Most tutorials describe what a tool does. This one walks through what actually happens when you use it, end to end: a trend surfaces, you validate it, a draft appears in your voice, you sharpen it, and it ships to every channel you publish on.
By the end you will know exactly what each stage produces and where your judgment fits in.
The Starting Point: Watchtower Catches a Signal
Watchtower monitors news, social, forums, and newsletters continuously. Every article that comes through is scored on three axes: momentum (how fast coverage is growing), sentiment (how the coverage leans), and source credibility (who is saying it).
You do not see raw articles. You see trend cards: clustered topics that have crossed a momentum threshold. A card shows the topic title, the composite score, and the underlying articles so you can verify the signal yourself.
What you do here: open the trend list, scan for topics in your lane. A card scoring high on momentum with solid source diversity is usually worth a look. If the underlying articles all trace back to one original story, the signal is thinner than the score suggests. If three or four independent outlets are covering it, the signal is real.
The point of this stage is judgment, not speed. Watchtower removes the scanning work. It cannot remove the decision of whether the topic is yours to own.
From Signal to Brief
When a topic is worth writing about, the handoff to drafting starts with a brief. The brief is short, and it is the highest-leverage 90 seconds of the whole workflow.
A good brief contains three things:
- The angle in one sentence. Not "AI regulation." But "The EU AI Act's latest guidance quietly changes what counts as a high-risk use case, and most teams have not noticed."
- What you know that others do not. Your experience, your data, your stake in the outcome.
- The takeaway you want the reader to leave with.
The brief matters because the draft is only as good as the direction. An AI twin can structure, phrase, and format. It cannot know which angle is yours unless you tell it.
Drafting in Your Voice
With the brief in place, Content Suite generates the first draft. This is where the voice twin does its work: the draft uses your phrasing patterns, your sentence rhythm, and your typical argument structure, learned from the content you have already published.
The draft arrives with a voice-match score. Drafts below your threshold never even reach your queue, which keeps the noise down.
What you do here: read the draft once end to end without editing. You are checking for three things. Does the argument hold? Is the angle actually yours? Did anything important get invented? That last one matters most. AI drafts are strong on structure and can be confidently wrong on facts. Verify anything that names a number, a source, or a claim before it survives to the next stage.
Sharpening: The Edit That Matters
The edit is not a polish pass. It is three specific moves.
Cut the throat-clearing. The first paragraph of most AI drafts restates the prompt. Delete it and start where the argument starts.
Replace general claims with your specifics. If the draft says "many companies struggle with X," replace it with the concrete version: which companies, which struggle, what changed. Your specificity is the moat. Generic text is the one thing readers reliably punish.
Tighten the ending. The draft will end politely. Endings should do work: a question, a challenge, or a clear next step. Pick one.
A good rule for the whole pass: if a sentence does not sound like something you would say out loud, it is not yours yet. Rewrite it or cut it.
Formatting for Every Channel at Once
One draft, many destinations. This is where most workflows break down, because native formatting is the part nobody wants to do manually.
Content Suite reshapes the approved draft per platform automatically:
- LinkedIn: the hook-first, 150-300 word version, ending in a question
- X: a thread with one claim per post and a quotable close
- Instagram: a carousel where each slide carries one idea in large text
- Newsletter: the fuller version, with sections and links
You review the set as a batch. The voice stays constant because every version draws from the same approved draft. Only the format changes.
Scheduling and Shipping
Social Deck queues each formatted version for its channel, at the times the audience for that channel is actually active. You can accept the suggested slots, pin your own, or publish immediately.
Two details worth knowing:
Approval gates scale with your team. Publishing solo, the queue is one click. With a team, drafts route to whoever holds sign-off for that channel, with revision history intact.
Analytics feed back. After publication, performance flows back into the system. Posts that performed well sharpen the voice profile and the hook scoring. The loop is what makes the workflow faster every month you run it.
Where the Time Goes
The full pass, signal to scheduled, is roughly 20 minutes for someone practiced, and under an hour the first time. The rough breakdown:
- Reviewing trend cards and picking a topic: 5 minutes
- Writing the brief: 2 minutes
- Reading and verifying the draft: 5 minutes
- The sharpening edit: 5-10 minutes
- Reviewing the per-channel formats and scheduling: 3 minutes
The stages where you spend real attention are the beginning (is this topic mine to own?) and the middle (is this true, and is it mine?). Everything else is mechanics that the workflow absorbs.
The Compounding Part
The first cycle is the slowest. Every pass after it gets faster for two reasons.
First, the voice twin improves. Every edit you make sharpens future drafts, because corrections are signal about how you actually write.
Second, your own judgment gets faster. You learn which trend cards are yours, which angles fit your voice, and which edits you always end up making. The workflow does not just publish posts. It trains you on your own publishing patterns.
That is the entire loop: signal, draft, sharpen, ship, learn. Run it enough times and the question stops being "can I keep up with the feed?" and becomes "which of these waves do I want?"