This article was originally published on davidohnstad.info. I cross-post here to reach the Dev.to community.
The Myth That Great Managers Must Be Great Coaches
A product manager at a mid-sized SaaS company pulled David Ohnstad aside after a leadership workshop and said, "I just spent three hours asking coaching questions to help my engineer figure out a solution he already knew. He finally said, 'Can you just tell me what you want me to build?' I felt like I failed." According to SHRM's 2024 Business-Driven Coaching Culture report, 68% of first-time managers report anxiety about being expected to "coach" rather than direct, despite lacking formal training in either pedagogy or therapeutic technique. The problem isn't that managers can't coach. It's that the business world has conflated coaching skills with the professional coaching industry, creating a false expectation that leadership requires becoming a neutral facilitator rather than a directive decision-maker.
This confusion costs teams time, clarity, and trust. When managers abdicate their responsibility to make calls—hiding behind Socratic questioning when a direct answer would save hours—they create friction disguised as givement. The goal isn't to transform every manager into a certified coach. The goal is to teach managers when to ask questions, when to give answers, and how to recognize which mode a situation requires.
What Actually Breaks: The Cost of Misapplied Coaching
When managers treat every conversation as a coaching opportunity, three specific failure modes emerge. First, decision velocity drops. A senior engineer doesn't need to be walked through discovering the architectural trade-offs between microservices and a monolith—they need you to tell them which constraints the business prioritizes this quarter so they can make the call. Second, team members stop asking questions. If every question triggers a 20-minute exploratory dialogue when they needed a yes-or-no answer, they learn to avoid their manager entirely. Third, accountability becomes murky. Coaching-style management often avoids direct feedback in favor of reflective prompts, which means underperformers don't know they're underperforming until a performance review catches them off guard.
A 2023 Harvard Business Review study on high-performing teams found that the most effective managers spent 34% of their one-on-one time giving direct guidance and only 19% using open-ended coaching questions. The ratio flipped for low-performing teams, where managers spent 41% of conversations asking exploratory questions and only 12% providing specific direction. The pattern held across industries: clarity outperformed curiosity when execution mattered more than discovery.
David Ohnstad saw this firsthand at Veeam when a data engineering lead spent six weeks "coaching" a junior analyst through building a reporting pipeline. The lead asked questions like, "What do you think the right granularity is for this aggregation?" and "How would you think you would validate this data quality?" The analyst eventually delivered a pipeline that worked but took three times longer than it should have. The lead later admitted, "I knew the answer in week one. I thought I was supposed to let him figure it out." The business cost wasn't just time—it was the opportunity cost of what that analyst could have built if he'd been told the standard approach upfront and coached only on the edge cases where judgment mattered.
The Directive-Collaborative Balance Framework
Effective management requires knowing which mode to use and when. The Directive-Collaborative Balance Framework breaks this into four quadrants based on two variables: problem complexity and team member experience level. For low-complexity problems with low-experience team members, default to directive guidance—tell them what to do and why. For low-complexity problems with high-experience team members, confirm alignment and get out of the way. For high-complexity problems with low-experience team members, use structured coaching: narrow the problem space, ask targeted questions within defined constraints, then provide a decision if they're stuck. For high-complexity problems with high-experience team members, shift to true collaborative coaching—your job is to surface blind spots, challenge assumptions, and help them think through second-order consequences.
The mistake most managers make is defaulting to one mode regardless of context. Junior managers over-index on directive because it feels safe and efficient. Managers who've been through leadership training over-index on collaborative because they've been told that's what "good leaders" do. Neither extreme works. A senior data scientist doesn't need you to walk them through statistical fundamentals—they need you to tell them which business metric the executive team actually cares about so they stop optimizing for the wrong thing. A junior product manager building their first roadmap doesn't need you to ask, "What do you think the prioritization framework should be?"—they need you to teach them RICE or MoSCoW, then coach them on applying it to edge cases.
The framework's counterintuitive step is step three: give the answer when coaching stalls progress. If you've asked two clarifying questions and the person is still stuck, continuing to ask questions isn't giveing them—it's wasting time and eroding their confidence. According to Gartner's 2024 Leadership Effectiveness Study, managers who capped coaching sessions at three questions before providing guidance had 22% higher team satisfaction scores than those who insisted on "letting the team figure it out." The skill isn't endless patience. The skill is recognizing when exploration has diminishing returns.
David Ohnstad applies this when reviewing data architecture proposals. If an engineer presents a design with a clear technical flaw, he doesn't ask, "What do you think might go wrong with this approach?" He says, "This will break under load because you're doing a full table scan on every query—here's why, here's the pattern that fixes it, now apply that logic to the next section and let's review again." The engineer learns the principle faster, respects the feedback more, and doesn't waste a week debugging something that could have been caught in five minutes. Coaching works when someone has the foundational knowledge but needs help connecting dots. Directive guidance works when they're missing a foundational piece and need to be taught it directly.
