Carrd.co: $9 Toy to $2M Biz

Carrd’s pages look almost like demos: a headshot, two lines of copy, a strip of buttons that point elsewhere. Yet this single-page builder now handles roughly 15 million visits a month—and is still growing more than 10 % month-over-month, according to Similarweb’s May 2025 panel. That traffic flows through link-in-bio buttons on Instagram, résumé cards, political “ways to help” hubs, and the “coming soon” teasers sprinkled across product-huntland. Before admiring the art direction, we have to start with the till, because Carrd’s economics are the reason it exists at all.

A price so low it disappears

Anyone can publish three sites for free under a carrd.co sub-domain. The instant you need a custom domain, bigger file uploads or a white-label footer, Carrd steps in with a paywall that feels like spare change: US $9 per year for the entry plan, then US $19 and US $49 tiers most users never reach. The trick is architectural. Rather than run application servers, Carrd compiles every page into static HTML, CSS and a few kilobytes of vanilla JS, then serves it from an object store behind a CDN. Compute cost rounds to zero; bandwidth is the only bill, so the margin on nine dollars is almost all profit.

Because the fee is negligible, the conversion question flips from “Is it worth paying?” to “Why not?”. Deciding costs more mental energy than entering a credit-card number, so churn implodes. Carrd’s price, not its features, is its most powerful growth lever.

Why one page still matters in 2025

Single-page sites thrive because they are navigation gateways, not destinations. Influencers paste one Carrd link in a TikTok bio and instantly route fans to merch, Discord, Patreon and Spotify. Job seekers replace PDF résumés with interactive cards recruiters can open on mobile. Grassroots groups launch donation funnels overnight. In each case speed trumps sophistication: Carrd’s editor renders in milliseconds, its template gallery is small enough to browse end-to-end, and every preset works on phones first. Competing no-code suites like Webflow or Framer can reproduce the final HTML, but they cannot reproduce the experience without shedding half their toolbars.

Personal bio navigation page

Carrd.co monthly traffic (from similar web)

Ten years, one founder, many strokes of luck

2015 – AJ, a Tennessee-based designer already making a living from HTML template shops HTML5 UP and Pixelarity, sketches a builder dedicated to the template his customers download most: the single landing page.

March 2016 – He ships the beta to his Twitter following and a Product Hunt thread; the existing audience supplies the first thousand sign-ups within days.

2017-2019 – Growth plateaus at “hundreds of new users a day.” The footer tag “Made with Carrd” does the marketing while AJ and his ops partner Doni handle everything else.

30 May 2020 – A Black-Lives-Matter resource hub built on Carrd goes viral when Kim Kardashian tweets it. Daily registrations jump from the hundreds to the thousands. Carrd’s static stack shrugs off the surge; celebrity luck compresses a year of user acquisition into a weekend.

2021 – During a Reddit AMA AJ discloses 3 million live sites and US $1 million ARR—all with two full-timers and no paid marketing.

September 2024 – Indie Hackers estimates US $167 000 MRR (~US $2 million ARR), head-count four, advertising budget still zero.

2025 – The one-page niche is now flanked by Webflow, Framer, V0 and half a dozen AI layout generators, yet Similarweb still records 14–15 million monthly visits to Carrd’s domain. Scope discipline, not feature velocity, keeps the moat wide.

Under the bonnet: why $9 works

Carrd’s client-side editor is written in plain JavaScript plus a hint of jQuery—no React, no build pipeline angst. When a user presses Publish, the system bundles the page into a static package, uploads it to an S3-style bucket and invalidates CloudFront. No database rows per page, no server-side rendering to scale, and therefore no surprise AWS bill when a celebrity fires off a tweet. With outbound traffic under 350 GB a month for ten-thousand daily publishes, the infrastructure sits inside AWS’s lower pricing tiers. At that scale the nine-dollar plan still delivers SaaS-class gross margins.

Could you copy it today?

Technically, yes, you could stitch together commodity components:

Frontend editor – Start with React (or even Svelte) but export to plain HTML/CSS. The key constraint is WYSIWYG fidelity, not framework choice.

Static site generation – When a user hits “Publish”, package the page into a zip, store it on AWS S3, and serve via CloudFront; cache invalidation becomes a solved problem.

Back-office – Stripe Billing for annual plans, Cognito for auth, SES or Resend for transactional mail.

Templates – Your real moat. Carrd’s library is hand-drawn by AJ; you’ll need designers who can sculpt tasteful variants rather than shovel hundreds of random themes.

Short links & custom domains – API hooks into Route 53 or Cloudflare, plus an in-house aliasing service that rewrites username.carrd.co to the client’s root domain.

Operational economics – At 1 MB average page weight and 10 000 daily publishes, monthly egress is still under 350 GB—well within the AWS free tier with room to spare. Charge $12-15/year and gross margins stay north of 90 %.

Follow that blueprint and your principal hurdle is not code but community: without AJ’s decade-long Twitter following your launch lacks the instant seed audience that powered Carrd’s early momentum.

Dodging the feature arms race

Every rival adding multi-page navigation, carts, memberships or AI copywriters increases both power and complexity. Carrd enforces the opposite rule: nothing ships if it elongates the publish-to-live cycle. Users who outgrow the one-page format graduate to Webflow or Framer, and AJ is happy to let them leave; a nine-dollar customer who opens a support ticket every week is a net loss. Constraint doubles as customer filter.

Founder philosophy: craft over conquest

AJ remains pseudonymous, grants few interviews and declines most partnership overtures. Freedom to tinker outranks valuation. No venture capital, no quarterly targets, no “AI site wizard” bolted on for hype. The company behaves less like a start-up and more like a sustainable craft studio—yet it prints software margins many venture-funded teams never achieve.

Personal takeaway — 小而美，一招鲜，吃遍天 (small, focused, and resilient)

Carrd proves that in 2025’s AI-saturated, VC-inflated landscape a deliberately narrow product can still carve out a durable franchise. By refusing to build anything beyond “one page, published instantly,” AJ kept code, infrastructure, pricing and support microscopic—small enough for four people to run a service powering millions of URLs.

For makers tempted by all-in-one dreams, Carrd offers a sharper question: What can we delete until only the indispensable remains? The joke project that charged nine dollars a year now earns two million dollars annually, simply because it solved one problem better than anyone else and never got distracted. Master one move, and the market will feed you—一招鲜，吃遍天.