1. Name the question you keep coming back to
Write the recurring question in your own words before opening the add-topic form. What would you otherwise search for again next week? What new information would change a plan, a decision or your understanding? If you only need a definition or a single answer, do that search once. Monitoring is more useful when the answer can develop over time.
- Illustrative work question: What implementation guidance affects our use of AI in EU healthcare?
- Illustrative personal question: What route and access information should I check before an Alpine trail-running trip?
- These are planning examples, not real customer results or case studies.
2. Set boundaries around the subject
A topic called ‘Technology’ can include almost anything. Narrow it with a technology, location, audience or type of change. Choose boundaries that affect your decision rather than packing in every possible keyword. If two questions have different audiences or would lead to different actions, create separate topics when your plan limits allow.
3. Use a short name and a specific description
The name should be easy to recognise among other topics. Put the detail in the description: subject, angle, context and the kinds of material you want to follow. Use plain language. This guides automated relevance judgments; it is not a guarantee of exact filtering or exhaustive coverage.
- Name: ‘EU healthcare AI guidance’.
- Description: ‘Follow implementation guidance for healthcare providers using AI in the EU. Focus on obligations and timelines, rather than general product announcements.’
- Name: ‘Alpine trail-running trip’.
- Description: ‘Follow route guides and seasonal access information for trail running in the Alps. Focus on practical trip planning rather than race results.’
4. Select Work or Personal and start research
Use Work for a professional decision and Personal for your own interests or projects. TopicWeaver keeps the categories separate. Free includes 5 topics in each category, with up to 6 visible resources per topic. Adding a topic and its first research are free. The progress panel reports the outcome; a search can finish with no resources, or report a discovery problem rather than producing a useful first set.
5. Assess the first results against your purpose
Open a few original sources. Ask whether each resource matches your scope, provides useful evidence and could change your understanding. Read ‘Found’ as when TopicWeaver discovered the page, not when the underlying event happened. Where available, ‘Published’ helps distinguish recent material from older background. A checked link is usable at the time of the check; it is not proof that the source's claims are correct.
6. Diagnose noise or an empty set before expanding
If most resources are off-topic, compare your description with what they discuss. Is an important region or audience missing? Is the topic serving several unrelated goals? If nothing was added, candidates may have been irrelevant, already saved or unusable, and search retrieval may be incomplete. Do not assume the subject has no activity. Use refined wording when setting up a replacement topic if needed; do not repeatedly refresh a vague topic hoping for different results.
7. Choose a review cadence and an end point
Automatic research runs about once a day for active accounts and costs no Credits. Use a Daily or Weekly digest, or turn email Off and check the app. An optional manual refresh costs 1 Credit and does not send an instant email. Revisit whether the topic still supports your project: delete it when it no longer does, after saving anything important elsewhere because deletion removes its saved resources and history.
A quick topic-description checklist
Before starting, check that you can identify the subject, explain why updates matter, name any essential context and describe what useful material would look like. Avoid promising yourself complete coverage or a verified answer. A good topic reduces ambiguity; your judgment of the original sources remains the final step.