Newsletters are one of the last channels you actually own — and a part-time job to maintain.

A newsletter is one of the few audience channels that survives platform churn — your list belongs to you, not Meta or X or a recommendation algorithm. The catch: owning that channel costs hours every week. Pulling source material, drafting, formatting, getting it into your ESP without breaking the layout. It adds up to a part-time job most operators can't afford to keep doing.

The original Newsletter Pilot tried to solve that by exposing the full prompt surface to the user. If you were technical enough to dial in the right prompts, you could get great output — and then copy-paste it into your mail client and reformat it anyway. If you weren't, you got generic drafts that needed a rewrite before they could go out. Either way, the part-time job didn't actually go away.

A rebuilt Newsletter Pilot that traded prompt surface for prescription. Tell it about your brand once, point it at the source you draw from, and it produces a publish-ready newsletter. No prompt engineering, no copy-paste-and-reformat round trip.

Brand-first setup

Configure your voice once. Output stays on-brand without per-issue prompting.

Source-driven drafts

Point it at a URL, an RSS feed, or your own archive — it pulls the substance, you stay in the loop.

Publish-ready output

Formatted for the major ESPs. No copy-paste-and-fix between draft and send.

Picked it up as a micro-acquisition in 2023 — a few thousand dollars and a working codebase. Rebuilt the product around brand-first setup, ran it for a year, then sold it in 2024 as the AI-content space got crowded. It made sense as a single bet; it didn't make sense as my single focus.

1. 01Prescriptive beats flexible for non-technical users. The first version handed users a steering wheel and a manual; the rebuild handed them a destination.
2. 02Micro-acquisitions are a real shape of work. A few thousand dollars, a focused rebuild, a year of running, a clean exit — the math only works if you treat it as a complete arc, not a stepping stone to something bigger.
3. 03A crowded category isn't necessarily a bad place to be — but it's a bad place to be your only bet. Owning one product in 'AI for newsletters' didn't leave room for the other problems I wanted to look at next.