- 01Turn your own bug reports into public content
- 02Ship one engineering deep-dive to Hacker News every month
- 03Stagger your Product Hunt and Show HN launches instead of stacking them same-day
- 04Recruit from your own traction feed
01 Snapshot
The one-glance read on who they are and how they grow. Each point is verifiable from the receipts above.
An open-source (MIT) browser automation library that lets AI agents click, type, and navigate the web, plus a paid Browser Use Cloud hosting layer, sold to developers building agents.
Launch-led open-source virality carried forward by founder-led build-in-public and a recurring Hacker News engineering-blog cadence.
~412K monthly visits, over 105,000 GitHub stars as of July 2026, and a combined 33,600+ founder following across two personal X accounts.
Co-founder Magnus Müller traced a Perplexity bug back to a hardcoded test string in Browser Use's own code, a detective thread that became one of the company's biggest organic hits.
The GitHub repo is the entire top of funnel. Virality came from becoming a dependency inside other viral AI products, not from marketing spend, which converted developer curiosity directly into branded search.
02 How it unfolded
The order the channels came online. Sequence is strategy: what they did first, and what they layered on once demand existed.
03 Traffic & channel mix
Estimated demand, the channel split behind it, and the keywords and referrers doing the work. Directional modeling, not audited analytics.
Traffic sources
Top search keywords
| Keyword | Volume | Weight | CPC |
|---|---|---|---|
| browser use | 43k | $2.38 | |
| browser-use | 12k | $2.92 | |
| browseruse | 9.2k | $2.90 | |
| browser use api | 910 | - | |
| browser harness | 14k | $3.00 |
Top referring sites
Closest competitors by audience
Browser Use pulls an estimated ~412K monthly visits, down 24.8% over the last 3 months. Direct traffic leads at 41.8%, meaning most visitors already know the product and type the URL in directly, followed by Search Organic at 36.1%, the second-largest channel and a sign that branded search demand alone carries the site. Social Organic sits third at 7.8%. The remaining channels, referrals, Gen AI answer-engine traffic, email, and paid social, together total about 14.2-14.3% of visits; the 5.8% Gen AI share, meaning visits referred from AI answer engines (ChatGPT, Perplexity, AI Overviews) citing the docs, is worth noting on its own as an early but real signal for an AI-developer audience.
The keywords they rank for are almost entirely their own name: "browser use" pulls ~43K monthly searches, "browser-use" ~12K, and the category term "browser harness" ~14K, meaning search demand is dominated by people who already know what they're looking for rather than generic category shoppers. Top referring sites are AI-tool discovery directories, aixploria.com, saaspo.com, and aiengineerpack.com, meaning inbound traffic comes from being listed in "best AI tools" catalogs rather than press or peer backlinks. The closest competitors are browserbase.com, a funded direct rival in hosted browser infrastructure, axiom.ai, a no-code automation tool serving a less technical buyer, and hyperbrowser.ai, another browser-infra-for-agents entrant, evidence the category is getting crowded fast.
04 Organic content
The specific pages earning their organic search traffic, and the pattern behind why they rank. Adapt the format, not the topic.
Their top organic pages are the product itself: the homepage draws the large majority of organic visits, with cloud.browser-use.com and three docs pages (a cloud quickstart, an MCP-server integration guide, and the open-source introduction) rounding out the rest. There's no blog post, comparison page, or "alternative to" page anywhere in the top five, and the top-ranking keywords are the company's own name and near-name variants plus category terms like "cloud browser" and "ai browser," where it already ranks #1-#6. The pattern reads as branded-demand capture through the product surface itself, docs and app pages absorbing search traffic that content marketing would normally chase, rather than a built content flywheel.
Top pages by organic traffic
05 Channel map
Not traffic share. How much weight the growth system actually puts on each channel, with a one-line read on the role it plays.
What drove growth
06 Who they’re hiring
Open roles read like a roadmap: the functions they’re staffing show where the company is investing next and what stage it’s at.
3 open roles
07 The motions doing the work
For founder-led SaaS the breakdown shifts from ads to traction: where the first users came from, how the founder grows it in the open, and the compounding organic surface.
Browser Use's growth started before its official launch: the library went viral on Hacker News on its own, and only after that did the founders apply to YC, run a Product Hunt launch, and post an official Launch HN.
