How do you coach a sales leader to coach their own managers in 2027?

Two paths: teach the frame, or observe the reps

There are really only two credible ways to make a sales leader better at coaching their own managers, and most organizations blur them into a mush that accomplishes neither. The first path is frame-first: you teach the leader an explicit coaching model — GROW, or a question hierarchy, or a simple diagnostic ladder — in a compressed workshop or a series of role-plays, then hand them a template and let them run. The second path is observation-first: you skip the theory almost entirely, sit in on the leader's actual 1:1s with their managers, and debrief what you saw. No framework taught up front; the framework emerges from repeated feedback on real behavior.

Frame-first is cheap and scales across a wide leadership bench. You can put twelve directors through the same two-day intensive, hand them a laminated card, and walk away with everyone speaking the same vocabulary. Its failure mode is well-known to anyone who has run enablement: vocabulary adoption without behavior change. The leader learns to say "what options have you considered" and then, four questions later, tells the manager exactly what to do anyway. They can describe GROW on a whiteboard and still run every 1:1 as a status interrogation. Frame-first teaches the *language* of coaching, and language is easy to fake — to yourself most of all.

Observation-first is expensive and does not scale past what your own calendar can hold. Sitting in on one 1:1 costs you 45 minutes, and the debrief costs another 30. Do that with five leaders monthly and you have burned roughly six hours plus prep. But it is the only path that produces reliable behavior change, because it operates on what the leader actually *did*, not what they believe they do. Every leader believes they ask more questions than they do. Watch one session, count the question-to-statement ratio, and hand them the number — that single data point does more work than a day of theory.

The trap in the comparison is treating them as mutually exclusive. The right answer for most revenue organizations is frame-first as the floor, observation-first as the escalation. Give everyone the shared vocabulary so debriefs have something to hang on, then spend your scarce observation hours on the leaders where the gap is real — the newly promoted director, the leader whose managers keep churning, the one whose team hits number but produces nobody promotable. A third variant deserves mention because it is increasingly common in 2027: instrumented coaching, where conversation-intelligence tooling pulls talk-ratio and question-count off recorded 1:1s and surfaces them automatically. That is not a replacement for either path. It is a cheap sensor that tells you where to point your observation hours. The data flags that a leader talked 78% of a session; only a human can tell you whether that was a bad habit or a necessary termination conversation.

One more distinction that matters more than people admit: coaching a leader to coach *managers* is a different skill from coaching a leader to coach *reps*, and the second does not transfer to the first. Coaching a rep is largely coaching for execution — the deal is real, the clock is running, the coach has more pattern recognition than the coachee, and a directive nudge is often correct. Coaching a manager is coaching for development — the timeline is quarters not days, the manager owns a team you cannot see into, and a directive nudge actively destroys the manager's authority with their own reps. The leaders who struggle hardest are usually the ones who were *excellent* rep coaches. Their instincts are calibrated for a game they are no longer playing.

How to decide which path a given leader needs

The decision is a diagnosis, not a preference. Before you spend a single observation hour, work out which of three gaps you are actually looking at, because the intervention for each is completely different and applying the wrong one makes things worse.

Skill gap. The leader wants to coach their managers, has the time, and does it badly. Diagnostic move: ask them to walk you through their most recent coaching session with a manager. Not "how is Priya doing" — that gets you a performance summary. Ask: *what were you trying to achieve in that conversation, and how would you know if it worked?* A leader with a skill gap cannot answer the second half. They will describe what was discussed, not what they were trying to shift. That is the tell. Skill gaps respond well to role-play — you play the manager, they coach you, then you break character and debrief. Twenty minutes of that beats two hours of framework.

Will gap. The leader knows how to coach and does not. Sessions get cancelled for pipeline emergencies. Feedback gets deflected. The diagnostic is behavioral: look at the last eight weeks of their calendar and count how many coaching blocks survived contact with the quarter. If more than a third got bumped, you have a will problem masquerading as a time problem. Underneath, will gaps are usually one of three things — the leader doesn't believe manager development is their job, the leader is being measured only on number so coaching feels like unpaid work, or the leader is quietly threatened by a manager who is visibly better than they were at that stage. That third one is real and almost never said out loud. It surfaces as nitpicking a strong manager's work, or as a leader who keeps the best accounts inside their own span of control. Address it with career pathing and public credit for developing talent, not with more coaching training.

