Case snapshot

Product: Carrd (one-page website builder)
People: AJ (founder) + Doni (joined later)
Launched: 2016
Public figures (SaaS Club interview): about $100K MRR / $1.2M ARR, roughly 2.6M users, approaching 4M hosted sites—still run by a tiny core team.

A lot of startup stories celebrate “catching a wave,” “finding a huge market,” or “raising fast.”

Carrd is the opposite.

It is a deliberately small product: help ordinary people make a clean one-page site quickly. It never tried to beat every website builder, hire a big team early, or buy growth with ads. And yet it became a business a tiny team could run for years.

What is worth studying is not the idea of “one-page sites.” It is that AJ controlled one question from day one:

What is worth building—and what will we refuse to build?

That boundary later shaped the product, acquisition, pricing, and headcount at the same time.

1. The product didn’t shrink the website—it shrank what users have to do

Many builders start from: whatever website you want, we’ll give you the tools.

That sounds powerful. For a small team it is also dangerous: every new request can become a capability you must maintain forever.

Blog? Build blogging. Store? Build commerce. Collaboration? Build permissions. Integrations? Maintain every connector. More features mean more product surface—and more support.

Carrd never took that road. The problem was narrow:

If you only need a good-looking, simple site you can publish today, you do not need that much.

The path users walk is short

Roughly:

Pick a template → edit content → adjust sections → publish

The important design choice: users do not start on a blank canvas. They start with a template.

The template already answers layout, image placement, heading size, color, and mobile comfort. Templates are not just “pretty skins.” They do work for the user—and for Carrd: fewer stuck users means fewer support tickets.

That is the first place product design and operating cost connect.

“One page” mainly constrains Carrd itself

Users still ask for more. “Can I add an About page?”

Open multi-page support and you soon need page management, navigation, menus, relationships between pages, URLs, more settings… “One more page” can become a different product.

Carrd’s move: if people only need different kinds of content, add sections on the same page—intro, product, services, case studies, contact. Users get a complete-feeling site; Carrd keeps the one-page boundary.

Do not implement the feature as requested. Ask what problem they are trying to solve. “I want a second page” may really mean “I want contact details separate from the rest.” That does not require a second page.

The skill that matters:

Meet the need without casually expanding the size of the business.

Why this saves people

Templates reduce choices
  ↓
One page limits site complexity
  ↓
Sections cover most content structure
  ↓
Common elements are enough
  ↓
Publish when done

Users feel they own a full website. Carrd does not have to maintain a “build anything” platform.

So the logic is not “few features → two people can handle it.” It is:

Limit scope first → problems cluster → less to fix and support → a tiny team can last.

Indie Hackers summarized it as: Constraints are superpowers. For small teams, constraint is sometimes what makes the business possible.

2. From 0→1: AJ did not bet on inventing a new market

AJ did not wake up and decide to “start a one-page SaaS.”

He had spent years on templates—free HTML5 UP and paid Pixelarity. Before Carrd, he already knew people wanted this kind of site and would pay for design.

So the real question was not “Does anyone need a website?” It was:

“If people who already download my templates could edit online instead of editing code and hosting themselves—would they prefer that?”

Those are different bets. One educates a market. The other changes how an existing demand is fulfilled.

Old path: download → edit → deploy yourself

New path: open the site → pick a template → edit online → publish

The lesson for ordinary founders: you do not always need a brand-new need. If you already have customers, traffic, content, experience, audience, a product, or a channel, the opportunity may be recomposing what you already have into something easier to sell.

That is much of what Carrd was.

3. At first he only needed to prove one thing

Early on, the job was not dozens of templates or a complex admin. It was:

Can a real person finish a site online and publish it?

The core path: create → edit → publish. If nobody walks that path, more features do not matter.

A common mistake: “I’ll launch when the product is complete.” Months pass, the product grows, and you still do not know whether anyone wants it.

Carrd did the opposite: let real people finish the job first—through Alpha and Beta with small groups—then widen.

For this kind of product:

1,000 signups matter less than 10 people who actually publish.

Signups mean someone clicked. Publish means the product delivered.

Per Indie Hackers, the earliest users and about $1,000 in revenue mostly came from AJ’s existing template audience. He did not spend years hunting a first cohort—old users became the new product’s first users.

4. Why free users were not pure cost

A sharp growth choice: free users can truly publish.

In many SaaS free tiers: sign up → try → consume resources → wait for conversion. If they never pay, they are mostly cost.

Carrd is different. Publish for real, with “Made with Carrd” on the site:

User finishes a site
  ↓
Site goes live on the open web
  ↓
Someone sees Carrd
  ↓
Someone clicks in
  ↓
Next user

Free users stop being only consumers. They can become a channel.

If free only burns servers and support, more free is more danger. If free also brings the next user, free is not just a promotion—it is acquisition.

5. Charging happens at the last step, not at the door

Carrd does not say “pay before you look.” It lets people build first—pick a template, write content, finish the site, even share it with friends.

By then the user’s mindset has changed: from “trying a tool,” to “I already built this,” to “I want to use it for real.”

