What Is a Good Talk-to-Listen Ratio on a Sales Call?

Ask around and you will get one number: about 43%. The seller should talk a little under half the time, the buyer should do the rest. It gets repeated in sales training decks so often that it has stopped sounding like a finding and started sounding like a law.

It is worth knowing where it came from, because that tells you how much weight it can carry.

Where the 43% figure comes from

The number most people are quoting traces back to Gong, which analysed a large volume of recorded B2B sales calls and reported that on deals that closed, sellers talked around 43% of the time, with the buyer doing the majority of the talking.

Two honest caveats travel with that, and they usually get dropped:

It is a correlation drawn from one vendor’s dataset. The calls are the ones their customers recorded — a particular slice of B2B software sales, not sales in general. Whether it holds for a mortgage broker or an enterprise procurement cycle is not something that dataset can tell you.

It describes what winners did, not what caused winning. Sellers talk less on calls where the buyer is engaged, and buyers are engaged on deals that were going to close. Talking less does not make an uninterested buyer interested; it just stops you burying the signal that they are not.

None of that makes the number useless. It makes it a direction rather than a target.

Why the ratio alone is the wrong thing to optimise

You can hit 43% and still run a bad call. Two sellers with identical ratios can have completely different conversations:

One talks in 20-second stretches, asks a question, listens, responds to what was said.
The other delivers a nine-minute monologue, then says almost nothing for the rest of the call while the buyer fills the silence.

Same ratio. Only one of them was having a conversation. This is why the more useful measurement is your longest unbroken stretch of talking. Past about a minute without a break you have stopped talking with someone and started presenting at them, and the buyer’s attention has almost certainly gone elsewhere.

Two more numbers matter as much as the ratio:

Turn count. How many times control of the conversation actually changed hands. A 45-minute call with eight turns is a presentation. The same call with sixty turns is a discussion.
Average turn length. Short turns mean you are responding to what was just said rather than working through a script.

The right ratio depends on the call

Applying one number to every stage is the most common way this advice goes wrong.

Discovery. This is where talking less genuinely matters, and where the seller should be well under half. The entire job is finding out what is true. Every minute you spend explaining your product is a minute you are not learning whether it fits.
Demo. Legitimately higher. Someone has to explain the thing. The failure mode is not talking too much overall, it is talking for eleven minutes without checking whether any of it landed.
Negotiation. Back down, and closer to even. Silence after you state a price is a tool, not an awkward gap.
Renewal or check-in. Heavily in the customer’s favour. If you are doing most of the talking on a renewal call, you are pitching to someone who already bought.

How to measure yours

Most people guess, and most people guess wrong — sellers reliably underestimate their own talk time, because the parts where you are talking feel productive and the parts where you are listening feel slow.

You can measure it without buying anything. The talk-ratio analyzer runs in your browser: record yourself practising a pitch, or load a call recording you already have, and it reports talk share, longest unbroken stretch, turn count and average turn length. The audio never leaves your machine — it is analysed in the page, with no upload and no transcription, which also means there is no consent problem with a recording of a real customer call.

One honest limitation worth understanding before you read too much into any tool’s number: a single mixed audio track cannot reliably tell two speakers apart by loudness. If your recording has both sides on one channel, treat “speech versus silence” as what it is. Either record only yourself and compare against the call length, or use a recorder that puts each side on its own channel.

How to actually talk less

Knowing the number does not change it. These do:

Ask, then stop. The most common mistake is asking a good question and then answering it yourself after two seconds of silence. Count to four. They will fill it.
Cut the preamble. “So the way we usually think about this is…” before a question is you talking. Ask the question.
Follow up instead of moving on. “What do you mean by that?” is the highest-yield sentence in discovery, and it costs you four words.
Stop confirming that you understood. Summarising back is useful once. Three times is filler.
Give the answer, then stop. When asked a direct question, answer it and stop. The urge to justify the answer is where monologues come from.

Measuring it live, rather than afterwards

Reviewing a call afterwards tells you what you did. It does not help you while you are doing it, and the moment that matters — the fourth minute of an unplanned monologue — has already passed by the time you review the recording.

That is the case for watching it during the call. Craqly’s sales assistant tracks talk ratio live with genuine dual-channel separation, your microphone on one channel and the call on the other, and flags a monologue while it is still happening rather than in a report the following week.

Frequently asked questions

What is a good talk-to-listen ratio for a sales call?

Around 43% talking for the seller is the figure most often quoted, from Gong’s analysis of recorded B2B calls on deals that closed. Treat it as a direction rather than a target: discovery calls should sit well below half, demos legitimately run higher because someone has to explain the product, and negotiation should be closer to even. What is consistently bad regardless of stage is a long unbroken monologue.

How do I measure my talk-to-listen ratio?

Record a call or a practice run and measure the speech. The free talk-ratio analyzer does it in your browser and reports talk share, longest stretch, turn count and average turn length without uploading the audio anywhere. Call-recording platforms report it automatically if you already use one. Guessing does not work — sellers consistently underestimate their own talk time.

Is talking less always better in sales?

No. On a demo somebody has to explain the product, and a seller who says almost nothing while a buyer tries to work it out alone is not doing their job. The reliable signal is not total talk time but whether the conversation is going both ways: how often control changes hands, and how long your longest unbroken stretch was.

What counts as a monologue on a call?

A continuous stretch of talking with no meaningful pause. A common line is 60 seconds — past a minute without the other person speaking, you have moved from conversation to presentation. Craqly uses that threshold to flag monologues live during a call, and the free analyzer reports your longest stretch so you can see how often you cross it.

Does talk ratio matter on discovery calls specifically?

More than anywhere else. Discovery exists to find out what is true about the buyer’s situation, and every minute spent explaining your product is a minute not spent learning whether it fits. If you leave a discovery call having talked more than the buyer, you have usually pitched rather than qualified.