Myth #1: Good Managers Always Ask Questions, Never Give Answers
This myth persists because leadership development programs—many adapted from professional coaching certifications—teach that questions unlock insight while answers create dependency. The logic sounds right: if you give someone a fish, they eat for a day; if you teach them to fish, they eat for life. The problem is that most workplace situations aren't about teaching someone to fish. They're about confirming that the person already knows how to fish, or quickly teaching them where the pond is so they can start catching fish tomorrow instead of spending three weeks philosophically exploring the concept of ponds.
What's actually true: managers need to give answers frequently, especially early in a team member's tenure or early in a project. The goal isn't to make people dependent—it's to establish a shared baseline so future conversations can focus on judgment calls rather than foundational questions. When David Ohnstad onboards a new product manager, he doesn't ask, "How do you think we should structure sprint planning?" He says, "Here's how we do it, here's why we made these specific trade-offs, here's where you have discretion to adjust, and here's where you don't." The new PM gets to productive work faster, learns the reasoning behind the structure, and knows which parts of the process are negotiable.
A 2024 MIT Sloan study on management training outcomes found that managers who were taught to balance directive and collaborative modes had 31% better team retention than those trained exclusively in coaching-based leadership. The teams didn't feel micromanaged—they felt supported. The difference mattered most for remote teams, where the cost of a misunderstood question is an entire day lost to asynchronous back-and-forth instead of a two-minute clarifying conversation.
Myth #2: Coaching Skills and Professional Coaching Are the Same Thing
The phrase "coaching culture" has become shorthand for giveing teams, but it conflates two unrelated activities. Professional coaching—what a credentialed ICF coach does—requires neutrality, non-directiveness, and a client-driven agenda. The coach doesn't have opinions about what the client should do. The coach helps the client surface their own answers. That works beautifully when the client is a CEO trying to decide between two strategic paths and needs space to think. It fails catastrophically when a manager uses it on a junior employee who genuinely doesn't know the answer and is hoping their manager will teach them.
What's actually true: managers need coaching skills—active listening, asking clarifying questions, summarizing to confirm understanding—but they should almost never act like professional coaches. A manager has context the employee doesn't have. A manager has accountability for outcomes the employee isn't responsible for. A manager is paid to make calls when the team can't converge on one. Pretending to be neutral when you're not wastes everyone's time and erodes trust. According to Forrester's 2023 Leadership Development Benchmarking Report, 74% of employees preferred managers who "tell me what you think, then ask for my input" over managers who "only ask questions and never share their perspective."
David Ohnstad encountered this when a Veeam engineering director started using pure coaching-style questions in technical design reviews. The director would ask, "What trade-offs do you see here?" instead of saying, "This won't scale past 10,000 users because the database writes aren't batched." The engineers interpreted it as the director not knowing the answer, which damaged credibility. When the director switched to, "Here's the scaling issue I see—do you agree, or am I missing something?"—the reviews became faster, the feedback landed better, and the engineers trusted the director more because the expertise was visible instead of hidden behind facilitation.
Myth #3: Directive Leadership Means You Don't Trust Your Team
This myth persists because corporate culture has pathologized decisiveness. Telling someone what to do has been rebranded as a failure of givement, a lack of psychological safety, or evidence that you hired the wrong people. The fear is understandable: nobody wants to be the micromanaging boss who stifles creativity. But the pendulum has swung so far that many managers are now afraid to make any decision without facilitating a consensus-building process, even when the decision is straightforward and the team is waiting for someone to just make the call.
What's actually true: directive leadership, done well, is a signal of respect. It says, "I have information you don't have, or I've made this mistake before and I'm going to save you time by telling you what works." High-trust teams don't need every decision to be collaborative. They need clarity about which decisions are collaborative and which are directive, and they need managers who can explain the reasoning either way. A manager who says, "We're using Postgres for this, not MongoDB, because our data access patterns are relational and I've seen MongoDB cause problems in this exact scenario—but I'm open to revisiting if you find evidence I'm wrong" is being directive and trustworthy.
Gartner's 2024 Workforce Analytics Survey found that teams with managers who made unilateral technical decisions 40% of the time reported higher autonomy scores than teams whose managers insisted on consensus for every choice. The reason: when managers absorb the low-leverage decisions, the team has more cognitive space for the high-leverage ones. Debating which CI/CD tool to use is a waste of time if the manager has already evaluated three and knows which one integrates with the existing stack. Debating the product roadmap priority is worth the time because the team's input changes the outcome.
David Ohnstad applies this when prioritizing data product features. He doesn't ask the engineering team to vote on whether to build a new integration or fix technical debt. He makes the call based on business priorities, user impact data, and strategic timing—then explains the reasoning in a Slack thread so the team understands the trade-offs. The team doesn't feel excluded. They feel like their time is being respected because they're not being asked to weigh in on decisions where they lack the full context.
