# AI for Sales Prompt Pack Prompts from the AI for Sales webinar, grouped by stage of the sales funnel. They are lightly cleaned up from what was said live. Replace anything in [brackets]. ## 0. Privacy settings to check first - ChatGPT: click your name > Settings > Data controls > turn off "Improve the model for everyone". - Perplexity: Account > Settings > Personalization > turn off "AI data retention". - Claude: Settings > Privacy > turn off "Help improve Claude". ## 1. Lead sourcing with a $20 subscription Use deep research (ChatGPT, Claude Research, Perplexity or Gemini). Find the decision makers at one company (paste the company URL): ``` [company URL] Can you help me find the names of the decision makers in this company, please? ``` Then reuse an email pattern: ``` Is there an email that you found? ``` ``` If [name]'s email is [email], what would be the emails of the rest of the decision makers? ``` Find many companies at once: ``` I want to find the names of the CEOs of consulting firms, marketing agencies and law offices in [Paris / the US / Mexico]. ``` Find local businesses (ChatGPT agent mode): ``` I want to contact the flower shops in [Paris]. Go to Google Maps and bring me back a list of 20. ``` Research one person before a meeting, to find things you have in common: run deep research on the person's name. ## 2. Lead sourcing, cleaning and research with Claude Code + Apify ``` What Apify actors are in your tools folder? ``` ``` Let's use Leads Finder. I want to find some people. ``` (Answer its questions: job level, company size, location, industry.) ``` Can you download that data and clean it up for me, please? Remove duplicates and people with no email. ``` ``` Now I need you to use the Perplexity API and check if all of these people appeared in some media coverage recently. Please make sure the script is fast. ``` ``` Please create an HTML dashboard for me to visualize the three databases. ``` ``` Now let's save this workflow, and every time I tell you to get some leads, you are also going to do all of these steps afterwards. ``` ## 3. Intent from LinkedIn comments Save a LinkedIn post with lots of comments, then: ``` Please run the comment scraper from Apify on this link: [post URL]. Check your tools. ``` ``` Now I want you to use the Anthropic API and analyze all of the comments. Which people were curious? Make sure it's fast. ``` ``` Create an HTML dashboard and open it. ``` ## 4. Enrich an exported LinkedIn contact list ``` Use the Apify LinkedIn Profile Scraper (No Cookie) with my API key to enrich this list of profiles. ``` ``` Now I need you to qualify all of those people against my ICP: [describe your ideal customer]. ``` ## 5. After a cold email campaign (Instantly) ``` Run the Instantly campaign analytics on the [campaign name] campaign. ``` What my workflow does with the results: - Non-repliers go to a list marked "Contact in three months". - Out-of-office replies go to the CRM with a follow-up date for when the person is back, plus any phone number or LinkedIn profile found in the email thread. - Positive replies go to the CRM with next steps. - Opt-outs go to a block list and bounces go to a bounced list. ## 6. From meeting transcript to proposal ``` Do you have access to my Fireflies account? Check your tools folder. ``` ``` Can you check my last 10 meetings? Look for one with [name]. ``` ``` I want you to analyze the entire conversation and find the key pain points from [name]. ``` ``` Let's create a business proposal for [name] using the exact same pain points that he mentioned. ``` Tip: if Claude Code reaches for the wrong tool, be direct: "Use the API." For free, clean-looking PDF proposals, have it format them with Typst. ## Rule to keep Automate sourcing, cleaning, research and reporting. Don't automate LinkedIn messages. That part stays human.