System gap. The leader has the skill and the will and the organization makes it impossible. Twelve direct reports instead of six. Comp tied entirely to bookings with nothing for manager retention or promotion. A forecast cadence that eats nine hours a week. This is the gap most VPs misdiagnose as will, and it is the one you must fix before coaching anything, because coaching a person inside a broken system just teaches them that your feedback is noise.

mermaid flowchart TD A[Weeks 0-2: Audit span, comp, meeting load] --> B{System supports coaching?} B -- No --> C[Fix span / comp / cadence] C --> A B -- Yes --> D[Weeks 2-4: 90-min shared frame session] D --> E[Weeks 4-8: Protect monthly dev blocks] E --> F{Blocks survived 8 weeks?} F -- No --> G[Surface it: what is pulling on the time?] G --> E F -- Yes --> H[Observe one real 1:1 per leader] H --> I[Debrief: notice, best question, redo] I --> J[Give ONE behavior to change] J --> K[Log it and reopen next month] K --> H </parameter>

Two adjacent workflows deserve a place in the sequence because they either reinforce it or quietly undo it. The first is peer circles: leaders meeting monthly without you to compare coaching problems. This works because it removes the evaluative charge — a leader will admit to a peer that they have no idea how to handle a manager who won't hold their team accountable, and will not admit that to the person who writes their review. The second is the forecast meeting, which is where coaching cultures go to die. If your weekly forecast call is a public interrogation, every leader in the room learns the actual house style regardless of what the workshop taught. Coaching behavior is downstream of how you run the room.

Where does RevOps sit in all this? Further forward than most teams use them. RevOps owns the instrumentation that makes coaching-the-coach measurable rather than anecdotal: manager tenure and promotion-rate reporting, coaching-session completion pulled from calendar data, and the per-team leading indicators — pipeline hygiene, stage-conversion consistency, forecast accuracy by manager — that tell you which manager is actually improving. Without that, a VP's read on their leaders' coaching is pure impression, and impressions correlate with likeability far more than with capability. Ask RevOps for one view: manager-level trend lines on a small set of leading indicators, refreshed monthly, that you and the leader read together. It converts "I think Dana's team is tightening up" into something you can coach against.

What this looks like in adjacent situations

The core pattern generalizes, and seeing it generalize helps a leader internalize it faster than another round of role-play.

A newly merged team. After an acquisition or a reorg, a leader inherits managers they did not hire, running a motion they did not design. The coaching-the-coach cadence is the same, but the first two months should be pure observation with almost no feedback. The leader needs to learn what "normal" looks like on that team before they can tell dysfunction from unfamiliar. Coaching too early after a merger reliably produces the leader importing their old playbook onto a team where it does not fit.

A player-coach leader. Common in companies under a few hundred people: the director carries a small book alongside their managers. The trap is that carrying quota is urgent and coaching is not, so coaching loses every week. The only durable fix is structural — cap the book, or accept a reduced coaching cadence explicitly rather than pretending. What you must not do is expect full coaching output from someone with a full quota.

Partner and channel leaders. A leader who manages partner managers faces the same three-layer problem with an extra layer of indirection: the partner manager coaches a partner rep who does not work for you at all. The coaching moves are identical, but the leverage is lower and the feedback loop is slower, so weight the cadence toward the partner manager's *process* — enablement delivered, joint pipeline reviewed — rather than toward outcomes you cannot attribute cleanly.

Customer success and post-sales. The same model transplants directly to a CS director coaching CS managers, with one adjustment: the leading indicators are retention and expansion rather than pipeline, and those move on much longer cycles. Because the lagging metric is slow, the leading behavioral indicators — session frequency, question ratio, manager retention — carry proportionally more weight in evaluating whether coaching is working.

A leader who is better than you at something. Eventually you will coach a leader whose technical command of the sales motion exceeds yours. The instinct is to avoid coaching them at all, or to compensate by finding something to critique. Do neither. Your value in that relationship is not superior domain knowledge — it is the outside observation. You can see that they interrupted their manager four times in twenty minutes; they cannot. That observation is legitimate regardless of who knows the product better, and saying so plainly usually defuses the awkwardness in one conversation.

Related questions

How is coaching a manager different from coaching a rep?

