How do I coach managers to coach their reps (vs sell deals for them)?

Flip the incentive before you flip the behavior. Managers default to closing because their comp plan still pays them to close. Strip individual quota, pay on team attainment + rep ramp + documented coaching reps, and the "selling vs coaching" debate ends inside one quarter. Until then, every "I'll just hop on this call" is a rational response to a broken plan - and every coaching framework you bolt on top is theater.

The Real Reason Managers Sell Instead of Coach

The Sales Management Association's 2024 Research on Sales Coaching shows only 47% of frontline sales managers spend the recommended 25%+ of their week on coaching, with the median closer to 9% (~3.6 hours/week) [1]. CSO Insights / Korn Ferry's World-Class Sales Practices Study has shown for a decade that reps under managers delivering 3+ hours of coaching per rep per month hit quota at 94%, vs 84% at 2 hours and a sharp drop below 1 hour [2]. The lever is huge. The behavior is rare. Why? Because the manager still carries a number, the org still celebrates the closer, and the coaching reps don't show up on the leaderboard. Behavior follows incentives, not posters.

Cultural piece: in most orgs the highest-paid frontline person is the manager who "saved the deal," not the one who promoted three reps. Until your story changes, the behavior won't.

Step 1: Fix the Comp Plan (or nothing else matters)

• Remove individual close credit. Manager variable comp = 60% team attainment, 20% rep ramp-to-quota time, 10% pipeline coverage, 10% coaching reps logged in Gong/Chorus [3]. Anchor "ramp-to-quota" at a defensible number - SDR/AE benchmark is 5-7 months to first full quota attainment per Bridge Group [7] - so managers cannot game the metric by lowering the bar.
• Cap "manager-named" deals at 15% of team revenue per quarter. Hard cap, finance-enforced. Above 15% triggers a clawback into the rep's commission - reps see the policy and stop deferring deals upward.
• Exclude managers from SPIFFs and President's Club qualification on closed-won. Put them on a separate "Coach of the Year" track tied to rep promotions, rep-to-quota ramp time, and coaching consistency.
• Publish a "Manager Coaching Scorecard" monthly. Names, numbers, calendar mix. Sunlight does most of the enforcement - you should not need to fire anyone over this if the dashboard is honest.
• See [/knowledge/q129](/knowledge/q129) for why great ICs almost never make this transition without a comp redesign, and [/knowledge/q120](/knowledge/q120) for the broader accountability culture that makes the scorecard work.

Step 2: Run the Coaching Audit (monthly, 30 minutes)

Pull last 30 days of Gong, Chorus, or Salesloft Conversations [3] and answer five questions:

1. What % of recorded calls had the manager speaking >40% of the airtime? Target: <10%. Above 25% = they're selling. (Gong's own 2024 conversation benchmark says top reps speak ~46% on discovery and ~65% on demo - managers should be well below those numbers because they are observers.)
2. What % of closed-won deals show the manager as a "key contact" in Salesforce? Target: <15%.
3. How many 1:1s converted into "let me jump on the call" within 24 hours? Target: <2 per rep per quarter.
4. How many documented coaching sessions per rep per week? Target: >=1, 30 min minimum, with a written action item logged in Lattice [4].
5. Coaching-to-deal-review ratio in the manager's calendar? Target: 2:1 coaching to forecast/deal-desk.

If three of five fail, the manager is selling, not coaching. Period. Run this against every manager on the same Tuesday each month - same day, same template, no exceptions. The rhythm is the lever, not the report.

Step 3: The GROW + Sandler Hybrid Coaching Framework

Force every coaching session through a written template (Notion or Lattice [4] both work). The manager fills it before and after the call:

• Goal - what is the rep trying to learn this week? (Not "close the Acme deal." That's a deal review.)
• Reality - what happened on the last 3 calls? Pull the transcript, tag the exact moment, paste the timestamp into the template.
• Options - what are 3 ways the rep could have handled it? Rep generates the options, manager only adds one.
• Will - what is the rep committing to try on the next call? Manager records it. Reviewed next session, with the transcript pulled to verify it was actually attempted.

Layer on the "3 questions before any answer" rule from Sandler [5]: the manager is not allowed to give an answer until they have asked the rep three diagnostic questions. This single rule kills 80% of "let me just tell you what to say" coaching.

For diagnosing whether coaching is even the right intervention vs. training or termination, see [/knowledge/q128](/knowledge/q128). PIPs are a separate beast - see [/knowledge/q123](/knowledge/q123). For a sales kickoff that actually reinforces this framework instead of treating it like a slide, see [/knowledge/q126](/knowledge/q126). For ongoing training cadence at scale, see [/knowledge/q127](/knowledge/q127).

