How do you coach a rep to talk less and listen more on calls?

Two coaching paths: fix the behavior or fix the system around it

There are two credible ways to get a rep to listen more, and managers routinely pick the wrong one because the symptom looks identical either way.

Path A — behavioral coaching. You treat over-talking as a personal habit. The rep knows the product, has decent rapport, but fills every gap with words. You drill pauses, question stacks, and statement-to-question conversion until silence stops feeling like failure. This is one-to-one work: recorded call reviews, role-play, a weekly ratio check. It's slow but durable, because you're rewiring what the rep does under pressure, not what they know.

Path B — system coaching. You treat over-talking as a rational response to a broken environment. If your discovery script is a 22-question compliance checklist, the rep has no choice but to march through it — every question is a statement in disguise, and the buyer becomes a respondent instead of a participant. If the territory is cold and the "discovery call" is really a first-touch pitch, there's nothing to discover and the rep fills the vacuum. If comp pays on demos booked, the rep will steer to demo-booking language in minute four. Fixing any of those changes talk ratio across the whole team overnight, without a single 1:1.

The distinction matters because Path A on a Path B problem is the most common failure in sales coaching. You spend six weeks drilling a rep on pauses, their ratio improves in role-play, and it snaps back the moment they open the real discovery template — because the template is the thing generating the monologue. The rep isn't resisting you. They're obeying a stronger instruction.

There's a useful diagnostic here that costs nothing: look at the talk-ratio distribution across the whole team, not just the rep you're worried about. If one rep sits at 70% while the team median is 48%, that's an individual habit — Path A. If the team median is 68% and your worry-rep is merely the worst of a bad set, you have a system problem and coaching one person is theater. RevOps can pull that distribution in a few minutes from any conversation-intelligence platform, and it should be a standing chart in the weekly forecast review rather than something you request ad hoc.

The honest answer for most teams is that both paths run in parallel. You fix the discovery framework at the team level while drilling the individual moves with the two or three reps who are outliers even after the framework improves. What you should never do is start with Path A because it's easier to schedule.

Choosing the right path before you spend a single 1:1

Diagnosis takes two recorded calls and about twenty minutes. Do it before you book coaching time, because the wrong opening conversation costs you credibility with the rep — a rep who's told to "listen better" when the real problem is a broken script will correctly conclude you didn't look closely.

Work through four root causes, in this order:

Skill. The rep genuinely doesn't know how to build a follow-up question off an answer. Listen for the tell: the buyer says something rich — "onboarding has been a mess since the reorg" — and the rep responds with the next scripted question instead of "what changed in the reorg?" That's not arrogance, it's a missing motor pattern. Skill gaps move fast, usually one to two weeks of drilling.

Will. The rep can do it and doesn't. Silence feels like losing control of the call, so they fill it. Listen for the rescue: they ask a good question, the buyer pauses for one second to think, and the rep jumps in with "…or is it more of a budget thing?" — closing off the exact answer they wanted. This is a mindset reframe plus desensitization, not information. Adding more training content makes it worse.

Knowledge. The rep over-explains because they're insecure about the product and feature-dumping is where they feel safe. The tell is direction: their monologues are almost entirely about your product, not about the buyer's world. Fix product confidence first — a certification, a shadow week with a solutions engineer — and the talk ratio drops as a side effect.

System. The framework, the territory, or the comp plan is generating the behavior. Tells: every rep on the team has the same shape of monologue at the same minute mark, or the monologue starts exactly where the script says "position the platform."

One nuance worth naming: the most common cause on modern remote-heavy teams is anxiety-driven filling on video. Video strips out the micro-cues that tell you a person is thinking rather than disengaged — no shifted posture, no glance away, just a still face and dead air on a 200ms-latency connection. Reps who came up entirely on video calls have never learned that a three-second pause reads as thoughtful in a room. That's a will-plus-skill blend, and it responds better to desensitization drills than to any amount of explanation.

Run this once per rep, write the answer down, and re-check it at day 30. Diagnoses drift — a skill gap that gets fixed can reveal a will gap underneath it that was invisible while the rep had a legitimate excuse.

The numbers behind each path, and what "good" actually means

Set the benchmark by call type or you'll coach reps into passivity.

