Capture problems before you look for topics.
Keep a small insight log beside your delivery notes. Record questions prospects ask, decisions clients struggle with, and misunderstandings that cause rework. Remove identifying details and respect confidentiality. You need the pattern behind the situation, not the client's private story.
For each entry, note who faces the problem, when it happens, what they currently do, and what would help them make a better decision. Choose an entry you can explain from actual knowledge. A topic being popular does not make it relevant to the people you serve.
Write a brief with one audience and one job.
Before opening an AI tool or a blank editor, fill in these five lines:
- Reader: Who is this specifically for?
- Situation: What are they trying to do right now?
- Point: What is the single idea worth remembering?
- Support: What example, process, or evidence makes the idea useful?
- Next step: What can the reader try after reading?
Reader: a freelance designer starting a project with several stakeholders. Point: identify one final approver before the first review. Support: a sample approval checklist. Next step: add an approver field to the kickoff brief.
This example can become a helpful post without promising more revenue or presenting an invented client result.
Draft with AI, then check the substance.
Give the model your approved brief, audience, tone, and boundaries. Ask it to organize the explanation, suggest a clearer opening, or identify unanswered questions. Tell it to leave missing facts as questions instead of filling them with plausible details.
Review every claim against the source material. Remove fabricated statistics, unsupported client stories, generic recommendations, and instructions that do not fit the reader's situation. Check whether the final piece gives someone a decision, a checklist, or a usable example.
A fluent draft can still be inaccurate or irrelevant. Keep a human review step before publication, especially when discussing client work, business claims, or confidential information.
Adapt the insight to the channel.
Use the same underlying insight with different structures. An X post might explain the mistake and show a three-line fix. A LinkedIn post can add the scenario and a short checklist. A video script can demonstrate the before-and-after structure of the brief.
Do not paste the same block everywhere. Remove context the channel does not need and add context the reader needs to act. Keep the central claim consistent. If you change the claim while repurposing, review the evidence again.
Set a cadence you can maintain alongside paid work. A small queue of reviewed pieces is more useful than a calendar full of topics nobody has developed.
Separate publishing from learning.
Use clear workflow states: idea, brief, draft, review, ready, published, and learning recorded. Decide who moves a piece into ready and where you will save the final version. Scheduling a piece and publishing it are different events.
After publication, record relevant questions, qualified replies, saves where available, profile visits, and clicks to a useful next resource. Track conversions separately when you have a working checkout and reliable measurement. Early impressions alone do not tell you whether a product is validated.
Once a week, review which questions deserve a follow-up, which examples need clarification, and which ideas attracted the wrong audience. Use that evidence to choose the next brief.
Build on this
Connect the next part of your workflow.
The AI Content Operating System connects insight capture, briefing, review, and publishing. The member workspace includes a Content Brief Builder and editable Content Repurposer.
Explore the Systems Stack