Not every valuable source has an API. Some of the best demand signals are already sitting in places your team touches every day.
A founder saves a forum thread. A sales rep drops a customer complaint into Slack. A developer notices the same GitHub issue appearing across three repositories. A support lead sees a pattern in tickets before the product team does.
Manual import turns those scattered observations into structured demand intelligence.
What to Import
Good manual import candidates include:
- GitHub issues and discussions
- Product Hunt comments
- Community threads
- Support tickets
- Sales call notes
- Customer interview snippets
- Competitor review pages
- Newsletter or forum links
The source does not need to be perfect. It needs enough context for a human or AI model to judge whether there is real pain.
The Minimum Useful Shape
Each imported item should include:
- Title
- Body or notes
- Source community
- Author or account label when available
- URL
That is enough for scoring, deduplication, and follow-up planning.
Why Manual Beats Waiting
Teams often delay demand discovery because they want the perfect connector first. That is backwards.
Manual import helps you validate the workflow immediately:
- Collect real examples from sources you already trust
- Score them for intent
- Review the highest-quality signals
- Decide which source deserves automation next
If manual imports from GitHub issues produce five useful conversations in a week, then a GitHub connector is worth building. If a source produces noise, you learn that before investing engineering time.
A Simple Team Workflow
Create one shared intake habit:
- Save the link
- Add one sentence of context
- Tag the source
- Let the system score it
- Review high-score items daily
This keeps founder intuition in the loop. The machine organizes the queue; the team still judges the market.
Where Glean Fits
Glean's manual import provider accepts source, community, title, body, author, and URL. Those items go through the same Claude scoring pipeline as automated sources.
That makes manual research reusable instead of trapped in Slack threads and browser bookmarks.