Step 4: Call Review Cadence by Rep Tier

If a manager has more than 3 reps in "bottom quartile" the org has a hiring problem, not a coaching problem - see [/knowledge/q125](/knowledge/q125) for the manager-doesn't-scale signal set.

Step 5: When the Manager Should Step In

Coaching purism breaks expensive deals. The Force Management MEDDICC playbook [6] says the manager should engage as executive sponsor (not closer) when:

• Deal is >3x the rep's average ACV
• A C-level on the buyer side is engaged and the rep has never sold to that level
• Procurement or legal is escalating and the rep is <12 months tenured
• The deal is in the top 5 forecast and slipped a quarter already

Even then, the manager mirrors, doesn't lead. The rep runs the agenda. The manager's role on those calls is one specific thing: validate executive alignment and unblock procurement. They are not there to handle objections the rep should be handling. See [/knowledge/q121](/knowledge/q121) for the full step-in decision tree.

Bear Case: Why This Plan Fails

Plan-honest, not plan-evangelist. This approach breaks in four predictable ways and you should plan for them before you launch:

1. The comp redesign nukes Q1. Strip individual quota and your best player-coach manager will mentally check out for a quarter while they figure out the new game. Bridge Group's 2024 Sales Compensation report flags 18-22% short-term attainment dip in the first cycle after manager comp redesigns [7]. Plan for it. Don't reverse the policy in week 6 - that's the most common failure mode and the one that destroys credibility for any future redesign. Pre-fund a one-quarter "transition guarantee" at 80% of OTE so managers don't panic and revert to closing.
2. Coaching theater. Managers learn to log "coaching sessions" in Gong that are really just deal reviews with the words "what do you think?" sprinkled in. Audit the content, not the count. Pull 5 random transcripts per manager per quarter and have a peer manager grade them against the GROW template - peer grading kills theater faster than top-down audits.
3. The bottleneck moves to the manager. Strong coaching cultures still cap out around 7-9 reps per manager because real coaching takes 3-4 hours/rep/month [2]. If you scale to 12 reps under one manager you re-create the "I don't have time, just close it" problem. The fix isn't more coaching tech, it's more managers - and that requires a player-coach to non-selling-manager promotion path most orgs don't have.
4. You promoted the wrong person. No coaching framework rescues a manager who fundamentally enjoys closing more than developing people. About 30% of newly promoted frontline managers demote themselves back to IC within 18 months per CSO Insights [2], and that's a feature, not a bug. See [/knowledge/q129](/knowledge/q129).
5. The data layer is missing. All five audit questions assume Gong/Chorus + clean Salesforce contact roles + Lattice 1:1 logging. If two of those three are not in place, the audit can't run and you're back to anecdote. Don't launch the program until the telemetry is real - 30 days of "we'll fix it later" becomes 18 months.

The 90-Day Implementation

• Days 1-30: Audit current state (manager calendar mix, Gong airtime %, deal records, 1:1 conversion-to-deal-jump rate). Publish baseline scorecard. Walk every manager through their numbers 1:1 - no surprises in the all-hands. The conversation is "here's what changes, here's why, here's the comp protection during transition."
• Days 31-60: Roll out new comp plan with finance. Train managers on GROW + Sandler 3-question rule with a 4-hour workshop and live role-play, not a deck. Stand up call-review cadence with a real cadence calendar (Tuesday transcripts, Thursday role-plays). Pair every manager with a peer-coach for the first 60 days.
• Days 61-90: First coaching audit under the new comp plan. Adjust. Promote two manager success stories internally - storytelling is the cultural lever and it's free.

Bridge Group benchmark: orgs with strong coaching cultures see 91% rep quota achievement vs 73% for orgs where managers still chase personal numbers [7]. Eighteen percentage points is the difference between hitting plan and missing it - and the entire delta is comp design plus a 30-minute weekly habit.

TL;DR

Coaching is a comp-plan problem before it's a skill problem. Pay managers on team attainment and rep development, cap their personal deal involvement, audit the content of coaching sessions monthly, and design the program for the realistic 7-9 reps per manager span of control. Anything else is a poster.

Sources: [1] Sales Management Association 2024 Research on Sales Coaching, [2] CSO Insights / Korn Ferry World-Class Sales Practices Study, [3] Gong / Chorus / Salesloft conversation intelligence platforms, [4] Lattice 1:1 + Performance, [5] Sandler Selling System ("3 questions before any answer"), [6] Force Management MEDDICC playbook, [7] Bridge Group SaaS AE Metrics & Compensation Reports.

TAGS: manager-coaching, team-development, sales-leadership, quota-achievement, skill-transfer, comp-design, gong, sandler, meddicc, frontline-management, grow-model, span-of-control