When I research and send in the same sitting, the last messages in the batch always get worse. Fatigue shows up as generic hooks — and generic hooks get ignored.
Splitting the work sounds slower. It’s often faster net because I don’t burn attention on both tasks at once.
Batch A: research only
- Build or trim the list against the ICP sentence.
- For each row: one hook + one note on why they’re on the list — no draft yet.
- Stop when quality drops; that’s your real batch size.
Batch B: write drafts
Turn hooks into messages in a separate block — same day or next. The gap lets bad hooks look obviously bad.
Batch C: send
A shorter session: paste, personalize last mile, send. I treat this as execution, not thinking — thinking already happened.
Why it helps
Research rabbit holes don’t steal send time. Send anxiety doesn’t steal research depth. Each mode has one job.
Minimum viable split
Even “research in the morning, send after lunch” beats stacking everything in one blurry hour.