Test a Claude-written sales page headline against your own (copywriter reports a win, and a loss on subject lines)

Sales page headlines carry most of a page's persuasive weight, and founders often rely on their own wording without testing alternatives. The question is whether an AI-written headline can beat a practised copywriter's on a live page.

1. Give Claude (Opus 4.6 in this case) basic context about the offer: who it is for and what it promises. 2. Ask for several alternative headlines for the sales page. 3. Pick one and run it against your existing headline on the live page. 4. Compare conversions and keep the winner. The author also ran AI against his own email subject lines, where his versions won, which suggests testing each asset separately rather than assuming AI wins everywhere. The exact prompt, test setup and full breakdown are in his linked video, not in the written post.

The author says the Claude headline won and conversions rose by about 50%. The $30,000 a year figure in the title is a hypothetical framing, not a measured result. He also says AI lost to him on email subject lines.

One test on the author's own page. The written post gives no traffic numbers, test split, duration or significance; the results appear in an image and a video that were not checked here. The exact prompt is not given. The author sells coaching for freelance copywriters (a Skool community offer), so the post also promotes his services. The subject line comparison is mentioned but not detailed.

An existing sales page with enough traffic to compare two headlines, and a way to split or sequence the test.

• Storey Time newsletter (Brandon Storey) ↗ Firsthand write-up, published February 19, 2026

Source published: February 19, 2026. Last reviewed here: October 22, 2026. Spot a mistake? Tell us.