How do you decide which reps to coach first?

The outcome you should expect

When you get the prioritization right, the change shows up in a specific, measurable way within one quarter — and it does not look like "everyone got a little better." It looks like two or three named reps moving one stage-conversion metric each, while the rest of the team holds steady. That asymmetry is the signal that you spent your hours where they compounded.

Concretely, a manager with seven reps who picks three core performers and coaches one skill each should expect: a visible behavior change in call recordings by week 3–4, a movement in the targeted stage-conversion rate by week 8–10, and a bookings impact somewhere in the following quarter. The lag is real and it is why managers abandon coaching — they check the lagging number at week 4, see nothing, and conclude coaching does not work. It works; you measured the wrong thing at the wrong time.

The second outcome is subtler and arguably more valuable: you stop spending your week on emotional triage. Most sales managers can name the rep who consumes 40% of their attention, and it is almost never the rep who returns 40% of the upside. A deliberate priority list ends that. You are no longer reacting to whoever escalated loudest on Monday; you have two to four names, a skill attached to each, and a cadence. Everyone else gets deal help, enablement, and a monthly check — which is honest and sufficient.

The third outcome is that your bad news gets faster. When you formally exclude a rep from the coaching priority list because the gap is will, fit, or territory rather than skill, you have implicitly made a decision. That decision either becomes a performance conversation, a territory fix, or a comp fix — but it stops being deferred under the comfortable label of "I'm coaching them." A lot of underperformance survives for two extra quarters because a manager mislabeled a structural problem as a coaching project.

What you should *not* expect: a lift in the bottom decile. Reps in the bottom tier of a well-designed comp plan and a fair territory usually have a will or fit problem, and skill coaching does not move those. If your bottom reps improve after you deprioritize them, the honest read is that the improvement came from the accountability conversation you finally had — not from coaching you did not do.

What drives that outcome

Everything hinges on one upstream diagnosis: is this a skill gap at all? Four causes produce nearly identical symptoms on a dashboard, and only one of them responds to coaching.

Skill — the rep does not yet know *how*. Weak discovery, cannot multithread past one champion, fumbles the pricing conversation, never asks for a next step. This is the only cause that drills, call review, and role-play actually fix. Test for it by asking them to demonstrate the behavior in a safe setting. If they genuinely cannot do it when watched, it is skill.

Will — they know how but do not. Low outbound activity, avoids the hard second call, coasting on inbound. Coaching here becomes accountability and motivation work, which is a different conversation with a different escalation path. The tell: they *can* do it in role-play, cleanly, and still do not do it live.

Knowledge — missing product, market, competitive, or buyer context. Fixed far faster by enablement, competitive battlecards, and shadowing two strong reps for a week than by 1:1 coaching. Do not burn coaching hours on something a two-hour enablement session solves for the whole team.

System — bad territory, broken comp, no lead flow, a genuine product gap. No amount of coaching fixes a structural problem. The diagnostic: if two or three reps are failing in the same identical way, it is almost certainly not three simultaneous skill gaps. It is the system. This is where RevOps earns its keep — pipeline coverage by territory, lead-to-opportunity rates by segment, and comp attainment distribution will tell you in twenty minutes what six coaching sessions would not.

Beneath the diagnosis sit three ranking inputs that decide the order once you have filtered to genuine skill gaps:

• Coachability — do they act on feedback? The cheapest test: give a small, specific assignment ("ask for a next step on every call this week"), then check three calls. A rep who executes a trivial ask executes a hard one. A rep who does not will not, regardless of talent.
• Revenue upside — how much number sits behind the gap? A rep with $600K of open pipeline stalling at evaluation is a bigger prize than a rep with $150K stalling at the same stage, all else equal.
• Deal exposure — is there a live deal to practice on this month? Coaching without a real deal to apply it to decays fast. Prefer the rep who can use the skill on Thursday.

