In short
A talk-to-listen ratio compares how much you spoke in a conversation with how much the others did. Tools measure it either by speaking time or by words spoken. It is most useful in sales discovery, interviews and coaching, as a prompt to ask more and listen longer.
A talk to listen ratio compares how much you spoke in a conversation with how much everyone else did, measured either by speaking time or by words. It is used most in sales discovery calls, interviews and coaching, where the point of the conversation is to learn from the other person, and a high share of talking on your side is a sign you were telling rather than asking.
This guide covers what the ratio is used for, the two ways tools measure it, what makes the number wrong, and how to use it without turning it into a target. It is a coaching aid for your own calls. Measuring other people's speech in meetings raises its own questions; the section near the end covers those.
What the talk-to-listen ratio is for
The ratio is useful where listening is the job:
- Discovery calls. The aim is to understand the other person's problem in their words. If you did most of the talking, you probably learnt less than you could have. Discovery call questions covers questions that keep the other side talking.
- Interviews. Research interviews and hiring interviews both go better when the interviewer talks less.
- One-to-ones. A manager who does most of the talking in a one-to-one is running a status meeting.
- Coaching. Comparing calls over time shows whether a habit, such as waiting longer before answering, is changing.
It is much less useful in conversations where talking is the point: a demo, a presentation, a training session. A high share there is expected.
Two ways to measure it
Tools measure the ratio either by time or by words, and they give different answers for the same call.
| Method | What it counts | What it misses |
|---|---|---|
| Speaking time | Seconds each person's voice was detected | Speed: a fast talker says more in the same time |
| Words | Words each person said in the transcript | Pauses, and speech the recogniser missed |
By time is closer to "who held the floor". A slow, careful speaker and a fast one can say very different amounts in a minute, but time treats the minute the same.
By words is closer to "who said the most". It depends on the transcript: words the recogniser missed are not counted, and a long pause before an answer counts for nothing.
Neither is wrong. What matters is to compare like with like: the same method, the same tool, the same kind of call.
What makes the number wrong
Any talk-time figure depends on knowing which voice is whose. Three things commonly break that:
The other side leaks into your microphone. On speakers rather than headphones, your microphone picks up the other person's voice from your own speakers. A tool that attributes by microphone can then credit their words to you, which inflates your share. Echo in a meeting recording explains why this happens and how to prevent it. Headphones fix most of it.
Speakers are not told apart. A tool that separates voices after the fact can split one person into two or merge two people into one. With several people on the other side, that moves words between them, and between them and you if your voice is involved.
Crosstalk. When two people speak at once, attribution gets harder, whatever the method.
A recording that keeps your microphone and the call's audio on separate tracks avoids most of the attribution problem for you versus them: your side is whatever came in on your microphone, their side is whatever the Mac played. It still needs the leak from your speakers handled.
Using it without chasing a number
You will find benchmark figures quoted online for the "ideal" ratio. Treat them with care: they rarely say how they were measured, on what calls, or by what method, and a figure measured by time does not transfer to one measured by words. A more reliable way to use the ratio:
- Measure a few of your own calls of the same kind. That is your baseline.
- Pick one habit to change. Asking a follow-up question before answering; pausing after the other person finishes; cutting the product overview short.
- Measure again after a few weeks with the same tool and method.
- Read the transcript, not just the number. A lower share means little if you talked less but still missed the problem. The transcript shows where you interrupted, where a question went unanswered, and where the other person said the thing that mattered.
A ratio is a prompt to look at the conversation, not a score for it.
Talk time and the other people in the call
Measuring your own share of a conversation is a self-coaching exercise. Measuring and reporting other people's speech, as team conversation-analytics platforms do, is different: it is a record about them, and they should know it is being kept. Tell people when a call is recorded, whatever the tool; how to ask for consent to record has wording. Team call analytics, coaching dashboards and forecasting are a different kind of tool from a personal notetaker; Gong vs a personal notetaker sets out the difference.
How Notey shows it
Notey records your microphone and what the Mac plays on two tracks, so each transcript line is labelled You or Them by the track it arrived on. After a meeting, the Notes tab shows meeting stats: Length, Words, Your words and Their words.
Two things about those figures:
- They are a share of the words, not of talking time. A slow speaker who said less will show a smaller share than their minutes on the call suggest.
- Lines your microphone picked up from your own speakers are not counted, and the stats say how many were left out. Without that step, the leak described above would count the other side's words as yours and make your share look higher than it was.
The figures come from the transcript on your Mac. Nothing is sent anywhere to work them out. It is one of several habits that make meetings shorter and more useful; the others are in how to reduce meetings.
Frequently asked questions
What is a talk-to-listen ratio?
The share of a conversation you spent talking compared with the share the other people did. It is usually shown as a percentage or a ratio, such as 40:60.
What is a good talk-to-listen ratio?
It depends on the conversation. In a discovery call or an interview, where your job is to learn, you want the other side talking most of the time. In a demo or a presentation, you will talk more. Compare yourself with your own past calls of the same kind rather than a universal target.
How is talk time measured?
Either by the seconds each person spoke or by the number of words each person said. The two give different numbers, because people speak at different speeds and pauses count only in time.
How can I tell how much I talked in a meeting?
Use a recorder that knows which voice is yours, and read its talk-time or word-share figure. A recording with your microphone on its own track makes that attribution reliable.
Why would a talk-time figure be wrong?
The usual causes are the other side's voice leaking into your microphone from speakers, speakers who were not told apart, and crosstalk. Each moves words or seconds to the wrong person.