Carrd: $2M ARR With Zero Marketing Spend and an Anonymous Solo Developer — the “Don’t Add Features” Strategy

Why This Case Matters

Site builders are a market ruled by Wix (2023 revenue over $1.56B) and Squarespace, supposedly the last place an indie developer should enter. Yet in that market, Carrd reached $2M ARR (roughly ¥300M/year) with zero ad spend and zero promotion. What’s more, its developer AJ remains anonymous, showing neither his face nor his real name. How did one person survive in a market of giants? The answer lies in a contrarian bet, “don’t add features”, and an invisible ten-year runway.

The Numbers First

A Timeline That Includes a Ten-Year Runway

What Was He Actually Doing?

Carrd is a site builder that can only make one-page sites. It carved out a unique position not by adding features but by cutting them. As AJ himself put it, “I couldn’t build a Squarespace-class builder on my own, so it made sense to aim a bit lower”, before being an aesthetic, this was a sober assessment of what one person could actually finish. Because of its simplicity, it is cheap, fast, and has almost no learning curve. It spread explosively among individual creators for link-in-bio pages, profiles, and landing pages.

The experience design is also relentless about removing friction. Without creating an account, you can jump straight into the full feature set at carrd.co/build. Sign-up is required only at the moment of publishing. AJ says, “I first decide on the user experience I’m aiming for, then work backward from there. UX is never an afterthought.”

Notably, developer AJ is anonymous. He says hiding his face and real name is not a strategy but a holdover from 2000s internet culture. Yet the result is 2.2 million users reached through word of mouth alone.

Acquisition — the Invisible Asset of a “Previous Audience”

The secret behind zero marketing spend lies in html5up. Years of giving away high-quality templates for free had accumulated visitors and fans worth 14 million downloads, backlinks, and a #1 search position. Carrd was announced to that audience with a single tweet, getting its initial momentum for free. “Zero marketing” more precisely means “the marketing had already been done by the previous project.” A free giveaway becomes the distribution network for the next paid product.

Ongoing traffic also runs on a self-sustaining mechanism. Roughly 90% of traffic is direct visits and social, and the “Made with Carrd” footer link on free-plan sites plus an opt-in referral program turn the product itself into its own advertising. Putting a Carrd page in an Instagram profile link became exposure in itself.

The Failures and Limits Are on Record Too

This is not an undefeated story. First, the direct commercialization of the previous project was close to a failure. Pixelarity offered 91 templates at $19 but “never amounted to much”, he had built an audience, but couldn’t monetize it through template sales. Carrd was a second attempt to re-monetize that audience with a different product.

The ceiling on scale also deserves a cold look. $2M is only about 0.1% of Wix’s revenue. The feature-limiting strategy creates an arena where one person can compete, but it also means giving up most of the market from the start.

And AJ himself acknowledges the limits of “one person.” After several years solo, he realized he had become the bottleneck, hired a developer, and his friend Doni from the html5up days joined as co-founder, making a team of 4. Even the 2021 seed round, raised while profitable, was not for the money but for advice and connections on “scaling infrastructure, content moderation, hiring, and support.” At 2.2 million users, the non-technical problems, monitoring illegal content, running support, outgrow one pair of hands. Carrd answered the same kind of limit that led to Peing’s sale, but with fundraising and hiring instead.

What This Case Teaches

“Feature limitation” is not a weakness but a positioning. In AJ’s words, “Constraints are superpowers.” The one-page-only constraint kept the development workload to a single person’s capacity, lowered the price, created a “no decisions to agonize over” experience, and built an arena where he wouldn’t be compared to the giants. Four effects flow from one constraint. Against Bannerbear’s “it’s the target you should niche down on”, Carrd demonstrates the other way to narrow: narrowing the scope of the product itself.

The free-giveaway-then-paid-product two-stage rocket is a repeatable pattern for indie developers. The html5up → Carrd relationship is the same “build the audience first” play as Irie’s public content → MENTA’s 8,000 visits on day one and Tony Dinh’s followers → TypingMind’s $1,000 on day one. The difference: instead of social media, he made a free product itself the audience-gathering machine. Even developers who dislike posting online can execute this. But as the Pixelarity lesson shows, monetizing an audience rarely works on the first try. The winner wasn’t a product too contiguous with the giveaway (template sales) but one that abstracted the audience’s problem one level up (a site builder).

Anonymity can sell, if the product is sufficiently self-explanatory, a personal brand is not mandatory. While building a personal brand is considered standard practice, Carrd substituted the clarity of a product you understand the moment you touch it. Read together with Lively Table’s “founder-dependence gets no price at exit”, investing in a personal brand is a choice, not a requirement.

Conditions for Replication — and the Limits

• What’s replicable: the sequence of “first give away something valuable for free to build an audience, then release the paid product that audience wants” works in any domain — templates, assets, tools. The no-signup-before-trying funnel and the self-propagating “Made with” link on free plans can be ported directly into any indie SaaS
• The limits: html5up’s accumulation took years — about 4 years before Carrd launched, about 10 if you count through to virality. Replicating “zero marketing” requires an equivalent invisible upfront investment. Exogenous viral moments like the Kardashian usage cannot be engineered, and moderation and support at scale ultimately demand people and money

Related Cases

• Photopea — the other one-person mega-product, built on a free model
• Plausible — a different route to $1M ARR with zero ad spend (conviction-driven)
• Bannerbear — the other right answer to “how to narrow down”