Project · Testing by Prompting an Agent to Research Your Product

Prompt 3 different AI assistants to research your product. Document what they say, what they get wrong, and what sources they cite. Identify the single most important content change you would make based on what you find.

• Verbatim transcript of each agent's response to 'What is [product] and how much does it cost?'
• A table comparing accuracy, completeness, and cited sources across all three agents
• A list of factual errors or omissions found, each traced to a specific content gap on your site
• One concrete content change recommendation with a before/after rewrite of the affected copy

• All three agents tested with identical prompts and responses recorded verbatim
• Cited sources identified and categorized (own domain, review site, third-party, none)
• At least one factual error or omission identified and traced to a specific page
• Content recommendation is specific: names the page, the sentence, and the proposed replacement