Half of managers say they give their people meaningful feedback every week. Only 20% of employees say they get it. That 30-point gap, from Gallup’s 2024 study of 2,729 managers and 12,710 individual contributors, was the largest perception gap in the whole survey.

Part of the explanation is simple. A lot of what managers count as coaching is really answering. Someone brings a problem, you solve it in ninety seconds, they thank you, and you log it mentally as a development conversation. They log it as getting told what to do.

The fix is not a coaching certification. It is a handful of coaching questions for managers, asked before you give an answer, and the discipline to stay quiet long enough to hear the reply.

The question that does most of the work

Before you answer any question a direct report brings you, ask one back:

“What would you recommend?”

Four words. They move the thinking to the person closest to the work, and they tell you something you cannot learn any other way: how that person reasons when you are not in the room.

Three things usually happen.

• They already have a workable answer. Most people who bring you a problem have a solution in mind and want permission, not direction. Your question gives them the permission without you taking the decision away from them.
• You find the real gap. If the recommendation is off, you now know whether the problem is missing information, poor judgment about risk, or not knowing who else to involve. Each of those needs a different response, and handing over your answer fixes none of them.
• They own what happens next. A plan someone proposed and you approved gets executed differently from a plan you dictated. People defend their own ideas.

Michael Bungay Stanier built an entire book around this instinct. The Coaching Habit, now in a tenth-anniversary edition, gives managers seven questions, and the one he leans on hardest is the shortest: “And what else?” The first answer someone gives is rarely the full picture. Asking once more, without adding your own view, regularly surfaces the thing they were actually worried about.

Why asking beats telling (and why more coaching is not the answer)

Gallup calls ongoing, high-quality manager conversations “the coaching habit” and describes it as among the best predictors of employee engagement it has ever studied. In an earlier analysis covering 1.2 million employees, Gallup found that only 26% of employees strongly agree the feedback they receive helps them do better work. Most feedback, in other words, is not landing.

The tempting conclusion is to coach more. The research says be careful.

Gartner surveyed 7,000 employees and sorted managers into four styles: teachers, cheerleaders, “always on” managers who coach and give feedback constantly, and connectors who coach in their own area of strength and then link people to others who know more. As HR Dive reported from Gartner’s own presentation of the findings, the always-on style actually degraded employee performance. Connectors improved it by 26%.

That finding matters for how you use coaching questions. The goal is not to turn every hallway exchange into a Socratic seminar. A person who gets interrogated every time they ask for a password reset will stop asking you anything. The goal is to ask on the decisions that build judgment, and to point people somewhere useful when you are not the right source.

Herminia Ibarra and Anne Scoular make a similar point in their 2019 Harvard Business Review piece, “The Leader as Coach”: managers are increasingly expected to help people find answers because no manager has all of them anymore. Coaching questions are a tool for that, not a personality you put on every morning.

When the answer is “I don’t know”

Sometimes you ask “What would you recommend?” and get a shrug. That is information too. A few follow-ups, depending on what the shrug means:

They are afraid of being wrong. Lower the stakes. “If you had to guess, which way would you lean?” A guess is easier to offer than a recommendation, and it is usually the same answer.

They are new and genuinely lack context. Narrow the field. “I see two options here, A or B. Which fits better, and why?” You are still making them reason; you have just shrunk the search space. Widen it as their judgment develops.

They have not thought about it yet. Send them back with a deadline. “Think about it and bring me your recommendation at three.” This one feels slow the first time. By the third time, people show up with the recommendation already in hand.

They think it is not their call. Then the problem is not coaching; it is unclear ownership. If people do not know which decisions belong to them, no question will make them comfortable answering. Settle who owns which decisions first.

And when the recommendation is wrong, resist the urge to correct it outright. Ask what would happen if a specific risk showed up: “What do we do if the vendor misses the date?” If they cannot answer, they will usually revise the plan themselves. If the downside is small and recoverable, consider letting them run with it. Some lessons only stick when they are earned, as long as the mistake is reversible.

Some problems you should just answer

Coaching questions are for the daily decisions where developing someone matters more than saving three minutes. They are the wrong tool in a few situations, and pretending otherwise makes you look evasive.

• A live incident. When a service is down and customers are calling, give direction. Debrief the reasoning afterward.
• A question of fact. “Which cost center does this go to?” deserves an answer, not a question back.
• Someone who is overwhelmed. A person drowning in a bad week does not need a growth moment. Take something off their plate, then coach next week.
• A decision that is genuinely yours. If you are going to overrule whatever they say, asking for their recommendation is theatre. Tell them it is your call and explain it.

What changed on my own team

I learned this the expensive way. Running operations teams at a large telecom, I had a stretch where my calendar filled up with people bringing me decisions they were perfectly capable of making: which change window to use, whether a vendor ticket justified escalation, how to word a customer outage notice. I answered all of them quickly and felt useful doing it.

The cost showed up when I took a week off. Decisions stalled until I got back, not because people lacked the skill but because I had taught them that the fastest route to a decision ran through my desk.

So I started asking “What would you recommend?” before answering anything that was not urgent. The first couple of weeks were awkward. People were annoyed, a few thought I was testing them, and some conversations took longer than just answering would have. Tone fixed most of that. “You’re closer to this than I am, what would you do?” lands very differently from the same question asked with folded arms.

Within about a month, the shape of the conversations changed. People started opening with “Here’s the problem, and here’s what I think we should do.” Within a quarter, a good share of the routine questions stopped coming to me at all, because people already knew what I would ask and had answered it for themselves.

That is the actual payoff. Not that your team needs you less in some abstract sense, but that your one-on-ones shift from approving small decisions to talking about the hard ones, and the 20% of employees who say they get weekly feedback starts to look a lot more like the 50% of managers who think they are giving it.

Start with the next person who brings you a problem today. Ask what they would recommend. Then wait, even when the silence runs past five seconds. They will fill it.