The "Find the Context First" Prompt That Closes Open Loops Before They Cost You a Deal

What you'll learn in this article:

• Why drafting follow-ups from memory is the slow, error-prone way
• The exact prompt that makes Gemini Personal Intelligence retrieve commitments before drafting
• How to adapt the prompt for a single client versus your whole inbox
• The one setup step that makes this work at all
• Three task types where this pattern saves the most time

The most expensive dropped ball in business is the one you never see. A client asks for a revised number. You say you will send it. Then a fire comes up, the thread scrolls off the screen, and three weeks later the deal has gone quiet for a reason you cannot reconstruct. Nobody decided to drop it. It just slipped through the gap between intention and inbox.

The usual defense is to scroll back through your email before every reply, reconstructing what you promised. That works, barely, and it costs twenty to thirty minutes a day across all your relationships. It also fails exactly when you are busiest, which is when commitments get dropped in the first place.

Gemini Personal Intelligence, launched January 14 for AI Pro and Ultra subscribers, makes that defense automatic. With Gmail connected, one prompt reads your actual sent and received mail, surfaces every open commitment, and drafts the email that closes the loop. The retrieval you used to do by hand happens in seconds, grounded in what you actually said.

Why Drafting From Memory Fails

When you write a follow-up from memory, two things go wrong. First, you forget items. The commitments you remember are the recent and the loud ones; the quiet promise from two weeks ago is exactly the one that slips. Second, you get details wrong. You misremember the number you quoted, the date you named, or which of two clients you told. A confident follow-up with the wrong detail is worse than no follow-up.

Reading the thread first solves both, but it is slow and you skip it under pressure. The fix is to make the AI do the reading. Personal Intelligence grounds its answer in your real Gmail, so the draft references real promises with real dates instead of your best guess. The errors disappear and the search time collapses to nothing.

The one prerequisite is setup. Personal Intelligence is off by default. In the Gemini app on an AI Pro or Ultra account, enable it and connect Gmail. That single step is what gives the prompt below something real to read.

The Prompt

This is the full prompt. It does three jobs in order: retrieve, organize, and draft. Paste it as-is and replace only the bracketed client or person name.

FIND THE CONTEXT FIRST PROMPT

Using my connected Gmail, list every open commitment I have made to [Client or Person] in the last 60 days. For each one, give me: what I promised, the date I said it, whether I have followed up, and what the next action should be. Then draft a short status email I can send today that closes the loop on the items still open. Keep the email under 150 words and plain in tone.

What you get back is a structured list of what you actually promised, with dates, separated into followed-up and still-open, and a ready-to-send email that addresses only the open items. The draft is grounded in your real messages, so the dates and figures are correct. You review, adjust a sentence, and send.

Adapting the Prompt

For a whole-inbox sweep instead of one relationship, widen the scope and shorten the window:

WHOLE-INBOX SWEEP PROMPT

Using my connected Gmail, list every thread from the last 7 days where someone is waiting on a reply or an action from me. For each, give me who it is, what is outstanding, and how many days it has been waiting. Rank by urgency and draft the three most urgent replies, each under 120 words.

For a single high-stakes deal, tighten the focus and ask for risk:

DEAL RISK PROMPT

Using my connected Gmail, reconstruct the timeline of my conversation with [Client] over the last 90 days. Flag anything I committed to that I have not delivered, and tell me where this deal is most at risk of going quiet. Then draft a re-engagement email that addresses the highest-risk gap.

Three Task Types Where This Saves the Most

Client follow-ups. This is the core use. Grounding the draft in real promises removes errors and the search step. Typical time saved: 20 to 30 minutes per follow-up session.

Weekly pipeline reviews. Run the whole-inbox sweep every Monday to catch every deal waiting on you before it stalls. Typical time saved: 30 to 45 minutes per review.

Recovering quiet deals. Use the deal-risk version to reconstruct a relationship that has gone silent and find the gap that caused it. The cost of catching one stalled deal early is far higher than the minutes saved.

Action Steps Summary

• Enable Personal Intelligence and connect Gmail: In the Gemini app on an AI Pro or Ultra account, turn it on and connect Gmail. This is the one prerequisite.
• Run the per-client prompt before any follow-up: Paste the find-the-context-first prompt, replace the name, and review the grounded draft.
• Run the inbox sweep every Monday: Use the 7-day version to surface everything waiting on you and draft the three most urgent replies.
• Use the deal-risk version on quiet accounts: Reconstruct the timeline and find the gap before the deal goes cold.
• Send, do not just read: The drafts are ready to go; the value is in closing the loop today, not filing the insight for later.