Then needs appear: custom domain, remove branding, forms, more “official” capabilities. Unimportant at signup; urgent after the site exists.

The paywall is not “want to try premium?” It is:

“You’re done—one last step.”

The early/long-running Pro path was about $19/year (later more tiers; check the site for current pricing). Public retrospectives also note conversion jumps on deep sale discounts—this audience is price-sensitive. The bet was not maximizing ARPU; it was keeping the step small enough that people who already got value would cross it.

6. Free and paid each do one job

Users	Main job
Free	Use the product, publish, create exposure
Paid	Custom domain, remove branding, etc.—pay the bills

Both sides matter. Cripple free and you lose real sites and word of mouth. Give everything away and nobody funds the business.

Carrd sits in the middle: free is valuable enough to deliver a real outcome; pay becomes natural when someone is ready to use that outcome formally. More natural than “trial → nag to upgrade.”

7. Acquisition stopped depending only on AJ

Early growth leaned on AJ’s Twitter and prior audience. Later the pattern flipped:

Was: AJ shares Carrd → users arrive

Became: users publish sites → others see them → others arrive

The product started carrying part of acquisition.

2020 brought a sharp spike—more online events, organizers on Carrd, and public figures (including Kim Kardashian) sharing Carrd-built sites. In the SaaS Club interview, AJ said growth was so strong that infrastructure meant to last years hit capacity early.

That is not a playbook you can copy. Celebrity shares do not repeat on demand. What is transferable: when outside traffic arrives, the product can absorb it. Real free publishing, brand marks, and user-to-user discovery are the durable base.

8. When did AJ hire?

Why not staff up to dozens early?

Hiring was not “we’re big now, so we should hire.” It was:

“Which single job can one person no longer do?”

• Early: AJ builds the product.
• As usage grows: support, moderation, community eat his time → Doni takes that load.
• Later: AJ himself becomes the coding bottleneck → first developer hire.
• After the 2020 surge: scale breaks old infrastructure → rebuild and migrate. SaaS Club says about 2.5M sites moved in roughly a month with zero downtime.

Product unfinished → do it yourself
  ↓
Support overloaded → bring in Doni
  ↓
Dev overloaded → hire a developer
  ↓
Scale breaks → fix infrastructure

Every headcount add maps to a real bottleneck—not to looking like a “real company.”

9. Why the business could fund itself

Carrd stayed largely bootstrapped—grow on revenue first, not on a big raise. Later capital, in public retellings, was more about advice on scaling, moderation, and hiring than buying ads or hiring dozens upfront.

It works because earlier product choices already capped cost: one page, templates, free real publishing, pay-at-go-live needs, and a tiny team. That is the loop.

10. What Carrd got right was not “one page”

By now the story is not about one-page sites.

What AJ got right: one product choice that lowers cost somewhere else in the business.

One-page site
  ↓
Narrow product scope
  ↓
Lower build + support cost
  ↓
Sustain a real free tier
  ↓
Users publish publicly
  ↓
“Made with Carrd” spreads the product
  ↓
More users enter
  ↓
Some prepare to go live for real
  ↓
Domain / unbrand / forms create paid demand
  ↓
Revenue covers team + infrastructure
  ↓
No need for constant paid ads
  ↓
Stay small
  ↓
Keep the boundary

One page is not a moat. Neither is $19/year. Neither is freemium alone. The power is how they reinforce each other.

Tighter product → lower cost → freer free tier → more publishing → more organic exposure → some pay when they go live → revenue without endless hiring → staying small reinforces the boundary. A closed loop.

11. What founders should copy is not “build another Carrd”

“Niche markets can win” is too thin. Better questions:

1. What do you already have?
AJ started with templates, users, traffic, craft, and proof people would pay. Inventory old assets before hunting a brand-new idea.

2. Is the business too wide?
The failure mode is often serving everyone. If you are two or three people, ask: what can we do unusually well that people already pay for—and do only that first.

3. Are free users cost or channel?
If free only burns resources, freemium is dangerous. If using the product also recruits the next user, free is acquisition. Ask: when someone uses this, can they help me get the next someone?

4. When do you charge?
Not only “what feature do we sell,” but “when do they need us most?” Carrd’s moment: site finished → ready to go live → paid needs appear. Timing beats a longer feature list.

5. Do you actually need to hire?
Not “we have revenue, so hire.” Ask which task is already blocking the business. Without a clear bottleneck, another person is often another person to manage.

The deeper proof:

A business does not have to grow by endlessly adding products, users, employees, and capital. Sometimes real growth is a product choice that also lowers another cost.

Shrink the site to one page. Shorten the build path. Let free users publish for real. Let publishing create distribution. Charge after the work is done. Hire only when something truly breaks.

What you get is not a “small feature set.” It is a model a tiny team can run for a long time.

Don’t only ask what users still want. Ask whether the choice makes acquisition, pricing, support, and headcount easier at the same time.

If yes, you are not just shipping a feature—you are building a small business that is easier to earn from.

Product

• Carrd