The widely-cited conversation-intelligence research on discovery calls puts top performers a little over 40% talk time — Gong Labs' analysis landed near 43% — against a common default in the 65–75% range. Treat that as a direction of travel, not a target to hit on a single call. Three things follow from it:

The goal is not silence. A rep at 25% on a discovery call is usually not listening well; they're passive, letting the buyer ramble through irrelevant territory, and failing to steer. Good discovery is high-question, high-direction, low-word-count. The shape you want is many short rep turns and few long ones — which is why longest-monologue is a better coaching metric than the ratio itself.

Ratio varies enormously by call stage. Discovery should sit under 50%. A demo legitimately runs 55–65% because you're presenting. A negotiation call often swings back down to 45% or lower, because the person who talks least in a negotiation usually learns most. A technical validation call with a solutions engineer present will look bizarre on any ratio chart because the SE dominates. Setting one universal number across all call types is the fastest way to make the metric worthless, and it's the single most common RevOps implementation mistake with conversation intelligence.

Sample size beats precision. A single call's ratio is noise — one chatty buyer swings it 15 points. Trend on a rolling four-call or two-week window. Don't coach off Tuesday's number.

Now the cost-and-timeline side, which is where the two paths diverge sharply.

Path A economics. Budget roughly 45 minutes a week per rep: 20 minutes of call review prep by you, 25 minutes in the 1:1. Over a 90-day cycle that's about 9 hours of manager time per rep. With eight direct reports you cannot run this on everyone simultaneously — pick two or three, run the cycle, then rotate. Expect visible ratio movement in one to two weeks, layered questioning by week six, and stability under pressure around day 75–90. Win-rate movement, if it comes, shows up on your lagging dashboard around day 60–90, and it will be confounded by everything else happening in the quarter. Don't promise it upward.

Path B economics. Rewriting a discovery framework is one or two working sessions plus a week of pilot calls, and it moves the entire team's median at once. If you have six reps and the framework is genuinely bad, that is a dramatically better return per hour than six parallel coaching cycles. The catch: it requires admitting the framework you approved is the problem, and it needs RevOps to actually retire the old template from the CRM and the call-prep tooling — otherwise half the team keeps using the version bookmarked in their browser.

When you have no conversation-intelligence tool. You can approximate a talk ratio manually: play the recording and tally 10-second blocks by speaker, then divide. A 30-minute call takes about 10 minutes to score by hand. It's tedious enough that you won't do it weekly, but doing it once with the rep watching is often more persuasive than any dashboard — the tallying itself is the intervention. For ongoing tracking without tooling, count questions asked per call instead. It's a single number, the rep can self-report it, and it correlates well enough with the ratio to coach against.

One more number worth tracking, borrowed from customer-success practice: longest uninterrupted buyer stretch. If your rep's longest customer story grows from 40 seconds to two and a half minutes over the cycle, the behavior changed — regardless of what the aggregate ratio says.

Installing the behaviors: the sequence that actually holds

The mechanics matter more than the philosophy. Here's the order that survives quota pressure.

The coaching conversation itself. Run it in a 1:1 with a recording open, using a GROW structure — Goal, Reality, Options, Will — and keep your own talk ratio low, because the rep will notice if you monologue about listening.

Open on the *goal*, not the flaw: "I want your discovery calls to be ones where the buyer does most of the talking, because that's where deals get made. Under 50% ratio by the end of the month — does that make sense as a target?"

Then *reality*: play sixty seconds of a call where they ran over a buyer, and ask one question and stop — "What do you notice about airtime there?" Wait. Do not rescue the silence; you're modeling the exact skill. Follow with "what did we not learn in that moment because we were talking?" The objective is for the rep to name the cost. If you name it, it's criticism. If they name it, it's insight.

*Options*: "What's one thing you'd do differently on the next call to open room?" If stuck, offer three and let them choose one — pause three seconds after every question; ask "what else?" instead of pitching when they finish; convert your next statement into a question. Let them pick. Choice drives adherence.

*Will*: lock it with a date. "Next two calls, three-second pause after every discovery question. I'll pull both and we look at the ratio Friday." Coaching without a scheduled review is a conversation, not coaching.

