Getting Better LLM Responses

The quality of your responses with LLM is largely dependent on the kinds of questions that you ask. Here are some best practices when asking LLM-specific questions:

• Feed data and call back: If you want LLM to be factual in responses, initially, it’s always better to feed it with actual data. I’ve seen it can fetch data, figures, and facts if you provide it beforehand.
• Tell it what you want it to be: A pro tip is to set the proper context and expectation. If you want it to be a market researcher, tell LLM the same. If you see my first prompt, I’ve done the same. However, make sure to be clear in exactly what you want, else leave LLM to it.
• Be very clear about what you want: LLM is a trained generative AI model. So, you can’t expect to beat around the bush and get quality responses. Be specific and clear about what you want and see it doing the hard tasks without wasting your time.
• Don’t rely on it solely: No matter how good AI tools like LLM gets, it’s always good to go to real resources like online reviews from real users and interviews. Don’t just rely on LLM data. Instead, scan online reviews from real customers at scale to power up your market research

References :
1. https://gapscout.com/blog/how-to-use-chatgpt-for-market-research/