The best AI prompt for a sales follow-up email gives the model the specific context of what already happened — not just “write a follow-up email” — and asks for one short, single-purpose message instead of a generic template. Most reps lose replies not because they didn’t follow up, but because the follow-up reads like every other follow-up in that prospect’s inbox: vague, long, and asking for nothing in particular.
Why Do Most Follow-Up Emails Get Ignored?
Gong’s analysis of over 300,000 sales emails found that emails with a specific, deal-stage call-to-action got a 37% meeting-booking rate, compared to just 15% for a generic “let me know if you’re interested” close. The same research found that certain phrases actively hurt results: “Thoughts?” lowered meeting bookings by 20% despite getting more replies, “Just following up” lowered bookings by 5%, and “Never heard back from you” cut bookings by 14% — even though all three feel like safe, polite defaults. If you’re prompting an AI to write these emails, the fix is to explicitly tell it to avoid this exact language and to always specify one clear next step.
What Should You Put in the Prompt Itself?
Give the model real inputs, not a vague ask: the prospect’s name and role, their company, what was discussed or sent previously, how long it’s been since the last touch, and the one thing you want them to do next. Gong’s data also found that emails between 30 and 150 words significantly outperformed shorter ones for booking meetings — so it’s worth telling the model to write within that range rather than trimming to a couple of curt lines.
Five Prompt Templates for Common Follow-Up Scenarios
- No response after a demo: “Write a 60-100 word follow-up email to [name] at [company], sent 3 days after a product demo where we discussed [specific pain point]. Reference one specific moment from the demo, restate one concrete benefit tied to their pain point, and end with a single clear CTA to book a 15-minute call this week. Avoid ‘just checking in,’ ‘thoughts?’, and ‘following up.'”
- No response after a proposal: “Write a short follow-up email to [name] two days after sending a pricing proposal for [product/service]. Acknowledge that proposals take time to review, offer to answer one likely question (about [pricing/implementation/timeline]), and ask directly whether there’s anything blocking a decision. One CTA only.”
- Re-engaging a cold, dormant lead: “Write a re-engagement email to [name], who went quiet [X months] ago after [brief context]. Open with a specific, relevant update (a new feature, a case study, a changed price point) rather than an apology for the gap. Keep it under 80 words and end with a low-friction CTA, like a yes/no question.”
- The ‘breakup’ email: “Write a break-up email to [name] after [3-4] unanswered follow-ups. Keep it under 5 sentences, don’t guilt-trip them, state plainly that you’ll stop reaching out, and leave the door open with one easy way to restart the conversation later.”
- Post-conference or event follow-up: “Write a follow-up email to [name], who I met at [event name], referencing our specific conversation about [topic]. Suggest one concrete next step (a short call, a resource relevant to what they mentioned) and keep the tone casual, like a continuation of the in-person conversation rather than a cold pitch.”
How Many Follow-Ups Should You Send Before Stopping?
Gong’s research found reply rates drop steadily with each additional touch — from 31% at the second follow-up down to about 14% by the eighth. A common structure, outlined by Close.com for “breakup” sequences, is an initial email, a follow-up the next day, another two days later, then a final breakup email four to five days after that — roughly three touches before you close the loop. There’s no universal rule, but diminishing returns after several unanswered emails is well documented, which is exactly when a clear breakup email, not silence, tends to perform best.
These same prompting habits — giving the model real context instead of a vague instruction, and specifying tone and constraints up front — carry over directly to cold email prompts and to AI copywriting more broadly: the model writes a noticeably better email when it knows who it’s writing to and why, rather than being asked to fill in a generic template.
Frequently Asked Questions
How many follow-up emails should I send before giving up on a lead?
There’s no fixed number, but Gong’s data shows reply rates decline steadily with each additional email, from roughly 31% at the second follow-up to about 14% by the eighth. Many sales teams stop around three to five touches, closing the sequence with a direct “breakup” email rather than simply going silent.
What should I tell an AI model to avoid when writing a follow-up email?
Explicitly rule out filler phrases like “just checking in,” “thoughts?”, and “following up” — Gong’s research found these phrases correlate with lower meeting-booking rates even when they don’t hurt open or reply rates. Also tell the model to avoid guilt-tripping language in breakup emails, since that tends to read as passive-aggressive rather than professional.
How do I get an AI-written follow-up to sound personal instead of robotic?
Feed the model specific, real details from the actual interaction — a moment from the call, the exact objection raised, the specific document you sent — instead of asking it to write generically about “your product.” A follow-up that references something concrete from the actual conversation reads as personal by default, without needing extra instructions about tone.
Featured image: “Business man and woman handshake in work office” by perzon seo, licensed under CC BY 2.0.