The drills. Talking less is a motor skill, so practice it as one.

• *Three-second pause drill.* Rep asks a discovery question, then counts silently — one-Mississippi, two, three — before saying anything. You play a buyer who genuinely pauses to think. The point is desensitization: after a week the pause stops registering as a crisis. Pair it with a physical cue — a sticky note reading "Wait 3" on the monitor edge works better than it has any right to.
• *The "what else?" stack.* Give the rep one opening question. Every buyer answer must be met with a deepening follow-up — "tell me more," "what's driving that," "what else" — for five exchanges before they're permitted a single statement. Miserable and effective.
• *Statement-to-question conversion.* Hand them five things they habitually say as statements ("our platform integrates with Salesforce") and require conversion on the spot ("how are you handling the Salesforce sync today?"). Run it cold, repeatedly, until it's reflexive.
• *The mute drill.* On an internal practice call the rep is muted for two minutes after each question and types their intended next move into chat. Forces the listening-before-speaking sequence and gives you a written record of what they *wanted* to say — which is often the real diagnostic.
• *Self-scored call review.* The rep scores their own ratio, marks each spot they talked over the buyer, and writes the question they should have asked instead. Self-scoring beats you scoring them by a wide margin; it also takes work off your plate.

The question stack. The structural fix for a rep who runs out of things to ask and defaults to pitching. Three layers: primary, deepening, consequence. "What's your current process for onboarding new reps?" → "Where does that break down first?" → "If that bottleneck went away, what would it do to your ramp time?" Each stack buys two to three minutes of listening, and the rep is never stranded wondering what to say, because they're busy listening for the next layer. Drill it until they can build one without notes on an unfamiliar topic.

The pre-call commitment. Before every call the rep writes one specific listening goal: "three layered questions before I mention pricing," or "buyer speaks 60% of the first ten minutes." Post-call, they self-score and note what triggered any over-talk. This converts an abstract intention into a target and hands you a precise coaching hook: "You committed to three stacks and pitched after one — let's replay that moment."

The 30/60/90 cadence. Days 1–30: make it visible, install the pause only, weekly review, target under 60%. Days 31–60: add stacking and "what else?", introduce a peer call-swap so they hear an excellent listener who isn't you, target under 55% with three-plus layered follow-ups per call. Days 61–90: stress-test on a hostile or price-sensitive buyer and confirm the behavior holds without prompting, target sustained under 50%.

Adjacent effects: what better listening changes downstream

Worth knowing, because it changes who you get budget and buy-in from.

CRM data quality improves before win rates do. A rep running three-layer question stacks collects specifics — named systems, dollar figures, internal politics — and those specifics land in opportunity notes. Within a cycle, deal reviews get sharper because the notes contain the buyer's language instead of the rep's summary of the pitch. If you need an early proof point to justify the coaching time, this one arrives before the pipeline metrics move, and it's the argument that lands with a RevOps team.

Forecast accuracy follows. Over-talking reps close deals on their own theory of the buyer's problem. Listening reps surface the real decision process — the procurement step nobody mentioned, the competing internal project. Slippage drops not because deals move faster but because commit dates stop being fiction.

Onboarding gets a curriculum. Once you've drilled two or three reps, you have recordings of the same person before and after. That before/after pair is the single most effective onboarding asset you can hand a new hire, and it costs nothing to produce because you already made it.

The same drills transfer sideways. Customer success QBRs, renewal conversations, and partner calls all suffer from the identical failure mode — the CSM presents a deck for 40 minutes and learns nothing about renewal risk. The three-second pause and the question stack port over unchanged. Solutions engineers benefit most and are coached least; an SE who asks two consequence-layer questions before opening the demo environment cuts demo length substantially and hits the actual use case.

Where over-talking is legitimate. Don't over-correct. Enablement calls, technical explanations, negotiation-terms walkthroughs, and executive readouts all justify a high ratio. A rep who has *earned* the right to explain by asking well should explain. The failure mode of aggressive ratio coaching is a team of passive order-takers who won't assert a point of view — which shows up as long, drifting cycles and no differentiation. Coach the ratio on discovery, not on everything.