In the 50,000-star reflection post, Gregor Zunic wrote "a few weeks after Hacker News blew up, we applied to YC and got an interview," describing organic HN attention that predates any submission in the visible record and had already pushed the repo to 23,000 stars by January 26, 2025.
A February 13, 2025 tweet confirms a Product Hunt launch reaching 60 upvotes within four minutes of posting, twelve days ahead of the official Hacker News launch.
Dig deeper+2▾
On February 25, 2025, Magnus Müller posted "Launch HN: Browser Use (YC W25), open-source web agents," which hit 259 points and 100 comments, the largest single launch-day HN result in the visible record, four days after a low-key, non-founder HN post ("Enable AI to control your browser," 1 point) had already surfaced the idea.
By March 2025, public reporting tied a sharp download and star spike to Manus, a viral Chinese AI-agent product built on the library, and the founders' own tweets track the acceleration from "45K+ GitHub stars in 4 months" (March 17) to 50,000 stars (March 29).
Growth compounds through two personal X accounts, not the brand handle. The founders narrate milestones, bugs, and hiring needs in real time and let the audience do the amplifying.
On April 9, 2026, Magnus Müller explained that a known Perplexity glitch, randomly answering "what is the capital of France," traced back to Browser Use's own hardcoded LLM-connection sanity check, since Perplexity is built on the open-source library; the thread pulled 2,511 likes and 268,971 views.
Zunic's January 26, 2025 post announced the Browser Use Cloud launch alongside a 23,000-star milestone, and Müller's October 16, 2025 post disclosed "1M pip installations last month... since start 11 months ago that's +100% monthly growth," a smaller but concrete usage-scale disclosure.
Dig deeper+1▾
Zunic's March 17, 2025 hiring post led with the traction number, "45K+ GitHub stars in just 4 months," and the line "we are MELTING under customer demand," using the audience itself to recruit a senior engineer to what was then a two-person team.
The SEO motion runs through the product and the repo, not a blog. Search Organic is the second-largest slice of the 412K-visit channel mix, and the top organic pages are the homepage, the cloud app, and docs, the pages covered above, with no comparison or listicle content anywhere in the mix.
With 105,000+ GitHub stars and real integrations into products like Perplexity and, per public reporting, Manus, the open-source repository itself functions as a distributed discovery asset, generating the ~43K/mo of branded search volume without any content-marketing spend.
No Meta, Google, or TikTok ad footprint appears anywhere in the evidence, and no referral or affiliate program is visible on the site; every documented "ad creative" in this report is actually unpaid creator UGC, meaning the entire growth budget runs through product, community, and the founders' own channels.
08 The playbook
The channels are not separate. They are one system where each stage feeds the next. Here is the read, then the plays to run tomorrow.
The inferred playbook
Attention starts with the repo getting embedded in someone else's viral product, converts through branded search and direct traffic into a self-serve pricing ladder, and compounds through the founders' own feeds between spikes, with no mid-funnel content layer to catch what launches don't convert.
GitHub stars and the confirmed Manus dependency are the earliest traceable driver of attention, not ads or blog content, given the zero paid footprint and blog-free organic top pages covered above.
Self-serve signups land on a freemium ladder, Free at $0/mo (3 concurrent sessions, 10 agent tasks/mo), up through Dev at $29/mo, Business at $299/mo, and Scaleup at $999/mo, each a usage-credit pool, before Enterprise is quoted directly, matching the 41.8% Direct and 36.1% Search Organic channels that suggest most buyers already know the name before they land.
Dig deeper+3▾
The founders' 33,600+ combined X following and Zunic's recurring Hacker News technical posts, the 322-point Firecracker VM hit over a year after launch, keep resupplying awareness between viral spikes rather than relying on a single launch moment.
No blog, comparison, or "alternative to" content sits anywhere in the organic top pages, and the launch-driven spikes aren't being replaced by a compounding content layer; on the team side, only 3 roles are listed as open, suggesting operating leverage still runs through the founders' own channels rather than a built-out growth function despite the $17.5M raised.
how many of the ~412K monthly visitors convert past the Free tier's 10-task cap into the $29+ paid plans, the size of the Discord community that creator posts point people toward, and whether the Manus and Perplexity usage are commercial relationships or incidental open-source dependency adoption.
Turn your own bug reports into public content
The proofMagnus Müller's biggest organic hit wasn't a launch post, it was tracing a competitor's product bug back to a hardcoded string in Browser Use's own code, pulling 268,971 views.
