What I Was Trying to Accomplish
I run a small freelance business.
Not a big one. Just me. I do content strategy and copywriting for a handful of clients. I get leads from my website, from LinkedIn, and from referrals. Most of them are warm leads—people who've already expressed interest.
The problem is follow-up.
I know I should follow up. I know it's important. But I'm busy. I'm doing client work. I'm writing. I'm editing. I'm managing projects. Follow-up emails get pushed to the bottom of the list.
When I do follow up, I often write the same thing. "Hi [name], just checking in. Did you have a chance to look at my proposal? Let me know if you have any questions." Generic. Robotic. Not exactly a great representation of my business.
So I decided to test AI. Can it draft follow-up emails that don't sound like a robot? Here's what happened.
What I Used
Tool: A general-purpose AI assistant (free tier)
Task: Draft follow-up emails for five warm leads
Input: Lead name, original inquiry, and a few notes about where the conversation left off
Time available: About 30 minutes
Goal: Follow-up emails I could actually send
The Prompts
I tested three different prompts.
Prompt 1: The simplest one
"Write a follow-up email to [name]. They inquired about my services two weeks ago. I sent a proposal but haven't heard back. Keep it short and friendly."
Prompt 2: The one with context
"Write a follow-up email to [name]. They inquired about [service] on [date]. I sent a proposal on [date]. They said they were interested but needed to check with their team. It's been two weeks. Keep it short. Friendly but professional. Do not sound desperate."
Prompt 3: The one with rules
"Write a follow-up email to [name]. Here's the context: [context]. Rules: Keep it under 100 words. No 'just checking in.' No 'circling back.' No 'touching base.' Use plain language. Be specific. End with a clear question."
The Results
Let me show you what each prompt produced.
Prompt 1 result:
"Hi [name], I hope you're doing well. I wanted to follow up on the proposal I sent a couple of weeks ago. I know you're busy, so no rush. Just checking in to see if you had any questions. Let me know if you'd like to chat. Best, [my name]."
My reaction: Generic. "Just checking in" is exactly what I didn't want.
Prompt 2 result:
"Hi [name], I hope this email finds you well. I wanted to follow up on the proposal I sent on [date]. I know you mentioned you needed to check with your team. I wanted to see if you had any updates. I'm happy to answer any questions. Let me know if you'd like to schedule a quick call. Best, [my name]."
My reaction: Slightly better. It referenced the context. But it was still too long. And "I hope this email finds you well" is a cliché.
Prompt 3 result:
"Hi [name], following up on the proposal I sent two weeks ago. You mentioned you needed to check with your team. Any updates? Happy to answer questions if that helps. Let me know either way. Best, [my name]."
My reaction: Much better. It was short. It was specific. It didn't use clichés. It ended with a clear question.
What the AI Did Well
It was fast. The drafts took about 10 seconds each. I got five drafts in under a minute.
It followed the rules. When I gave it specific rules (no clichés, under 100 words), it followed them.
It gave me a starting point. When I had a draft, I could edit it. When I had a blank page, I had to create.
It was consistent. All five drafts had a similar tone and structure.
What Still Needed My Attention
It missed context. The AI didn't know the lead. It didn't know the relationship. It didn't know if the lead was warm or cold, friendly or formal.
It missed the personal touch. I added a personal detail to each email. "Hope your trip to Denver went well." "Hope the new hire is settling in." The AI didn't know those details. I did.
It missed my voice. The drafts were fine. But they didn't sound like me. I had to rewrite the opening and closing.
It missed the nuance. Some leads need a gentle nudge. Some need a deadline. Some need a question. The AI treated them all the same.
It missed the judgment. One lead was a good fit. One was not. One needed more time. One needed a different offer. The AI didn't know the difference.
The Math
Without AI: Writing five follow-up emails from scratch would have taken about 45 minutes.
With AI: Writing prompts, generating drafts, and editing took about 30 minutes.
Time saved: About 15 minutes. Not huge. But the drafts were more consistent than what I usually write.
What I'd Do Differently
I'm still refining my approach. Here's what I'm considering.
Adding a relationship line. Something like "This is a warm lead" or "This is a cold lead." That might help the AI adjust the tone.
Adding a personal detail line. Something like "Include this personal detail: [detail]." That might help the AI include the personal touch.
Adding a "do not" list. Something like "Do not use: just checking in, circling back, touching base, I hope this email finds you well." That might prevent clichés.
Using a different tool. The free tool is fine, but it doesn't remember context. A paid tool might. I'm testing.
My Question for You
How do you handle lead follow-up?
Do you use AI? Do you use templates? Do you write each one by hand?
I'm curious what works for other people. I've been using this approach for a few weeks. It's not perfect, but it's better than sending generic follow-ups.
