The 5-Minute Fix for Post-Meeting Follow-Ups

You just finished a demo, glance at your calendar, and see three more before the end of the day. Somewhere in there, you're supposed to send a follow-up, but by the time you sit down to write it, the specifics have already started to blur, so you end up sending something generic and forgettable.

Sound familiar? We blow right past tasks like this, too small, too routine to question. But now that AI's made things this easier, it's worth building the habit of asking, could AI do this?

Let me walk you through how I solved this for myself, using demo follow-ups as an example. And before you tune out because you don't work in sales, stay with me. Demos are just my world, but if you send any kind of email after a meeting, this applies to you too.

Instead of settling for the AI-slop your meeting tool spits out the second the call ends, take control.

Email is about as low-stakes as it gets. So start there.

Thinking Comes Before Doing

This is the part that's actually strategic thinking, more than deciding which tools to use. Before you automate anything, start with what you're actually doing today, step by step.

Mine looked like this: sit in the meeting, take notes, forget about them until the end of the day, reread them, try to remember what mattered, draft an email, send it.

Once that was laid out, the next question was obvious: what could this look like instead?

Get specific.

Where should the information come from? Do I want the email to be sent automatically, or do I want to review it first? How do I want to be notified when the draft is ready?

Answering those questions is where the real work is. Deciding what's worth recapping, what mattered in that meeting, is human judgment. But retrieving the info and formatting it into an email? That's not judgment. It's just tedious work. Before AI, you had no choice. You had to do it. Now you don't.

Here are the steps that got me there. Follow them, and you'll free yourself up for higher-value work.

Step 1: Hand Your Grunt Work to AI

Start using a meeting note-taker if you aren't already. Granola is my go-to, but use whatever you have access to. Your meeting notes become the input everything else depends on. Get this step right, and everything downstream gets easier.

But it's more than just one less thing to do. It changes what you should be doing during the meeting. You're no longer racing to keep up with every word. Now you can actually read the room.

Don't focus on writing down what was said. Your note-taker already has that.

Write down what you noticed, like the hesitation right before they mentioned budget or the feature that made them lean in.

You're doing less than before. You are.

But that's the point. The note-taker handles the low-value work. You handle what it can't.

Step 2: Define the Structure

Before you connect any tools, decide what the follow-up actually needs to say. Don't just say "summarize the meeting" unless you want a different email every time, which defeats the purpose. Instead, build a template with a few variables that get filled based on each meeting.

For a product demo, mine looks like this: their current process, the gap in it, what we can improve, and, if it came up, a quoted price. Yours might look completely different, and it should, since it depends on what you're trying to accomplish with the email. A status update needs one structure. A team meeting needs another.

See the template I built below as an example. Think about your own use cases and start with the one you spend the most time on.

Hi {{first_name}},

Thank you for meeting with me today. It was great learning more about your {{current_process}} and exploring how {{product_name}} can support your {{desired_process}}.

- Current process: {{current_process}}
- What {{product_name}} can do: {{improvements}}
- Team access: Supports collaboration for {{user_count}} users with {{desired_features}}.
- Pricing: {{quoted_price}}

I'll follow up next week after you've had a chance to think it over.

In the meantime, if you have any questions, send them my way, and I'd be happy to help.

Step 3: Review and Refine

When I mapped out my own process earlier, I decided I wanted to review the draft before it was sent, so that's the version I'll walk through here. But your process might look different, depending on what you actually want.

Once the process is built out, it runs in the background, freeing me up to focus on the parts of my job that actually move the needle. My process looks like this:

1. Get notified when a draft is ready
2. Review it
3. Make any small tweaks if necessary, then send

If I notice the same tweak coming up again and again, say the emails are too long, that's a signal to fix the template itself, not just continuing to edit each output manually. The truth is, most people give up if the output isn't perfect and go back to doing things the old way.

But this is no different than building any new skill. Like working out, when it's hard, that means it's working. Don't get frustrated when you can't "one shot" AI, especially when you're still new to it. Keep tweaking and learning, and you start to recognize patterns.

Once you've nailed down the process, never having to do something again is a pretty awesome reward.

The Real Obstacle Isn't Technical

The hardest part of this isn't the technical setup. It's getting past the voice that says this is your job to do, that you've always written your own follow-ups, that skipping the manual version means you're cutting corners. That voice sounds responsible, but it's worth questioning directly: how much of what you write every week actually requires your judgment, and how much is just retrieval and formatting?

You're probably not the only person in the meeting who could write a follow-up email. But you might be the only one who noticed the CEO's eyebrows raise when you introduced the campaign concept. That's the part worth protecting. Everything else is just busy work. Once you can tell the difference, "Can this be automated?" stops feeling like cutting a corner and starts feeling like the obvious question.

Thinking Has Always Been the Hard Part

Before AI, technical work was the hard part. Now that it's gotten easier, people rush straight to it. Instead of asking "How do we use AI here?" ask, "How well do we understand this work?"

The part that used to take the longest is now often the fastest.

It's probably faster than you think, because the tools are further along than most people realize. I use Claude Cowork, which supports native integrations called Connectors that connect directly to Gmail, Granola, and Slack. Before, you needed to understand API documentation, webhooks, or know how to use Zapier or n8n.

Now?

Just copy the Skill below, start a conversation in Cowork by pasting it, and watch as the heavy lifting is done for you in a few short minutes.