Two tiebreakers are worth naming because they overrule raw upside. First, ramp stage: a rep in month four of a six-month ramp beats a tenured flat rep at similar attainment, because early habits compound and bad ones take months to unwind. Delay coaching a ramper by three weeks and you have baked in a discovery pattern you will spend a quarter undoing. Second, shared stall pattern: if three reps all lose deals at the same evaluation handoff, that is one group session or one process fix serving three reps — strictly better ROI than a 1:1 with a rep whose stall pattern is unique to them.

Benchmarks and realistic ranges

Numbers here vary by motion, but the ranges below are the ones worth planning against, and each one is a lever you can actually set.

How many reps you can coach deeply at once: two to four. For a frontline manager with six to eight reps and normal forecast, hiring, and escalation duties, deep weekly coaching caps out around three. Past four, the sessions degrade into status updates — you stop preparing, stop watching calls beforehand, and start asking "how's the pipeline?" That is not coaching. If you carry ten or more reps, the honest number is two, plus peer coaching to cover the rest.

Time budget: five to eight hours a week. Budget roughly 30 minutes of live 1:1 plus 20–30 minutes of prep (watching one call, pulling the relevant clips) per priority rep, per week. Three priority reps is therefore about three hours of contact and ninety minutes of prep, plus a biweekly 45-minute deal review each. Managers who claim to coach eight reps weekly are almost always doing zero prep, which is where most of the value lives.

Cycle length per rep: six to twelve weeks. One skill, one rep, one cycle. Under six weeks you rarely see the leading indicator move; past twelve you are in hope-based coaching. Build the priority list quarterly and expect one to two reps to graduate off it mid-quarter and be replaced.

Time-to-signal by stage of the funnel of evidence:

• Week 1–2: the rep can describe the gap in their own words and name the specific moment it happens.
• Week 3–4: the behavior appears in role-play reliably and starts appearing in live recordings.
• Week 6–10: the targeted stage-conversion metric moves.
• Next quarter: bookings move.

If you are still at "describes the gap" in week five, the diagnosis was wrong or the rep is not coachable. Re-rank rather than push harder.

Consecutive-session cap: two. If you have coached the same rep two sessions in a row on the same skill with no measurable behavior change — no shift in discovery questions per call, no change in contacts per deal, no next-step ask — stop and rotate. Continuing is repeating advice and expecting a different result. Rotate to the next rep on the list; the space often does more than the third session would have.

Call-review volume: one call per rep per week, watched properly. One call reviewed with real preparation beats five skimmed. Skim five and you will coach generic feedback ("more discovery"), which changes nothing. Watch one closely and you get a timestamped moment — "at 14:20 the buyer said budget was tight and you moved to features" — which is coachable.

Stall concentration threshold: 30%. When you pull deal velocity by stage, flag any rep losing 30% or more of their deals at a single stage. That is a stage-specific, highly coachable gap. A rep whose losses are spread evenly across stages usually has a qualification or territory problem instead, and belongs in a different conversation.

Coverage floor before you coach at all: check pipeline coverage first. If a rep is carrying materially less pipeline coverage than the rest of the team on a comparable patch, coaching their closing skill is malpractice — they do not have enough at-bats for skill to be the binding constraint. Fix the top of the funnel or the patch first, then coach.

Risks, edge cases, and failure modes

The bottom-decile trap. The single most common failure is defaulting to the struggling rep because the pain is loudest. It feels responsible and is usually financially backwards. Before you commit weekly hours to a bottom performer, answer honestly: is this skill, or is this fit, will, or territory? If it is not skill, coaching is a way of postponing a decision you already know you need to make, and it costs the middle of your team the attention that would have paid.

The favorite trap. The inverse: over-coaching the top rep because the conversations are enjoyable and they engage well. Marginal return is near zero — they are already converting. Their highest-leverage role is teaching. Have your best discovery rep run a 30-minute teardown for the team once a month; that scales your hours in a way another 1:1 with them never will.

