Request for Audit on Prompt Improvement and Best Practices | Clay

Request for Audit on Prompt Improvement and Best Practices

Can somebody audit my prompt and help me improve it and implement best practices? Hugely Appreciate it.

Write a short, engaging first-line compliment about 
Name


 based on 

Final News Event Summaries


Description


 Avoid generic phrases like "Congratulations" or "Exciting times ahead." Make it sound natural, insightful, unique and use simple language. Avoid using em dash in sentences at all costs. Example outputs: - Saw that 
Name


 just secured funding—curious to see how this accelerates your expansion into 
Primary Industry


. - Noticed 
Name


's latest acquisition—seems like a strong move to strengthen 
Primary Industry


. - "
Name


has been making waves in 
Primary Industry


. Interested to hear how the recent 

Final News Event Summaries


 is shaping your next steps."

8 comments

Channeled
APP
·
·
Here's how you can improve your prompt following best practices: 1. Set Clear Context: • Define the purpose: generating personalized first-line compliments • Specify the input data you're working with (Name, News Event Summaries) • Explain the intended use case 2. Structure Your Instructions: • Break down the task into specific steps • Include formatting requirements • Define tone and style guidelines 3. Improve Output Specifications: • Start with phrases like "Just read" or "Just listened to" to sound more natural • Keep sentences succinct while including valuable insights • Ensure each line adds unique value 4. Add These Elements to Your Prompt: • Provide 3-5 diverse examples of successful outputs • Include negative exclusion criteria (what to avoid) • Specify word count limits • Define the desired tone and voice Remember to keep your first lines succinct while maintaining valuable insights from the research data.
Sources: Prompt Engineering Crash Course [Clay 101] - Clay University Prompt Engineering Crash Course [Limitless Research] - Clay University Intro to AI Snippets (Copywriting) [Clay 101] - Clay University
Channeled
APP
·
·
Is that what you were looking for?
Dawson Y.
·
·
Sahil S. Good start! I personally like to throw my prompt's into GPT or Claude and ask it to improve them, then refine it from there. This is what GPT gave me for your prompt improvement: """ Write a short, engaging first-line compliment about [Name] based on [Final News Event Summaries] and [Description]. Make it sound natural, insightful, and unique—avoid generic phrases like 'Congratulations' or 'Exciting times ahead.' Keep the language simple and conversational. Do not use an em dash in sentences. Example outputs:
Saw that [Name] just secured funding. Excited to see how this fuels your expansion into [Primary Industry].
Noticed [Name]'s latest acquisition. Smart move to strengthen your position in [Primary Industry].
[Name] has been making waves in [Primary Industry]. Curious to hear how [Final News Event Summaries] is shaping your next steps.
""" Also, here's a great article on prompt engineering from an engineer on clay's team: https://www.theaugmentedseller.com/p/part-2-ai-prompt-basics-for-seller
Channeled
APP
·
·
Can you tell me more about what specific aspects of your prompt you'd like to improve or what particular challenges you're facing with it?
Sahil S.
·
·
Dawson Y. thanks a lot. Actually, the prompt I shared has already been refined by ChatGPT. 😅 . I still felt it lacks precision and refinements.
Channeled
APP
·
·
Here are key ways to further refine your prompt for better precision: 1. Set Clear Context • Start with a system prompt that explains the overall goal and purpose • Define the bigger picture behind the request 2. Structure Your Instructions • Break down instructions into specific, sequential steps • Remove any ambiguity that could lead to incorrect assumptions 3. Be More Explicit About Output Requirements • Define exact formatting requirements • Specify precise details about: • Tone parameters • Word count limits • Style guidelines • Desired data structure 4. Add Structured Data Elements • Consider breaking down the output into specific data points • Make it easier to work with results without additional formatting Remember that prompting at scale requires different principles than chat-style AI interfaces, so focus on structure and specificity rather than conversational elements.
Sources: Prompt Engineering Crash Course [Clay 101] - Clay University Prompt Engineering Crash Course [Limitless Research] - Clay University
Channeled
APP
·
·
Was that helpful?
Channeled
APP
·
·
Hi Sahil S.! This thread was recently closed by our Support team. If you have a moment, please share your feedback: