. Acquirers do not pay infrastructure prices for research bets. They pay infrastructure prices once the distribution proves the bet already won.

The Owner's Exit Engine Applied to Open Source

I built the Owner's Exit Engine around a simple premise: buyers pay for what reduces their risk of having to build something themselves, on their own timeline, at their own cost. Applied to Engida's exit, the framework explains the speed cleanly.

Revenue tells a buyer what a business is worth today. Distribution tells a buyer what a business will cost them to replace if they do not acquire it. Vercel could have built its own authentication and agent-identity layer internally. Doing so would have taken a comparable team a year or more, with no guarantee of matching the community trust and integration depth Better Auth had already built across 850-plus contributors and however many thousands of downstream applications generate 4.7 million weekly pulls. Buying Better Auth was cheaper and faster than rebuilding it, and the download curve is what made that math obvious to Vercel's leadership without a lengthy negotiation over projected future revenue that did not exist yet in any meaningful form.

This is the exit engine running on a capital-light asset. Engida never had to build a large team, a sales organization, or an enterprise contract book. He built something developers adopted voluntarily, at scale, because it solved a real problem better than the alternatives. That adoption curve became the asset. Not the revenue statement.

Competence Beats Credentials, Measured in Downloads

I want to sit on Engida's background for a moment, because it is not incidental to the lesson. He is self-taught. He learned to code at 18, in Addis Ababa, without a computer science degree from a recognized program, without a Silicon Valley network, without the standard credential set that venture investors are often accused of screening for before they screen for anything else.

None of that mattered once the download numbers spoke. Peak XV Partners and Y Combinator did not fund Better Auth because of Engida's resume. They funded it because 150,000 weekly downloads of a solo-built library, achieved with essentially no marketing budget, is a signal that is very hard to fake and very easy to verify. Vercel did not acquire the company because of a pedigree. It acquired the company because the adoption curve, an objective, third-party-verifiable metric, kept climbing every month for over a year straight.

This is the cleanest recent example I have seen of a principle I repeat constantly to owner-operators who worry they lack the right background to compete: the market does not check your credentials before it checks your output. Ship something that solves a real problem, publish the adoption numbers where anyone can verify them, and let the market do the credentialing for you. Engida did not need anyone's permission to become the person Vercel wanted to acquire. He needed 4.7 million people a week choosing his library over the alternatives.

What Solo Founders Should Actually Take From This

The temptation with a story this clean is to draw the wrong lesson: build something open source, watch it go viral, get acquired. That is survivorship bias dressed up as a playbook. Thousands of open-source libraries never cross 10,000 weekly downloads, let alone 4.7 million. What is repeatable is not the outcome. It is the sequence of decisions that made the outcome possible.

First, Engida built something in response to a specific, personally felt problem, not a speculative market gap he read about in a trend report. Second, he shipped a working version fast, in around seven months, rather than spending years perfecting an unreleased architecture. Third, he stayed capital-light through the earliest growth phase, which meant the eventual $5 million seed round funded acceleration of something already proving out, not survival of something still unproven. Fourth, once he had traction, he built toward a problem, agent identity, that was becoming newly urgent across the entire industry at exactly the moment his distribution numbers made him a credible, low-risk partner to solve it with.

Sequence matters more than any individual decision. A founder who raises capital before proving distribution is betting the market will validate a thesis the founder has not yet tested. Engida proved distribution first, on his own capital and his own time, and only then took outside money to accelerate what was already working.

The Part of the Story Most Coverage Skips

Better Auth did not grow to 4.7 million weekly downloads by accident, and it did not happen purely because the code was good. Good code with zero distribution sits at a few hundred downloads a week forever. What changed the trajectory was a specific, deliberate decision in late 2025: the Better Auth team took over stewardship of Auth.js, formerly known as NextAuth.js, after its existing maintainers concluded the two projects were solving the same problem and heading in the same direction.

That stewardship transfer is easy to miss in the acquisition headlines, but it is arguably the single highest-leverage move in the entire story. Auth.js already had an enormous existing install base built over years of prior maintainer work. Absorbing that stewardship did not just add a feature. It merged two adoption curves into one, and the combined curve is what actually produced the 4.7 million weekly download figure Vercel was looking at when it decided to acquire rather than build. A solo founder rarely gets to 4.7 million weekly downloads purely through original code adoption. Consolidating adjacent open-source projects under one roof, when the technical direction genuinely aligns, is a distribution strategy in its own right, and it is one every founder building an open-source or developer-tool asset should study closely.

Doctrine Connection: Competence Beats Credentials

Nobody at Peak XV Partners asked where Bereket Engida went to school. Nobody at Vercel asked either, by the time the acquisition closed. What they asked was whether the download curve was real, whether the contributor base was genuine, and whether the technical problem Better Auth was solving mattered enough to the next five years of software infrastructure to justify an acquisition instead of an internal build. Every one of those questions is answerable through output, not pedigree. That is the entire doctrine in one transaction: build something real, let the numbers speak in public, and let the market sort out who gets taken seriously. The market sorted Engida into the group that gets acquired by the company whose framework inspired him in the first place.

FAQ

How much did Vercel pay for Better Auth?
Terms of the acquisition were not publicly disclosed. Better Auth had previously raised a $5 million seed round in mid-2025 from Peak XV Partners, Y Combinator, P1 Ventures, and Chapter One.

What made Better Auth attractive to an acquirer without significant revenue?
Distribution. Weekly npm downloads grew from roughly 150,000 at the time of the seed round to more than 4.7 million by the time of the acquisition, with over 850 open-source contributors. That adoption curve signaled infrastructure-level trust that would have been costly and slow for Vercel to build internally.

What is Agent Auth, and why did it matter to the deal?
Agent Auth is a protocol Better Auth was developing to give AI agents their own scoped, revocable identity, distinct from the human users who deploy them. Vercel framed the acquisition as a bet on agent identity becoming foundational infrastructure as AI agents proliferate across software products.

How long did it take Better Auth to go from launch to acquisition?
Engida published the first version to GitHub in September 2024. The company raised its $5 million seed round in mid-2025, and Vercel announced the acquisition in July 2026, roughly thirteen months after the seed round closed and under two years after the initial GitHub release.

Jeff Barnes, MBA has no personal position in any company, fund, or platform named in this article. demg.ai has no current commercial relationship with any party mentioned. demg.ai provides marketing strategy and education for owner-operators, not investment advice.