Rep coaching is execution coaching: short cycle, live deals, directive input often correct. Manager coaching is development coaching: multi-quarter horizon, and directive input undermines the manager's authority with their own team. Excellent rep coaches frequently struggle here because their instincts are calibrated wrong.

Should I ever skip a level and coach a manager directly?

Almost never for coaching. Skip-levels for listening and visibility are healthy and should be scheduled openly. But redirecting a manager's work around their leader teaches everyone that real authority sits a level up, and that damage outlives whatever problem you solved.

How long before coaching-the-coach shows in revenue?

Two quarters minimum for behavioral change to reach team results, often longer. Inside that window, track leading indicators: coaching session completion, question-to-statement ratio, manager retention and promotion rate, and per-team pipeline hygiene trends.

What if the leader simply doesn't want the role?

Some strong individual leaders genuinely don't want to develop coaches, and that is legitimate information rather than a failure. Name it directly and explore whether an individual-contributor-heavy or specialist path fits better than pushing them through a role they will resent.

Can AI tools replace observation?

No. Conversation intelligence surfaces cheap signals — talk ratio, question count, wait time — across far more sessions than you can attend. Treat it as a sensor telling you where to spend human observation hours, never as a verdict on whether a session was good.

FAQ

How do I know if a sales leader is ready to coach their own managers?

Look for two things: self-awareness and tolerance for slow results. A leader who asks for feedback on their own coaching, and who can describe a moment they handled badly, is ready. A leader who explains every manager's shortfall as the manager's fault is not — and the deficit there is not skill, it is the belief that developing people is part of the job. Start with that conversation, not with a framework.

What if the leader resists because they think coaching managers isn't their job?

Make the leverage arithmetic explicit. A leader personally coaching reps might touch eight people; a leader who develops five managers who each coach eight reps touches forty, indefinitely, without their calendar. If the resistance persists after that conversation, it is usually a role-fit question rather than a motivation question, and it is kinder to both parties to name it early.

How often should I observe a leader coaching their manager?

Monthly for the first two quarters, then quarterly once the behavior holds without you. Frame it as observation, not inspection, and be consistent about it — sporadic observation reads as surveillance triggered by suspicion, while a fixed monthly rhythm reads as normal. Tell the manager beforehand why you are in the room, too. A silent extra body changes the session.

A leader's manager is clearly underperforming — do I step in?

Not directly. Coach the leader on handling it: what have they already tried, what did they observe, is it skill, will, or system, and what is their next move. Offer support that runs through them rather than around them. The exception is genuine misconduct or a compliance issue, which is not a coaching situation at all and should go through the appropriate channel immediately.

Which metrics actually tell me the coaching is working?

Session completion rate against the protected cadence, question-to-statement ratio trending toward parity, manager tenure and internal promotion rate, and per-manager leading indicators like pipeline hygiene and forecast accuracy. Have RevOps own the reporting so it does not depend on self-assessment. Revenue is the lagging confirmation, not the signal you steer by.

What's the most common failure mode in the first quarter?

Coaching blocks getting eaten by forecast urgency, followed by everyone quietly agreeing not to mention it. The counter is to notice out loud the first time it happens and ask what pulled on the time. That single observation, made once and without accusation, does more to establish that the cadence is real than any announcement.

Sources

• https://hbr.org/2011/07/the-dirty-secret-of-effective-sales-coaching
• https://hbr.org/2019/11/the-leader-as-coach
• https://www.gartner.com/en/sales/insights/sales-management
• https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights
• https://www.gallup.com/workplace/236570/employee-engagement-drives-growth.aspx
• https://sloanreview.mit.edu/topic/leadership/
• https://www.ccl.org/articles/leading-effectively-articles/
• https://www.shrm.org/topics-tools/topics/organizational-employee-development
• https://corporatefinanceinstitute.com/resources/management/span-of-control/

Related on PULSE

• [How can I use open-ended questions to help a sales rep identify their own weaknesses in a deal review?](/knowledge/cg0809)
• [How do you coach reps to manage their own energy and time?](/knowledge/cg0149)
• [How do you coach a rep using their own call recordings in a 1:1?](/knowledge/cg0117)
• [How do you coach reps to set their own development goals?](/knowledge/cg0115)
• [How do you coach a rep to inspect their own pipeline?](/knowledge/cg0105)
• [How do you ask a rep to re-articulate a prospect's pain point in their own words to test comprehension?](/knowledge/cg0872)