ChatGPT prompt for drafting a newsletter issue

This is the prompt we run on newsletter day, when there is material but no draft yet. It produces a full issue: three subject lines, an opening that goes straight to the idea, one actionable takeaway, and any claims that need a source clearly flagged.

The prompt to copy

Draft issue #{n} of our newsletter. Audience: {who}. Format: {your structure, or "one main idea + 3 quick links"}. This issue's material: {paste}.
Rules:
- 3 subject line options — curiosity without clickbait.
- Open with the idea, not "welcome back" throat-clearing.
- One clear takeaway a reader can use today.
- Our voice: {tone}. Plain language.
- Use only the material provided; if a claim needs a source I didn't give, mark it [SOURCE?].

Replace the number, audience, and format, then paste your raw material: notes, links, loose ideas. The issue comes back structured, with subject lines to choose between.

When to use it

On send day, when you have gathered material but the blank page is waiting. The prompt makes the jump from loose notes to a readable issue in one pass, without you having to decide where to start.

What you get back

1. Three subject lines that play on curiosity without clickbait.
2. An opening that gets into the idea, no pleasantries.
3. A takeaway the reader can apply today.
4. Claims to source flagged [SOURCE?], with no invented reference.

Real example: raw material → structured issue

Input: the material Léa gathered at Brightside, an 8-person brand agency, for this week’s issue. An excerpt, as-is:

issue idea: a good visual identity is tested on mobile first
link 1: our Fernhill Coffee rebrand
most people meet a brand on their phone (to check, exact figure?)

Output, the part that matters:

Subject A. Have you ever looked at your logo on a phone? Subject B. “Mobile first” isn’t just a developer’s rule. Subject C. What a visual identity misses when you sign it off on a big screen.

Body. An identity is judged first where people see it: a phone screen. Most people meet a brand on mobile [SOURCE?] before any other surface.

Notice what the AI did not do: it did not turn “most people” into a precise percentage. The exact figure was missing from the material, so it marked [SOURCE?] instead of inventing one. Léa finds the source, adds the figure if it holds, rewords it if not. The newsletter goes out with a claim the team can defend.

This prompt ships inside the kit, alongside the other 49 in the library and the full Brightside worked examples.