More Resources
Prompt

Write a personalized follow-up based on what the prospect actually said

Time to build: 1 min
Difficulty: Easy
Tools: Claude ▪ Gmail ▪ Lusha

Images on this page are for illustrative purposes only. Example outputs in this play are illustrative — they reflect the structure, fields, and format of real Lusha connector and Gmail thread output, but were not pulled from a live session. Run the prompt with your own data and connectors to see live results. Personal details in any live examples are masked or abbreviated for privacy.

This Claude prompt pulls the last email thread with a prospect from Gmail, validates the prospect’s current title via Lusha, and drafts a follow-up that picks up exactly where the conversation stopped. The opening line references something specific from the thread. The closing asks one concrete question. Under 100 words, ready to send.

The prompt

Open in Claude Open in ChatGPT 
<context>
I need to follow up with a prospect, but I want the email to reference what they actually said — not a generic check-in. Pull our last email thread, confirm who I'm writing to, and draft a follow-up that picks up exactly where we left off.

My prospect:
- Name: [NAME]
- Company: [COMPANY NAME OR DOMAIN]
- What I'm selling: [PRODUCT / SOLUTION]
- How long since my last email: [DAYS OR "CHECK GMAIL"]
</context>

<task>
1. Search Gmail for the most recent thread with this contact or company:
   - What was the last topic discussed?
   - Who sent the last email — me or them?
   - What was the last specific thing they said, asked, or expressed interest in?
   - Any unanswered question or request on either side?
   - What was the tone — warm, hesitant, neutral?

2. Use Lusha to validate the contact:
   - Confirm they're still at the company in the same role
   - Flag any title change since the last email
   - Pull verified email and direct phone

3. Draft the follow-up email:
   - Subject line: references the specific topic from the thread — not "Following up"
   - Opening: picks up from the last thing they said or asked — references it directly
   - Body: advances the conversation one specific step
   - If they asked for something I didn't send: acknowledge it and include it or a timeline
   - Closing: one specific question or ask — not "let me know if you have questions"
   - Under 100 words

4. Return:
   - Thread summary: what was discussed, who sent last, tone
   - Any outstanding commitments on either side
   - Draft email, ready to send
   - One flag: anything about the thread worth knowing before hitting send
</task>

<constraints>
- The email must reference something specific from the Gmail thread. A generic follow-up is not acceptable output.
- If the contact can't be verified via Lusha, flag it before sending.
- If the last email was theirs and they're waiting on something from me, address that first.
- Under 100 words. Count them.
</constraints>

What you get

The situation: An AE needs to follow up with S.R. at Waverly Digital. Last email was 9 days ago. S.R. asked about a native HubSpot integration. The AE said he’d check and get back to her. He never did. Running the prompt now.

Thread summary

Last thread: 9 days ago · Subject: “RE: Data layer for the new SDR ramp” S.R. sent the last email. Her exact question: “One thing — does this integrate natively with HubSpot, or do we need a workaround?” No reply was sent. Thread went quiet on my side.

Tone: warm — she’s engaged, asked a specific technical question. The silence is mine, not hers.

Outstanding commitments:

  • She’s waiting on an answer about HubSpot integration. I owe her that before anything else.

Contact validated (Lusha)

  • S.R., VP of Sales, Waverly Digital ✓ — confirmed, same role
  • Verified email: s.r@[waverly].com ✓
  • Direct mobile: available (masked)

Draft email

To: s.r@[waverly].com Subject: HubSpot integration — confirmed

Hi Sofia,

Sorry for the delay — you asked about native HubSpot integration and I wanted to confirm before replying. Yes, it’s native. No workaround needed. I can share the connector documentation if it helps your technical eval.

Worth 20 minutes this week to walk through the setup?

[Name]

Flag before sending: She asked a direct question and waited 9 days for an answer. The opening acknowledges the delay — don’t remove it. Skipping it signals you didn’t notice she was waiting.

Thread structure based on live Gmail connector field format. Contact details illustrative — run with your own data to see live results.

Why use Lusha in Claude

A follow-up email written without reading the thread is a guess. Most reps write “just checking in” because the rep genuinely can’t remember what was said. Gmail in Claude reads the actual thread — who sent last, what was asked, what was promised — and Claude drafts a reply from that exact context. Lusha confirms the prospect’s contact details haven’t changed since the last touch. The result is a follow-up that shows the rep was paying attention — which is what gets a reply.

Data drawn from 300M+ verified contacts under GDPR, CCPA, SOC 2, ISO 27701, ISO 31700, and TRUSTe.

FAQ

  • What if the thread is very long?

    Gmail returns the full thread. For long threads, the prompt focuses on the last two or three exchanges — the most recent context shapes the follow-up, not the full history. If something important was said earlier that you want to reference, add it to the context field.

  • What if I sent the last email and the prospect hasn't replied?

    The prompt flags it and adjusts. If you sent the last email and the prospect hasn’t replied, the output is a soft re-engagement rather than a response to something the prospect said. The tone and opening adjust accordingly — the prompt handles this automatically based on what Gmail returns.

  • What if there's no prior thread?

    The prompt flags it as “no thread found” and switches to a first-touch email using Lusha context — current title, any account signal worth leading with. Different output, still useful.

  • Why 100 words?

    A follow-up longer than 100 words rarely gets read in full. The constraint forces the draft to contain one thing — the thread reference — and one ask. That’s all a follow-up needs to do.

  • How is this different from the discovery call follow-up prompt?

    The discovery call follow-up drafts the post-call email from a Zoom transcript — capturing what was said on a call, next steps, competitor mentions. This prompt drafts a follow-up to an email thread — the between-calls outreach that keeps a deal moving day to day.

Related Prompts

Ready to run this?

One data connection. Works in Claude, ChatGPT, your CRM, or any agent you build.

Start for free Read the Docs