How to scope a SaaS clone

How to scope a SaaS clone is the step most people skip. They paste a product name into a coding agent and type “clone it”. The agent aims at the marketplace.

Every card on this site already splits the work into three rows. The first row is the replacement you can cancel a subscription with. The last row is the company you do not want to become.

I would pin the first row before I paste anything. The Typeform page is the cleanest example. Two hours for a one-question form and a response table. Two weeks and up for a visual builder, payment fields, and an integrations gallery.

Earlier today I wrote about how Sentry works. That was a teardown. This one is a regular backlog topic. The useful move is reading the table before you write the prompt.

How to scope a SaaS clone

Open the product page for the app you already pay for. Read the audience line. Read the non-goal. Then look at the three estimates.

The first row is a usable product. A list of links, a form and a table, columns and cards. It matches how much of the paid app you use this week.

The second row is the longer version of the same job, like logic jumps or a second network. You add it after you have used the first row as the real tool.

The third row is labeled “the part you should probably skip” on every page. Multi-tenant hosting, a theme marketplace, every network API, CRM sync. That row exists so you can point at it and say no.

If you tell an agent to “clone Typeform”, it aims at the third row. Copy the prompt on the page, then pin it to the first one.

What happens if you say clone Typeform

Typeform costs $29–$99 a month. That is $348–$1,188 a year. The catalog line is “A polished form for the seven questions you actually ask.”

The basic row is 2 hours, one question at a time and a table of answers. The solid row is 6 hours, with logic jumps, theming, a CSV export, and webhooks. The honest row is 2 weeks and up, the visual builder, payment fields, and an integrations gallery.

The Typeform walkthrough ships the first row. Most people need seven questions and a thank-you screen. They do not need a builder other people can use.

The same v1 also replaces Jotform and Youform. One focused build covers whichever of those you currently pay for.

Here is the shape I would paste after the prompt. Nothing else has to exist at 2 hours.

const v1 = {
  product: 'Typeform',
  estimate: '2 hours',
  ship: 'one-question form and a response table',
  skip: 'visual builder, payment fields, integrations gallery',
}

The agent now has a stop condition, which “clone Typeform” never gave it.

How to read a product page

Every build page has the same blocks. Use them in this order.

The audience line tells you who the replacement is for. Linktree is one creator managing a single public profile. Canny is a product team collecting feedback from existing customers. Buffer is one creator publishing to two social accounts. If that line is not you, pick a different card.

The goal is one sentence. Publish a bio page, or collect requests and show a status, or draft posts and publish them at a chosen time. If you need more than that sentence, keep paying.

The non-goal is the most useful paragraph on the page. Typeform says do not create a general visual workflow builder or payment fields. Linktree says do not build multi-tenant billing, a theme marketplace, or enterprise analytics. Buffer says do not add AI writing, social listening, approval workflows, or every network in v1.

Paste the non-goal into the prompt. Then paste the first estimate. Then paste the feature list only after you have cut it down.

The first-project post is about which card to open. This post is about what you do after you open it.

The three estimate rows

The table on each page has three times and three scopes. The page calls them a basic version, a solid v1, and the full honest rebuild. One number would be a lie, so the table refuses to give you one.

The basic row

This is the cancel. Linktree is 30 minutes, a links list, an avatar, and one theme. Trello is 3 hours, columns and drag-and-drop cards. Canny is 10 hours, a feedback board, votes, and statuses.

You should be able to use this as the real tool on Monday. If you still open the paid app for the core job, the first row was too thin, or you picked the wrong product.

The Linktree walkthrough is that 30-minute row. Stop before the theme marketplace. The marketplace is the week-long third row, and you are not selling themes.

The solid row

This is the work you do after the first real use. Typeform adds logic jumps and a webhook. Canny adds comments, a roadmap, a changelog, single sign-on, and email notifications. Buffer adds a calendar you can drag, previews, retries, and a second network. Linktree adds drag reorder, drafts, and daily click counts.

Start on the first row, use it, then take the pieces you personally missed. The solid row is a menu.

The honest row

This is the company. Notion’s third row is 3 months and up: databases, live collaboration, granular permissions. The Notion teardown is that post. A notes tree is 10 hours. The block CRDT is the other product.

Buffer’s third row is every network API, approvals, analytics, and AI writing. Canny’s is AI deduplication, CRM sync, and a multi-workspace SaaS. Carrd’s is a visual editor, templates, multi-tenant hosting, and custom domains.

The third row is there so you can see the bill before the agent writes it.

Pin the first row in the prompt

The product page already has a complete build prompt. It is scoped for a useful v1. Copy it. Then add three lines of your own so the agent cannot wander into the marketplace.

Build only the first estimate: one-question form and a response table.
Do not build a visual builder, payment fields, or an integrations gallery.
Stop when I can publish one form and download a CSV of answers.

“Do not build X” saves more time than another feature bullet. If the agent starts a second workspace, a theme shop, or a HubSpot import, stop it and paste the non-goal again.

Scope a Canny alternative the same way

I wrote the Canny walkthrough yesterday. Same three rows.

$79–$99 a month. $948–$1,188 a year. The basic row is 10 hours for a searchable board, one vote per person, and four statuses. The solid row is 1 week for comments, a roadmap, a changelog, SSO, and email. The honest row is a month of AI, CRM, and extra workspaces.

The useful v1 is the board plus a status people can see. Search before the first post. Move one real request to planned. Then decide if comments are worth a week.

Buffer is the same shape with a different job. $18–$36 a month. 8 hours for a post queue, one network, and scheduled publish. 3 days for a calendar, previews, retries, and a second network. A month and up for every network API. Pin the queue. Leave the analytics suite on the skip row.

Trello is 3 hours for columns and cards. The third row is live sync, Power-Ups, and the automation engine. A board you can ship in an afternoon is the first row. The rest is a different product.

When the second row is enough

Use the first row as the inbox, not as a demo. A week of real submissions, real votes, or real scheduled posts will tell you what is missing.

If you miss a webhook, add the webhook. A second network or a color picker is the same move. A visual builder is a different request, and most people never make it.

The solid row is for the thing you reached for and did not have, not for the button the paid product still shows in the sidebar.

Notion is the warning. The first row is markdown pages, nesting, and search, about 10 hours. The third row is databases and live collaboration. If the team already lives in those databases, keep paying.

What not to put in v1

Leave out the marketplace and the billing system. Leave out the integration catalog the paid product spent a decade adding.

Leave out AI writing and AI deduplication. Those need a pile of other people’s data you do not have.

Leave out multi-tenant hosting unless you are selling the clone. You are cancelling a subscription.

The second workspace, team routing, and mobile apps can wait until one person has used the first row for a week.

The guardrail on each page stays in v1. Typeform wants server-side validation, a rate limit, and a honeypot. Linktree does not want visitor fingerprints. Buffer wants encrypted tokens and idempotent publish jobs.

Start now

Go to the audit. Open the app you already pay for.

Copy the first estimate. Copy the non-goal. Paste both after the prompt.

Ship that row. Use it as the real tool. Cancel when the paid tab has been closed for a week.