# Hypotheses for testing headlines — prompt for Gemma

Plan a test of one specific headline difference without combining several changes in one variant.

Canonical URL: https://boswidget.ru/en/hypotheses-for-testing-headlines-prompt-for-gemma/

Model: Gemma

Category: Marketing

Tags: headlines, hypotheses, validation

![Hypotheses for testing headlines](https://boswidget.ru/assets/images/text-preview-gm-gipotezy-testa-zagolovkov-1200.d52d416cae14.webp)

Hypotheses for testing headlines

## Prompt

Current headline and offer: {current}. Visitors and context: {audience}. Measurable outcomes: {measurement}. Propose three independent hypotheses about unclear or unconvincing parts of the headline. For each, provide a control, one modified version, the expected behavioral change and the risk of misinterpreting the result. Do not promise improvement or calculate a sample size without data. Change only the meaning under test in each version, preserving the facts.

## Template fields

- `{current}` — Title. Example input: Current wording and verified offer
- `{audience}` — Visitors. Example input: Where visitors come from and what they know
- `{measurement}` — Measurements. Example input: Available events and validation constraints

## Launch

[Open Gemma in YesAI](https://yesai.su/en/gemma)