How to find investment bankers with ChatGPT

A typical version of this starts with one prompt: “Help me find 10 investment bankers at Goldman Sachs.” ChatGPT is genuinely good at that part. The workflow below covers everything that happens after — researching them properly, verifying what ChatGPT told you, writing to them, and not losing track of any of it.

1. Define exactly who you're looking for

"Find me bankers at Goldman Sachs" is too broad — Goldman's Investment Banking Division alone spans half a dozen industry groups and several product groups, and ChatGPT will do a better job the more specific the request. Before searching, decide: which bank(s), which group or coverage area, and what seniority (analysts and associates are more likely to respond than MDs).

Worth deciding early how big this list needs to be — see how many bankers you should actually network with before you start generating names.

2. Find and research bankers with ChatGPT

ChatGPT can't browse LinkedIn for you directly, but it's useful for two things: generating precise search terms so your own LinkedIn search returns the right people, and helping you make sense of what you find once you have names. Here's the exact prompt for the Goldman Sachs example, and the general version underneath it.

If your school has any presence at the bank at all, start with finding alumni from your university — it converts better than a cold list. Once that's exhausted, building a fully cold list covers where to source names with no shared connection at all.

3. Verify before you trust any of it

This is the step people skip and shouldn't. Large language models occasionally generate plausible-sounding names, titles, or deal details that are wrong or outdated — sometimes called hallucination. For recruiting outreach, sending an email to someone who doesn't actually work at the bank, or referencing a deal they weren't on, damages your credibility more than not reaching out at all.

• Confirm the person exists and works there now — check their LinkedIn profile directly, not just ChatGPT's description of them.
• Confirm their current title and group — titles change (analyst → associate) and people move groups; LinkedIn is more current than anything ChatGPT was trained on.
• Don't reference specific deals or achievements unless you can verify them — a wrong detail in a first email is worse than a generic one.
• Confirm the email format separately — see the note on this in finding investment bankers.

4. Generate personalized outreach

Once you have verified names and one real detail about each person, ChatGPT is genuinely good at turning that into a first draft. The output still needs a human pass — but it's a much faster starting point than a blank page.

For five ready-made structures to adapt (alumni outreach, non-alumni outreach, follow-ups, post-coffee-chat notes), see the cold email templates guide.

5. Organize the resulting contacts

By this point you have, scattered across a ChatGPT conversation, a LinkedIn search, and a handful of drafted emails: names, titles, groups, and outreach drafts. The single biggest failure point in this whole workflow is right here — none of that lives anywhere durable, so it gets lost the moment you close the tab or move on to the next bank. Pull it into one list before you send anything, with at minimum: name, firm, group, connection type, and status. If you don't already have a spreadsheet set up for this, the free networking tracker template has these fields built in.

6. Track outreach as you send it

Once emails start going out, the list needs a status per contact — not sent yet, sent, replied, no response. This is the part a chat history genuinely can't do: ChatGPT has no memory of which of the 10 people you've actually emailed by day 4 versus day 11. See how to track investment banking networking for the fields that matter once you're past a handful of contacts.

7. Log the coffee chats you land

When someone says yes to a call, ChatGPT can help you prepare — but the conversation itself, and what you do with it afterward, is on you.

Full guidance on what to capture and when: tracking coffee chats. For more questions than the four above, see the free coffee chat question bank — 77 questions organized by who you're talking to.

8. Build the follow-up habit

The standard rule is a follow-up 7 days after no response, and a value-add follow-up 7-14 days after a coffee chat. Almost everyone knows this rule; almost no one does it consistently, because it requires remembering a date without a system prompting you.

How MarketIQ fits into this process

ChatGPT is the better tool for the generative parts of this workflow: search strategy, drafting, and prep questions. It has no memory of your recruiting pipeline between sessions, no way to remind you a follow-up is due, and no persistent record tying a contact to their status, notes, and history. That's the part MarketIQ is built for — not a replacement for ChatGPT, the layer underneath it:

• Contact discovery with resolved email details, so step 2 doesn't depend entirely on manual LinkedIn search.
• AI-drafted outreach generated per-contact from their actual background — the same idea as the prompts above, built into the pipeline itself.
• A pipeline view — drafted, sent, replied, coffee chat scheduled — for step 6, instead of a spreadsheet or a chat history.
• Coffee chat notes tied to each contact, and automatic follow-up reminders after 7 days, for steps 7 and 8.

None of this requires believing ChatGPT and MarketIQ compete — they handle different halves of the same problem. The guide above works completely on its own; MarketIQ is what keeps steps 5 through 8 from falling apart once you're managing more than a handful of contacts at once.