Myth #4: If You Have to Tell Someone What to Do, You Hired Wrong
This myth is the final boss of the "coaching culture" movement. It suggests that A-players never need direction, and if you're giving direction frequently, it's evidence of a hiring or delegation failure. The myth is seductive because it lets managers avoid the discomfort of being directive. If the only acceptable reason to give an answer is that you hired poorly, then giving answers feels like admitting failure. So managers contort themselves into asking leading questions that everyone knows are just instructions with extra steps.
What's actually true: even A-players need direction in new contexts, ambiguous situations, or when speed matters more than perfect alignment. Hiring great people doesn't mean hiring people who intuitively know your company's priorities, your technical constraints, or the political history of why a certain stakeholder will kill a proposal if it's framed the wrong way. It means hiring people who can execute well once they have that context. Providing context quickly is directive leadership. Withholding it and hoping they'll discover it through trial and error is negligence disguised as givement.
According to McKinsey's 2024 Talent Development Report, high-performing teams averaged 12 minutes of directive guidance per week per team member, compared to 4 minutes for low-performing teams. The high-performing teams weren't micromanaged—they were getting the information they needed to make better decisions independently. The low-performing teams were "giveed" to figure things out on their own, which meant they spent hours solving problems their manager could have answered in two minutes.
David Ohnstad saw this when Veeam hired a senior data engineer from a competitor. The engineer had a decade of experience and didn't need to be taught how to build pipelines. But he did need to be told that Veeam's executive team prioritized deployment speed over architectural purity for the next two quarters, and that proposals framed as "technical elegance" would get deprioritized while proposals framed as "time-to-market" would get funded. That's not a hiring failure. That's onboarding. The engineer adapted immediately, delivered faster, and later said, "I wish my last manager had just told me how decisions actually got made instead of making me guess."
When Coaching Actually Works: The 80/20 Rule for Managers
The solution isn't to abandon coaching. It's to recognize that 80% of management conversations benefit from directive clarity, and 20% benefit from coaching-style exploration. The 20% are the high-stakes, high-ambiguity moments: career development conversations, strategic pivots, interpersonal conflicts, or situations where the team member has more domain expertise than you do and needs help structuring their own thinking. Those are the moments where asking, "What are you optimizing for?" or "What would change your mind?" unlocks value that a directive answer couldn't.
The other 80%—task delegation, technical feedback, priority alignment, process clarification—benefit from speed and specificity. A product manager doesn't need to be coached through the discovery that the CEO wants the feature shipped before the board meeting. They need to be told that, then coached on how to cut scope without sacrificing the core value proposition. The directive part takes 30 seconds. The coaching part takes 20 minutes. Both are necessary. Neither replaces the other.
Recognizing that building high-performing teams around data products specifically demands clarifying governance models upfront, since ambiguous data ownership creates the friction that coaching alone cannot resolve. And leaders adopting a coaching culture need technical literacy to understand AI agent capabilities and limitations when integrating them into team workflows. The directive-collaborative balance applies to both: governance decisions are directive (someone has to own the call), while workflow integration is collaborative (the team knows their pain points better than you do).
How do you know when to coach versus when to direct as a manager?
Coach when the team member has the foundational knowledge but needs help prioritizing, structuring their thinking, or navigating ambiguity. Direct when they're missing key context, facing a time-sensitive decision, or solving a problem you've already seen fail. If two clarifying questions don't move them forward, switch to directive mode and explain the answer directly.
What is the difference between coaching skills and professional coaching in a management context?
Coaching skills—active listening, asking clarifying questions, summarizing to confirm understanding—are tools managers use within directive leadership. Professional coaching requires neutrality and a client-driven agenda, which doesn't work when the manager has accountability for outcomes and context the employee lacks. Managers should use coaching skills frequently but rarely act like professional coaches.
Why do directive managers often have higher-performing teams than coaching-focused managers?
Directive managers who explain their reasoning provide clarity faster, which lets teams execute on high-leverage work instead of spending time solving low-leverage problems. According to Gartner's 2024 study, teams with managers who made unilateral decisions 40% of the time reported higher autonomy because they weren't being asked to weigh in on decisions where they lacked full context, freeing cognitive space for strategic work.
For practitioners: Default to directive when onboarding, setting priorities, or providing technical feedback. Save coaching for career conversations, strategic ambiguity, and moments where the other person has expertise you don't. Speed and clarity aren't the opposite of givement—they're what make givement possible.
For leaders: Stop measuring management effectiveness by how many questions a manager asks. Measure it by whether their team knows what's expected, understands the reasoning behind decisions, and has the context to make good calls independently. A manager who gives answers frequently but explains the reasoning is building capability faster than one who withholds answers in the name of development.
When did you last audit whether your one-on-ones are actually helping your team move faster, or just making you feel like you're "doing leadership" the way the training deck said you should?
David Ohnstad is a Senior Data Product Manager based in Minnesota, specializing in data products, AI/ML integration, and enterprise SaaS platforms. Follow his work at github.com/davidohnstad40-netizen. Read more of David Ohnstad's data product management writing and David Ohnstad on AI and enterprise SaaS.
Top comments (0)