The adaptationScan your own support tickets or error logs weekly for a reproducible, specific technical anomaly, ideally one that touches a well-known product other people use, and write it up under your personal account with the exact string, log line, or error message as the hook, not a generic "case study." The mechanism: specificity (an actual hardcoded value, an actual bug ID) is what made the thread spread, so lead with the artifact, not the takeaway.
Cost: $0 · Time to signal: days to weeks · Works pre-PMF: yes
Ship one engineering deep-dive to Hacker News every month
The proofGregor Zunic posted roughly eight technical write-ups to HN since the February 2025 launch, on leaving Playwright for raw CDP, sandbox infrastructure, and booting VMs fast, eventually landing a 322-point front-page hit over a year after the original launch.
The adaptationPick the single most surprising performance number from your last month of engineering work (a speed-up, a cost cut, an architecture change) and write 800 words on how you got there, framed as a factual engineering claim in the title, mirroring "how we start browsers in under 1s" rather than a marketing headline. Submit it from a founder's personal account, not the company account.
Cost: $0 · Time to signal: weeks · Works pre-PMF: yes
Stagger your Product Hunt and Show HN launches instead of stacking them same-day
The proofBrowser Use launched on Product Hunt on February 13, 2025 (60+ upvotes in minutes), then ran its official Launch HN twelve days later on February 25, which outperformed it by far, 259 points versus a smaller Product Hunt showing.
The adaptationRun Product Hunt first to gather early comments and testimonials, then use those as social proof in the top comment of a Show HN post roughly a week or two later, titled as a plain factual claim about what the tool does rather than a hype headline. This is a one-time moment per product, not a repeatable weekly lever, so pair it with an ongoing posting cadence like Play 2 rather than expecting a second launch to work the same way.
Cost: $0 · Time to signal: days · Works pre-PMF: yes
Recruit from your own traction feed
The proofZunic's hiring post led with "45K+ GitHub stars in just 4 months" and "we are MELTING under customer demand" to recruit a senior engineer to a two-person team, using the audience itself as the recruiting channel.
The adaptationWhen you need to hire, post the ask from your own personal account and open with the single most impressive real number you have, whether that's users, revenue, or a growth rate, rather than a generic job description, and frame the role around a specific structural gap you're drowning in. This only works if you actually have one genuinely impressive metric to lead with; a role posted with no real traction to point to reads as a generic job ad, not a growth signal.
Cost: $0 · Time to signal: days to weeks · Works pre-PMF: conditional, provided you have at least one real metric worth leading with
Make your product embeddable enough that other builders' virality becomes yours
The proofBrowser Use's biggest star and download spikes came from becoming a dependency inside other products going viral, Manus and, per Müller's own detective thread, Perplexity, not from any campaign the founders ran themselves.
The adaptationKeep your core tool under a permissive license with a one-line install and a minimal API surface, then actively pursue integration into two or three adjacent open-source projects in your category by contributing the integration yourself rather than waiting to be discovered. This is a distribution bet you don't fully control since it depends on someone else's product taking off, so treat it as a slow-burn side channel, not a primary growth plan.
Cost: $0 · Time to signal: months · Works pre-PMF: conditional, only if you have an OSS component cheap enough for others to embed Not transferable at an earlier stage: the $17.5M in funding behind Browser Use's founding-engineer hiring push, and the fact that the Manus and Perplexity dependency spikes required another company independently choosing to build on their code, a mechanism a reader can pursue but can't manufacture on demand.
You just read how Browser Use does it. Tomorrow it is someone else.
Every morning we take one company that is actually growing and break down where its customers come from: the ads still running after a year, the channel doing the real work, and the play you can run this week.
Systemaic · directional intelligence. Traffic, spend, and reach figures are SimilarWeb-style estimates and qualitative reads of public data, not audited numbers. Built on real public receipts.
// teardowns with a similar GTM motion
sources & receipts (6)
Launch archives: Hacker News: History of the browser user-agent string (2008) · Hacker News: How we run Firecracker VMs inside EC2 and start browsers in · Hacker News: Launch HN: Browser Use (YC W25) – open-source web agents · Hacker News: Closer to the Metal: Leaving Playwright for CDP · Hacker News: An advanced browser fingerprint calculator aimed mainly at T · Hacker News: Tor Browser User Manual
Traffic, spend, and revenue figures are estimates as noted in the report; the links above are the primary public artifacts.