Writing a meeting summary after a client call: what actually works?

You just spent an hour with a client. Three things were decided, two things were promised and one thing was postponed. You hang up, you have ten minutes until the next call, and you think: I will write that up in a bit. In a bit never comes.

A meeting summary is a short record of what was discussed, what was decided and who is doing what. For an agency, that last part is the only bit that really matters. Most articles on this subject are about structure and layout, and that is not where it goes wrong. It goes wrong because the summary never gets written, and when it does, because nobody does anything with it afterwards.

What belongs in a meeting summary?

In short: who was there, when it was, what was discussed, what was decided, and which commitments follow from it with a name and a date attached.

The standard advice is fine as far as it goes. Note the attendees, the date and the subject. Summarise the main points. Number the action items and add an owner and a deadline to each one. Distinguish between an internal summary and one you send to the client.

What that advice leaves out: of those five ingredients, exactly one is the thing your client will hold you to later. That is the commitments. The rest is context. If you have to choose because you only have five minutes, write down only the commitments and skip the summary. A three-line record that exists is worth more than a complete one you never write.

Why do agencies send the summary to the client?

For two reasons, and almost nobody says the second one out loud.

The first is service. The client knows where things stand and does not have to remember anything.

The second is that it protects you. The moment you record what was agreed, and the client reads it and says nothing, the scope is fixed. Three weeks later, when something comes up that was not in there, you have a calm conversation instead of an awkward one. For agencies working on fixed fees, that summary is the cheapest piece of risk management there is. It is not a contract, but it saves a lot of discussion.

That only works if you send it quickly. A summary that arrives four days later gets read with far less attention.

Can you generate a meeting summary automatically?

Yes, and it works better than it did a few years ago, though not as well as the sales pitch suggests.

What goes well: turning speech into text, telling speakers apart, and distilling a summary with action items out of it. For a normal client call in clear language, that produces a usable record.

What goes less well, and what to watch for:

• Jargon and brand names. Your own terminology, client names and project names get misheard regularly. Calls that mix languages go wrong more often than calls in one language.
• Who said what. In a call with four people talking over each other, attribution gets lost. An action item assigned to the wrong person is worse than no action item.
• Implied commitments. "We will have a look at that" is a promise to you and just a sentence to the software. Those half-promises often do not make it into the summary.
• Bad audio. One person on a poor connection or on speaker in a car ruins the whole record.

What that means in practice: let the summary be generated, but read it before it goes to the client. Two minutes of checking is still much faster than twenty minutes of typing, and it stops you from locking in a mistake in your own scope. If you want to move from taking notes to something that happens by itself, meetings become tasks goes into that.

When do you not need a meeting summary at all?

For short calls without a decision. A five-minute call about a schedule that is not changing does not need recording. If you record everything, the archive gets so big that nobody searches it any more, and then you are worse off than when you recorded nothing.

The line that works in practice: if something was decided, promised or changed about budget, timeline or scope, write it down. Otherwise do not.

The real problem starts after the summary

This is where it goes wrong at most agencies, and no article about summary structure covers it. The summary gets written, it goes to the client, and then it sits in a folder. The commitment inside it is on nobody's task list. Three weeks later it turns out Sjef assumed Bep was picking it up.

A meeting summary that is only a document is an archive. One that turns commitments into tasks owned by a person with a date is a working system. That difference matters more than the difference between well and badly formatted notes.

Where Wecooking fits in

The meeting notetaker in Wecooking records the call, writes the summary and puts the commitments straight into tasks assigned to the right person and the right client. That last part is the point: the summary does not disappear into a folder, the commitments are simply in your list the next morning. From there it flows into the project, as described in from won deal to client project. Wecooking starts at 19 euros per month with everything included, and is currently free in beta.

What it does not solve: recording quality and jargon. Here too you should check the summary before it goes out. And if you mostly work solo with short calls and no decisions, this solves a problem you do not have.

If you want to see whether it works for your kind of client calls, you can try it for free and run it alongside your own way of working for a week.