The talk-to-listen ratio is the share of a sales conversation during which the rep is talking, compared with the prospect. The most quoted figure, 43% talking and 57% listening, comes from discovery-type conversations: on cold calls, reps who book the meeting actually talk a little more than the prospect, around 55% of the time according to Gong.
That gap explains a lot of contradictory advice. Reps are told to “listen 70% of the time,” then find that a cold call where they stay quiet dies in thirty seconds. This article puts the studies back in order, with their dates and sample sizes, separates cold calls from discovery calls, explains how to measure your own ratio and how to fix it.
What is the talk-to-listen ratio?
The talk-to-listen ratio, or talk ratio, is simple to compute:
rep talk share = rep talk time ÷ (rep talk time + prospect talk time)
Silence counts for neither side. A 4-minute call where the rep talks for 2 minutes and the prospect for 1 minute 30, with 30 seconds of silence, gives 2 ÷ 3.5, or about 57% rep talk time. It is usually written as 57:43.
Managers like this metric because it is easy to measure and reflects a behavior: a rep who dominates the conversation is pitching instead of understanding the need. But a single number says little on its own. It depends on the call type, the stage of the conversation and what fills that talk time.
What does the research actually say?
Nearly every figure in circulation comes from Gong, a conversation intelligence vendor that regularly publishes analyses of its recorded calls. Here they are in chronological order, with what each one actually measures.
| Study | Published | Sample | Call type | Main finding |
|---|---|---|---|---|
| Gong, ideal ratio | February 2017 | 25,537 B2B conversations, 17 customers, H1 2016 | Sales and discovery | 43% rep talk on the highest-converting calls; most reps talk 65 to 75% of the time |
| Gong, cold calls | March 2018 | 90,380 connected cold calls | Cold call | Successful calls (ending in a held follow-up meeting) show more rep talk than unsuccessful ones |
| Gong, cold call stats | August 2021 | ”Thousands” of cold calls | Cold call | 55% rep talk on successful calls; successful calls nearly twice as long |
| Gong, update | March 2025 | 326,000 calls longer than 10 minutes | All long sales calls | 60% rep talk on average; 57% on won deals vs 62% on lost deals |
Three takeaways stand out, checked on Gong’s pages on November 5, 2026:
- 43:57 is a discovery number. It comes from sales conversations where the prospect had already agreed to talk. It is often applied to cold calls, incorrectly.
- On cold calls, talking more is not a flaw. Gong states it plainly in 2018: the talk ratio is higher on successful cold calls than on unsuccessful ones. The 2021 recap gives 55%.
- The average ratio has barely moved in eight years. The 2025 update still finds 60% rep talk time on average, and notes that top performers keep roughly the same ratio whether they win or lose, while lower performers swing from 54% talk time on won deals to 64% on lost ones.
Why the ratio differs between cold calls and discovery calls
On a cold call, the prospect did not ask for it
On a cold call, the prospect does not know who you are or why you are calling. The rep has to carry the conversation: introduce themselves, give a specific reason for calling, show in a sentence or two that they understand the prospect’s situation, then propose a next step. Gong put it this way in 2018: the job of a cold call is not discovery, it is selling the meeting.
The same study flags two telling details. The rep’s longest uninterrupted stretch of talking averages 37 seconds on successful calls versus 25 seconds on unsuccessful ones: the rep who books the meeting takes the time to explain. And the prospect’s turns are shorter on successful calls, 3.5 seconds versus 8 seconds according to Gong. Long prospect monologues on a cold call are often objections or polite ways of getting off the phone.
On a discovery call, the prospect should talk
On a discovery call, the prospect agreed to the meeting to explain their situation. Every minute the rep spends presenting the product is a minute they learn nothing. That is where 43:57 makes sense: the rep asks, follows up, rephrases, and lets the prospect describe the problem in their own words. Our cold call discovery questions and our guide to SPIN Selling give you the material for those exchanges.
Benchmarks by call type
The table below gives working ranges derived from the studies above. They are not measurements: they are starting points to test against your own calls.
| Call type | Rep talk share | What matters beyond the ratio |
|---|---|---|
| Cold call, opener (first 30 seconds) | Majority, and that is normal | A clear reason for calling, then a short question |
| Full cold call | Around 50 to 60% | Turns under 40 seconds, one or two qualifying questions |
| Discovery call | Around 40 to 50% | Open questions, follow-ups, rephrasing |
| Demo | Higher, variable | Leave pauses for the prospect’s questions |
| Negotiation and close | Close to even | Hear the terms before you respond |
What matters more than the percentage
A 55% ratio can hide a great call or a monologue followed by an interrogation. Four signals tell you more.
