Import journey with AI

When the customer story already exists somewhere else, do not start from a blank canvas. Use AI import to get to a first draft quickly, then spend your time fixing scope, naming, and evidence instead of rebuilding the whole map by hand.

​

What this does

AI import creates a draft journey from one source you already trust. Custory can draft:

• a journey name
• likely stages
• likely steps
• initial structure your team can review

This is a drafting tool, not a final answer. The value comes from choosing a good source and cleaning the result up right away.

​

Choose AI import when the source is already clear

AI import is usually the best starting point when one source already explains the journey better than a blank journey would. Good fits:

• a public setup or onboarding flow on your website or docs
• one GitHub repository that reflects the product area you want to map
• one focused Notion page or database
• one Figma file that shows the flow clearly

If your material is mostly interview notes, support notes, or mixed internal docs, AI as a workspace member or Migrating to Custory may be the better first step.

​

Current starting paths

Custory currently supports these journey-creation paths from the dashboard:

• From website or URL
• From GitHub, when GitHub is connected
• From Notion, when Notion is connected
• From Figma, when Figma is available and connected in your workspace

If a connected source is not set up yet, Custory opens the connection flow first. If a source is not available in your workspace, use a URL import, a template, or AI chat instead.

​

Pick the source that tells the truest story

The best import is usually the narrowest source that already describes one real customer journey. Prefer:

• one repo over a broad monorepo
• one specific docs page over a generic homepage
• one focused Notion page over a mixed workspace
• one relevant Figma file over a large design library

If you give the importer a source that mixes several flows together, the draft usually comes back broad, generic, and noisy.

​

Add journey context before you run it

Custory lets you describe what the journey should cover before the import runs. Use that field. Good examples:

• Onboarding journey for new self-serve users
• Trial-to-paid flow for first-time workspace owners
• Activation journey for teams connecting Slack and inviting their first teammate

Weak prompts such as customer journey or main flow leave too much room for the importer to guess wrong.

​

Which import path to use

​

Website or URL

Use a URL import when the public product story already explains the journey well. This works best for:

• self-serve signup flows
• pricing and plan-comparison paths
• onboarding or setup docs
• public help content that describes activation clearly

​

GitHub

Use GitHub import when the most useful source is the product itself. This works best when:

• the product already exists
• one repository maps closely to the customer flow
• engineering reality should shape the first draft

Before importing from GitHub:

1. Connect the GitHub integration.
2. Select the repository you want to use.
3. Add a short note describing the journey scope.

​

Notion

Use Notion import when the clearest source is already documented. Good examples:

• research notes
• discovery summaries
• workshop writeups
• onboarding documentation

Choose the page or database that is closest to one real flow. Broad internal workspaces tend to produce weaker drafts.

​

Figma

Use Figma import when the structure of the experience is easiest to understand from the design itself. This is useful for:

• early-stage products that are still design-heavy
• new flows that are not fully shipped yet
• redesign work where the screens explain the journey better than the current implementation

Paste a Figma file URL or file key in the import flow. Custory reads the text in that file so the draft can reflect the flow shown in the design.

​

What to do right after the import

Do not treat the draft as ready for the rest of the team. Do this first:

1. Rename anything vague.
2. Remove stages that do not belong.
3. Rewrite steps from the customer’s point of view.
4. Add or correct the first items.
5. Link the right persona if one matters.

Most good imports need a 15 to 30 minute cleanup pass before they become the map the team should work from. Add screenshot of the AI import flow and first imported draft here

​

Example

Say your team wants to map trial-to-paid onboarding. You already have:

• a pricing page
• a help article for setup
• one repo for the onboarding flow

The best move is usually to pick the source that reflects the customer experience most directly. If the public setup path is already clear, start with the docs page or URL. If the product flow is clearer in the codebase, use the repository instead. After the draft appears, trim it until it sounds like your product and matches what customers actually experience now.

​

Import mistakes to avoid

​

Next step

Read Your first journey (step by step) to turn the imported draft into a map the team can actually use. If your source material is scattered across notes and docs instead of one clean import source, read Migrating to Custory.