Coaching the deal instead of the skill. Jumping into a live opportunity to save this quarter's number feels productive and teaches nothing transferable. It is also addictive, because it works — the deal closes, and you both feel good. Deal help is legitimate, but log it as forecast work, not coaching, and keep it separate from the priority list so you do not fool yourself about how much coaching you actually did.

Framing that reads as remediation. The capable middle disengages the second coaching smells like a performance improvement process. If you open with "we need to talk about your numbers," you have lost. Open with upside: *"I want to spend real coaching time with you this quarter because I think you're close to a level-up, not because something's wrong."* Then let them name the gap.

Stacking fixes. Handing a rep five improvements at once guarantees zero. One skill per cycle. If you have identified four gaps, pick the one closest to revenue and hold the rest; write them down so you are not tempted to mention them.

No follow-through. Assigning a skill in week one and never reviewing a call in week three trains the whole team that coaching is theater. Worse, it teaches your priority reps that your assignments are optional — which contaminates the coachability signal you rely on to rank.

Rescuing the rep in role-play. Taking over to "show them how" robs the rep of the practice. Let them struggle in the safe setting; that is the entire point of the safe setting.

Misreading a system failure as three skill gaps. If three reps fail identically, fix the system. Pull the RevOps view — conversion by segment, lead quality by source, comp attainment distribution — before you schedule a single session. Coaching a structural problem burns goodwill and delays the actual fix by a quarter.

Edge case — the high performer with one glaring hole. A rep at 110% who cannot multithread is a genuine exception worth a slot, because the hole caps their ceiling and the fix is narrow. Give them a short, sharp cycle: one skill, six weeks, then out.

Edge case — the rep who improves in calls but not in numbers. Behavior changed, results did not. Usually territory, lead quality, or pricing. Confirm the behavior change is real in recordings, then stop coaching that rep and escalate the structural issue — you have already done your part.

Edge case — the coachable rep with no live pipeline. Nothing to practice on. Pair them with enablement and prospecting work, get them at-bats, and slot them into the coaching list next cycle when there is something to apply the skill to.

A practical rollout plan

Week zero is a ranking exercise, not a coaching session. Pull three things: attainment and pipeline coverage per rep, stage-conversion by rep for the last two quarters, and your own notes on who acts on feedback. Score each rep on coachability (high/medium/low), revenue upside behind the gap, and whether a live deal exists to practice on this month. That produces a ranked list in under an hour.

Then run the diagnosis on the top six. For each: can they demonstrate the skill in role-play? Is the same failure showing up on other reps' deals? Is the territory and lead flow comparable to peers? Only the reps who come out as genuine, isolated skill gaps make the list. Cap it at three.

Week one, open each priority rep with the upside framing and run a GROW conversation — Goal, Reality, Options, Will. Ask what "measurably better" would look like to them by quarter end. Pull an actual recorded call and walk the exact moment the deal slipped. Ask them for two alternative moves rather than supplying one. Close by asking how they want to be held to it. You leave with one skill, one metric, and a check-in date.

Weeks two through ten, hold the rhythm: a 30-minute weekly 1:1 built around one reviewed call, a biweekly 45-minute live deal review against a qualification checklist, and one drill per week. Useful drills — mute a recorded call and have the rep narrate what they would do differently before hearing themselves; run a six-minute discovery role-play where you play a skeptical economic buyer and score against a written discovery scorecard; rapid-fire your five deal-killing objections with a ten-second response window; have the rep present one win and one loss to peers. Monthly, re-rank the list, graduate whoever moved, and rotate the next rep in.

Measure the leading indicator you coached, never the lagging one. If you coached discovery, watch discovery-to-demo conversion and discovery questions per call — not bookings. If you coached multithreading, watch contacts per deal. If activity rose and conversion stayed flat, your diagnosis was wrong: it was will or system, and you should re-route rather than run another cycle.

Two governance habits keep the system honest. Write the priority list down where your own manager can see it, with the skill and the metric next to each name — public commitment is what stops the list quietly reverting to whoever escalated on Monday. And keep a standing item in your RevOps review for reps you *excluded* because the cause was structural; that list is a queue of territory, comp, and lead-flow fixes, and it should shrink.

