← All AI writing use cases
Landing pages

Build your landing page from the exact words buyers use in reviews

Tools:
Claude, ChatGPT
Who it's for

Founders writing a landing or sales page who don't have many customers to interview yet

What they report

Joanna Wiebe of Copyhackers reports that a headline taken word for word from an Amazon book review, "If you think you need rehab, you do.", got more than 400% more clicks than the control headline on a rehab center's site, and more than 20% more lead-form submissions on the next page. She did the mining by hand, reading 500+ reviews in about 2 to 3 hours.

Example

Wiebe's client couldn't easily interview past patients, so she read reviews of 6 books about addiction and pasted reviewers' words, with as little summarizing as possible, into 3 tables. One reviewer's line became the winning headline. When rewriting the Google Play description for the Flow task app, she read reviews of Flow and of its competitor Asana, and recurring complaints (no offline mode, slow loading, no subtasks) shaped what the description needed to answer.

from source

How to do it

  1. Answer Wiebe's 4 questions first: who is the audience, what do they want, what are they using today to get it, and where online do people review those options.
  2. Collect 50 to 200 reviews of those options (books, competitor products, apps, services). Include the 2 and 3 star reviews. Keep the reviewer's exact wording.
  3. Paste them into the prompt below and have the model sort quotes into pains, desired outcomes, worries and sticky lines, with a count for each theme.
  4. Check every quote against the original reviews before you use it. Drop anything the model paraphrased.
  5. Build the page order from the themes people mention most, then write the headline and subhead from the strongest real quote.
  6. Test the review-based headline against your current one before you call it a winner.

The prompt

Fill in the [brackets]
Copy prompt

You are a voice-of-customer researcher. Below are [number] reviews of [products, books or services my buyers use today]. My buyers are [audience]. They want [outcome]. Step 1: Sort the reviews. Make 4 tables: - PAINS: what was wrong before or what frustrates them - OUTCOMES: what they want or what changed for them - WORRIES: doubts, objections, things they wish they'd known - STICKY LINES: phrases that are vivid enough to use as a headline Use the reviewer's exact words in quotes. Do not summarize or clean up grammar. Add the review number after each quote. Step 2: For each table, group repeated ideas into themes and count how many reviews mention each one. Sort by count. Step 3: Draft a landing page outline for [my product] using only these themes, in order of frequency: headline, subhead, 3 problem points, how it works in 3 steps, 3 objections answered, call to action. Write 5 headline options built from STICKY LINES quotes. Rules: - If a quote is not in the reviews, do not use it. - Do not invent results, numbers or testimonials. - Mark any claim about my product that needs proof with [NEEDS PROOF]. REVIEWS: [paste]

Prompt:
Legacy Builder version, based on the method in the source (Wiebe mined the reviews by hand)
Source:
Joanna Wiebe, 'Amazon review mining for copywriting' (Copyhackers)
Read the original ↗
Legacy Builder Lab

Want help putting these to work every week?

Legacy Builder Lab is Clifton Sellers' community for founders who write their own content with AI. Members get his prompts and bot setups, five new post templates every Monday and a monthly live teardown.

Join the Lab waitlist
Run by Clifton Sellers, who also curates this library.