When to stop coaching. If you've diagnosed a will problem, run the honest conversation twice, and the rep still won't own it, more coaching won't fix it — that's a performance-management path, not a coaching one. Coaching resolves skill gaps and most will gaps. It does not resolve someone who has decided not to engage, and pretending otherwise burns your time and the team's patience.

Related questions

What if the rep listens well but still loses deals?

Then listening wasn't the constraint. Check whether they're synthesizing — a rep can collect great information and fail to feed it back as a point of view. Coach the "here's what I heard, here's what I'd do" summary next; it's the skill that sits immediately downstream of listening.

Does talk ratio matter on outbound cold calls?

Less. A cold call's first 30 seconds is legitimately rep-dominated because you're earning the conversation. Measure question count and call duration instead — a cold call that runs past four minutes usually means the rep found something real.

How do I coach this when I can't record calls for legal reasons?

Live-shadow instead. Sit in on two calls, tally speaking turns on paper, and debrief within the hour while it's fresh. You lose replay but keep the diagnostic. Two-party consent regions make this the default in some markets.

Should I share the whole team's talk ratios publicly?

Publish the distribution without names first. Public per-rep leaderboards on a behavioral metric tend to produce gaming — reps go quiet to hit a number rather than to listen. Share individual numbers privately in the 1:1.

How does this change on multi-threaded calls with several buyers?

The metric gets noisier and more useful. Track which buyer talked, not just how much you did. A call where the economic buyer said 40 words is a bad call regardless of your ratio.

FAQ

How do I measure a talk-to-listen ratio without conversation-intelligence tooling?

Score a recording manually: tally 10-second blocks by speaker and divide. A 30-minute call takes about 10 minutes. It's too tedious for weekly use, but doing it once with the rep watching is unusually persuasive — the act of tallying is the intervention. For ongoing tracking, count questions asked per call instead.

What's a good talk-to-listen ratio for a discovery call?

Published conversation-intelligence research puts top performers a little over 40% on discovery — Gong Labs' analysis landed near 43%. Aim to get reps under 50% on discovery specifically. Demos run higher by nature. Set the benchmark by call type; a single universal number makes the metric useless.

The rep nails it in our 1:1 but reverts on live calls. What now?

That's pressure, not knowledge. Adding more explanation makes it worse. Role-play the exact scenario where they revert — the skeptical CFO, the price question at minute six — and increase review frequency to twice weekly for a fortnight. If they can do it in practice, they need repetitions, not information.

How long before this sticks?

Visible ratio movement in one to two weeks with weekly reviews. Layered questioning by week six. Durable performance under quota pressure takes the full 90 days. Habits formed under stress need repeated reinforcement, so don't declare victory after one good call.

Can AI call-coaching tools do this instead of me?

They handle detection well — flagging long monologues, surfacing the ratio, spotting missed questions at scale. They don't handle the will gap, which is the harder half. Use the tooling to find the moments and free your prep time; run the conversation yourself.

Does over-talking ever indicate something other than a coaching need?

Yes. Consistent over-talking across a whole team usually means a broken discovery framework or a comp plan rewarding the wrong call outcome. Check the team distribution before coaching an individual — fixing the system moves everyone at once and costs less manager time.

Sources

• Gong Labs — Talk-to-Listen Ratio Research
• Harvard Business Review — What Great Listeners Actually Do
• MindTools — The GROW Model of Coaching and Mentoring
• RAIN Group — Sales Coaching Tips
• Sandler — Sales Coaching Resources
• Winning by Design — Sales Resources and Frameworks
• Chorus by ZoomInfo — Conversation Intelligence
• Harvard Business Review — The Power of Listening in Helping People Change

Related on PULSE

• [How do you coach a rep to handle 'let me talk to my boss'?](/knowledge/cg0079)
• [How do you coach a rep who freezes on executive calls?](/knowledge/cg0794)
• [How do you coach a rep who gets ghosted after strong first calls?](/knowledge/cg0790)
• [Which single question helps a rep map their talk track directly to the buyer's pain points?](/knowledge/cg0833)
• [How do you coach reps to book more meetings from cold outreach?](/knowledge/cg0027)
• [How do you decide whether to push for a close or give the prospect more time?](/knowledge/cg0938)