Module 3 · Client Management & CRM

Prompt it: better client care today

One idea carries this whole lesson: the leap from “AI writes my emails” to real leverage is context. Give the model the thread, the notes, the history — and the quality of what comes back tracks the quality of what you fed it. Every prompt below is built on that move.

Walk in caught up: the pre-meeting briefing

This is the highest-value prompt in the module. Before any client meeting, gather whatever you have — old emails, call notes, the intake form — and ask for a one-page brief. Two minutes of reading, and you walk in as the professional who remembers everything.

I have a meeting with a client coming up and I want to walk in fully caught up.

Below is everything I have on them — notes, past emails, anything relevant. Build me a one-page prep brief with four sections:

1. Who they are and what they care about, in three or four lines.
2. Where things stand — what we last discussed, what was decided, and what's still open.
3. What they're likely to ask me, based on what's unresolved.
4. Three smart things I could raise that would show them I've been paying attention.

Keep it tight enough to read in two minutes. Flag anything that looks like a commitment I made and haven't delivered on yet.

My business: [YOUR BUSINESS TYPE — for example: "residential landscape design studio"]
The client and the meeting: [WHO AND WHY — for example: "the Hendersons, mid-project check-in on their backyard renovation"]

Here is everything I have (messy is fine):
[PASTE YOUR NOTES, EMAILS, AND HISTORY]

Follow-ups that don’t sound like mail-merge

The check-in email you keep meaning to send is a two-minute prompt: tell the AI what you did for the client, one detail you remember about them, and the tone you want — warm, brief, not salesy. “We finished their kitchen remodel in the spring; ask how it’s holding up” produces a note that reads like you wrote it, because the remembering — the part that makes it personal — came from you. Ask for it under 120 words; short reads as genuine.

Onboarding: the first two weeks, designed once

A new client’s first two weeks set the tone for the whole relationship, and it’s the same two weeks every time — which makes it perfect AI territory. Three drafts to ask for, one sitting each:

1. Day 0

   Welcome email

   What happens next, in plain terms.
2. Day 2–3

   The ask

   What you need from them, and exactly how to send it.
3. Week 1

   How we communicate

   Channels, hours, and response times.
4. Week 2

   The what-to-expect one-pager

   Answers the first ten questions before they're asked.

• A three-email welcome sequence. Email one: welcome, and what happens next. Email two: what you need from them and how to send it. Email three: how you communicate and how to reach you. Tell the AI who you are and what clients worry about at the start — it drafts, you make it yours.
• Your intake form. Describe your business and ask: “what should I ask every new client up front so I never have to ask twice?” It will suggest questions you’ve been asking mid-project for years.
• A what-to-expect one-pager. The welcome packet that answers the first ten questions before they’re asked — timelines, communication, billing, the works.

The hard conversations

A price increase, a boundary, a mistake you have to own — these are the messages owners put off for weeks, and the delay costs more than the conversation. AI is a remarkably good rehearsal partner: it gives you the calm version while you’re still in the heated one.

Help me handle a hard client conversation. I want to be firm, warm, and done apologizing for the business needing to work.

The situation: [WHAT'S GOING ON — for example: "a good four-year client keeps adding small requests outside our agreed scope, and I need to start charging for them"]
The outcome I need: [YOUR GOAL — for example: "new requests get quoted separately, and the relationship stays warm"]
What I'm worried about: [YOUR FEAR — for example: "I don't want this to read as greedy after four good years"]

Draft three things:
1. An email version — short, human, direct. No corporate stiffness, no over-apologizing.
2. A three-bullet script I can glance at if they call to talk it through.
3. The version of this message I should NOT send — so I can see where the line is.

Acknowledge the relationship, state the change plainly, and end on what stays the same.

Ask for the review at the right moment

There’s a moment in every engagement when the client is happiest — the case closes, the project ships, the problem disappears. That’s when the review request goes out, and AI will draft it in both flavors: a short email and an even shorter text, each naming one specific thing the client could mention and neither sounding needy. Save the good version as a template and the ask becomes a habit instead of an awkwardness.

Watch this step

Here's the follow-up-email move on screen: prompting ChatGPT for a personalised client reply, one of this lesson's headline use cases. It’s from an independent creator — credited below, so go give them a follow.