Length of each turn. A 30 to 40 second explanation works. A two-minute monologue on features never does. On a recording, look for the stretches where nobody interrupts you: they are often the moment the prospect checks out.
Number of questions. More questions are not always better. In its 2025 update, Gong counts 15 to 16 questions on average on calls from won deals, versus about 20 on lost deals. On a cold call, two or three well-chosen questions beat a checklist.
When the prospect gets to talk. The prospect should speak early, even briefly, for example to confirm they are the right person or react to your reason for calling. A rep who holds the floor for the first 90 seconds misses the chance to check they are being heard.
Silence after a question. Leaving two or three seconds after a question, without rephrasing it or answering it for the prospect, often changes the quality of the answer. What you then do with that answer is covered in our guide to active listening in sales.
How to measure your talk-to-listen ratio
With conversation intelligence software
Dedicated call analysis tools separate the voices and calculate talk ratio, monologue length and question count automatically. They usually sit on top of your existing phone system, with a per-user cost. Our guide to conversation intelligence for sales and our roundup of the best call transcription software cover the main options.
With Skipcall recordings and transcripts
Skipcall does not show a talk ratio metric. It records calls, transcribes them and writes an AI summary, which is enough for a manual measurement on a sample. The method takes about an hour per rep:
- Pick ten calls in the Recordings menu, using the Member, Sequence and Duration filters: five that booked a meeting, five that went nowhere, all longer than one minute. The help center explains how in listening to your call recordings.
- Time the rep’s talk while listening: start a stopwatch when the rep speaks, stop it when they go quiet. Divide by the call’s duration minus the silences.
- Read the transcript to count the questions the rep asked and spot the longest blocks of text: those are the monologues. See AI summary and call transcript.
- Compare the two groups instead of chasing an absolute number. If calls that book show 52% rep talk and 30-second turns, while the others show 70% with minute-and-a-half blocks, you know what to coach.
The dashboard rounds this out with total call time and number of conversations, which help connect in-call behavior to results. Recording, transcription and AI summaries are included from the Business VoIP plan, starting at $25 per user per month billed annually, per the pricing page.
How to fix an unbalanced ratio
If the rep talks too much:
- Cut the opener to two sentences, then ask an easy closed question such as “you’re the one who runs outbound over there, right?” Our cold call openers follow that format.
- Drop the product pitch from the cold call: save it for the meeting.
- Count to three after every question before speaking again.
- Review one call a week, note the longest turn, and aim for under 40 seconds.
If the rep does not talk enough:
- Write a specific reason for calling, tied to the prospect’s situation, and say it in full before asking anything.
- Replace rapid-fire questions with a statement followed by a question: “Teams like yours often lose time on X. Is that the case for you too?”
- Prepare 20 to 30 second answers to common objections, starting from our guide to cold call objections.
Either way: do not turn the ratio into a target imposed on the team. A rep who knows their talk time is being scored quickly learns to go quiet at the wrong moment. The ratio is a diagnostic for debriefs, not a grade. For coaching during the call, see our guide to live call coaching.
Talk ratio and call length
Read the ratio alongside call length. According to Gong (2021), successful cold calls last nearly twice as long as unsuccessful ones: a call that keeps going is a prospect who has accepted the conversation. But a long call where the rep talks 80% of the time is not a success. Our article on how long a cold call should last breaks down typical durations by call outcome. And if the problem sits before the conversation, in the hesitation to dial at all, read our guide to call reluctance.
What to remember
Talk-to-listen ratio divides the rep’s talk time by total talk time. Gong’s 43:57, published in 2017 from 25,537 conversations, describes discovery-type calls. On cold calls, successful calls show more rep talk, around 55% according to Gong’s 2021 recap, with longer but controlled turns.
The percentage alone is not enough: monologue length, question count, when the prospect gets to talk and the silence after a question tell you more. The “talk 30%” rule is not backed by any of the studies cited here.
To measure it, conversation intelligence software does it automatically; with Skipcall, you time a sample of ten recorded, transcribed calls and compare those that book with those that do not. Use the ratio to diagnose in debriefs, not to grade reps.
