How to take notes during a client call (without taking notes during a client call)

The fastest way to make a client feel unheard is to look at your keyboard while they talk. The fastest way to forget what they said is to not.

This is the unsolved problem at the heart of every paid meeting: be present, but also remember.

Most advice gives you one of two extremes. Either "be fully present, write down nothing, you'll remember the important parts" (you won't) or "use a structured Cornell-method template and capture every detail" (your client will hate you). Both are wrong for client work.

Here's what actually works.

The four strategies

There are exactly four real approaches. Everything else is a variant.

1. The AI notetaker

Let a transcription service capture everything, then have it summarize after.

How it works: You join the call, an AI assistant joins as a participant (or runs in your browser), it records and transcribes everything. After the call, you get a structured summary, action items, and a searchable transcript.

Pros:

• You're 100% present. No split attention.
• Nothing gets missed — every name, number, deadline is captured.
• Searchable archive across all your calls.

Cons:

• There's a thing in the room listening. Some clients object. Most don't.
• Quality varies wildly by vendor. Bad ones produce summaries you can't trust.
• Privacy: you're sending client conversations to a third-party model.

When to use it: Default for paid client work. The presence-recall tradeoff is too brutal otherwise.

2. The structured template

Pre-fill a template before the call, then check off and fill in fields as you go.

How it works: Open a Notion/doc template with sections like "Context," "Goal," "Constraints," "Next steps." During the call, you tab over briefly and capture into the right field. You're effectively filling out a worksheet.

Pros:

• No third-party recording. Fully private.
• Forces structured thinking — you can only capture what fits the template.
• Easy to skim later. Everything is in the same shape.

Cons:

• Split attention is real. You will miss things.
• Templates ossify — they make you ask only the questions on the template.
• Bad fit for free-flowing or exploratory conversations.

When to use it: Recurring meeting types where the structure is genuinely the same every time. Intake calls, status reviews, scoped check-ins.

3. Post-call dictation

Take zero notes during. Immediately after the call, spend 5–10 minutes dictating what you remember into a voice memo or transcription tool.

How it works: The instant the call ends, before context bleeds out, you walk away from your desk and talk through what happened. A transcription service captures it, then you (or AI) cleans it up.

Pros:

• Full presence during the meeting.
• The act of recall reinforces memory.
• No third-party in the room.

Cons:

• You will lose details. Names, numbers, exact phrasings get fuzzy fast.
• Requires discipline. The temptation to skip "just this once" compounds.
• Doesn't scale if you have back-to-back meetings.

When to use it: Solo practitioners with light meeting loads (2–3/day) and good memory. Falls apart at higher volumes.

4. The smart highlight

Stay present, but tag moments during the call — without writing actual notes.

How it works: You keep a single line of text in front of you: "highlights." When something important happens, you type one word or hit a key. Later, an AI tool (or you) expands those tags into real notes using the recording as context.

Pros:

• Minimal split attention — one keystroke per moment.
• Captures your judgment of what mattered, which a transcript alone can't.
• Works with any recording tool.

Cons:

• Requires a recording in the loop, so you still have the AI tradeoffs.
• Discipline to not over-tag.

When to use it: Hybrid approach. Combine with strategy #1 for the best of both.

Which one is actually best?

For paid client work where the meeting is the deliverable — coaching, advisory, design critique, strategy — the answer is some combination of #1 (AI notetaker) and #4 (smart highlight).

Specifically:

• AI notetaker captures the substrate (everything that was said).
• Highlights capture your judgment (what mattered).
• The combination is what's worth re-reading next month.

Pure transcripts are useless. They're too long, too unstructured, and they don't tell you what you thought was important. Pure presence-and-memory works for the first two meetings with a client and falls off a cliff by month three.

What to look for in a notetaker

Three things actually matter:

1. The summary quality. Ask yourself: would I re-read this two months from now? If the summary is just the transcript with bullet points, no. If it's structured around decisions and next steps, yes.
2. The disclosure UX. Some tools join as a participant ("Otter is now in the meeting"). Some run silently in your browser. Either is fine — but be deliberate about which one you want.
3. The retrieval. If the notes live in a place you'll never look (Slack DM from a bot, an inbox folder), they don't exist. They need to live attached to the person you met with.

That last one is the part most tools get wrong. Otter and Fireflies dump notes into their own database, where you have to remember to go look. The whole point is that next time Alex books a call with you, you should see the last three meetings with Alex automatically.

What we built

MeetingWith does the AI notetaker pattern, but the notes live attached to the person, not to a separate notes app. When Alex books their fourth meeting with you, your prep view shows the three previous summaries.

That's the part that pays off long-term: the difference between "what did Alex and I talk about last month?" being a 30-second search vs. a 5-minute archaeology dig.

It's free, and the whole workflow is in one tool. Claim your handle below.

FAQ

Are AI notetakers legal? Recording laws vary by jurisdiction. In two-party-consent states (most of the US in practice), you must inform the other party. Best practice: always disclose, regardless of jurisdiction.

Can I review the AI summary before sending it to a client? Yes, with any decent tool. Never send AI-generated notes straight to a client without reading them first.

Do clients hate the notetaker bot? Some do. The pattern that works: mention in your booking confirmation that "calls are recorded and AI-summarized so I can be fully present" and offer an opt-out. Most clients appreciate the transparency.

What if I take really good handwritten notes? Keep doing what works. The argument for AI notes isn't "yours are bad" — it's the attention cost. If you can write notes without disengaging, you're in the rare 1%.

Stop choosing between presence and recall. MeetingWith is free — claim your handle below.