Related questions

Should I coach the rep with the biggest gap or the biggest upside?

Upside, filtered by coachability. The biggest gap is often the least fixable — it usually signals fit or will rather than skill. A moderate gap in front of large pipeline, held by someone who acts on feedback, returns far more per hour.

How do I rank reps if everyone is roughly at quota?

Rank by ceiling instead of floor. Look for the single skill capping each rep's largest deals — multithreading, executive access, pricing — and pick whoever has the most pipeline blocked behind a narrow, teachable hole.

Does the priority list change every quarter?

Yes, and it should. Reps graduate when the leading indicator moves, typically in six to twelve weeks. Re-rank monthly, replace graduates, and re-run the diagnosis whenever territory, comp, or lead flow changes materially.

What if my manager tells me to coach the bottom performer?

Do the diagnosis and bring evidence. If it is will, fit, or territory, say so plainly and propose the right intervention. If it is genuinely skill, coach them — but negotiate which core rep drops off your list to make room.

How does RevOps help decide who to coach?

RevOps supplies the data that separates skill from system: stage conversion by rep, pipeline coverage by territory, lead quality by source, and comp attainment distribution. Without it you are ranking on impressions.

FAQ

Should I ever coach my top reps?

Yes, on a lighter "sharpen, don't fix" cadence — a monthly call review and one stretch skill. Pulling your best closer into weekly remedial-style sessions insults them and yields almost no marginal revenue. Their highest-leverage contribution is usually teaching the middle: a monthly teardown of one of their own deals scales your hours further than another 1:1 ever will.

What if an underperformer is genuinely coachable but low-priority?

Queue them behind the core and give them enablement, self-serve resources, and a clear monthly check-in so they are not abandoned. Coachability without near-term revenue upside earns a rotation slot, just not the front of the line this quarter. Tell them that directly — reps handle honest prioritization far better than vague neglect.

How do I tell a will problem from a skill problem?

Ask them to do the thing in role-play. If they can do it cleanly when watched but do not do it in the wild, it is will or system. If they genuinely cannot demonstrate it, it is skill. Only the skill case responds to drills and call review; the will case needs an accountability conversation with a defined escalation path.

How many reps can I actually coach well at once?

Two to four for deep weekly coaching, depending on team size and your other duties. Past four, prep collapses and sessions become status updates. If you carry ten or more reps, plan on two, and use peer coaching and group sessions for shared gaps across the rest of the team.

Does AI call-coaching software replace this decision?

No. Conversation-intelligence tools surface what to coach far faster than manual call review, and they make the diagnosis step much cheaper. They do not decide who deserves your scarce hours, and they do not hold the accountability conversation. They sharpen your prioritization rather than replacing it.

When is coaching the wrong tool entirely?

When the gap is fit, comp, territory, lead flow, or a product hole. Those are management decisions. Coaching a structural failure burns goodwill, delays the real fix by a quarter, and teaches the rep that their environment is their fault.

Sources

• Harvard Business Review — The Most Productive Ways to Coach Your Sales Team
• Gong Labs — Sales Coaching Research
• RAIN Group — Sales Coaching Best Practices
• MindTools — The GROW Model of Coaching and Mentoring
• MEDDIC Academy — What Is MEDDIC
• Winning by Design — Resources
• SHRM — Performance Improvement Plans
• McKinsey — Sales Growth and Performance Insights

Related on PULSE

• [How do you decide whether to push for a close or give the prospect more time?](/knowledge/cg0938)
• [How do you coach a brand-new sales manager on their first hire?](/knowledge/cg0802)
• [How do you coach a rep who gets ghosted after strong first calls?](/knowledge/cg0790)
• [Which sales metrics should you coach to first?](/knowledge/cg0162)
• [How do you coach a new rep through their first lost deal?](/knowledge/cg0020)
• [How do you coach a career-changer into their first sales role?](/knowledge/cg0018)