How to Find the Best Engineering Leadership Coach: 7 Tests

What the director of engineering seat does to the people in it, what the evidence says a coach can and cannot fix, and seven tests that tell you in one call whether this one can do the job. Facts first; my own answers last.

Who is the best engineering leadership coach? Nobody honest will give you a name. The right coach for a first-time engineering manager is the wrong one for a director who started managing managers last quarter, and neither is the right one for a VP of Engineering weighing the CTO title. So this is not a list. It is the part of the search results that is usually missing: what the director and VP seats do to the people who hold them, in numbers with sources; what the research says a coach can change and what it cannot; and seven tests you can run in a single intro call to find out whether a specific coach can do the specific jobs your seat needs done. It is meant to be worth reading whichever coach you pick, or if you pick none. My own answers to the seven sit in one section at the end, so you can weigh them as one candidate's. The lens is my own: I coach directors of engineering and engineering leaders and ran engineering as a VP over more than 1,000 engineers.

Who this is for: directors of engineering, heads of engineering and senior engineering managers running two or more teams through a manager layer, and VPs of Engineering who share a roadmap with a product leader. If you are a tech lead or a first-time manager of engineers, most of the tests still apply, but the marketplace mentors and the engineering-manager training programs are usually the better first call, and the numbers below say why. If you are the VP or CTO buying coaching for a director on your team, run the seven tests on the shortlist and compare the answers side by side.

What the seat does to the people in it, by the numbers

Start with what happened to the job in the last year. When 617 engineering leaders were surveyed in March 2025, two in three said their scope had widened and two in five had picked up more direct reports, while only 3% had less on their plate than before. More than a third were working longer weeks. Asked about burnout, 22% put themselves at the critical level and a further 24% at moderate. Two in five said their team was less motivated than it had been twelve months earlier.¹ The cuts, where they came, fell on the middle: among companies that reduced their manager count, 67% took out line managers and 56% took out the directors and other managers of managers. Nearly half of the people managing engineers, 47%, had thought about going back to an individual-contributor job in the previous year; among those managing managers the figure was 27%, and the reasons people gave for staying were pay, not wanting to lose their technical edge, and liking the work.¹

Then consider how much of the outcome runs through you. Gallup's two-decade dataset, 27 million employees across 2.5 million work units, attributes at least 70% of the variance in engagement to the manager. The same research estimates that roughly one person in ten has high talent for managing, that two more in ten can perform at a high level if coached, and that organizations get the manager hire wrong about 82% of the time.² Gallup's worldwide survey for 2025 adds the trend: manager engagement dropped five points in a single year, 27% to 22%, the sharpest fall it has recorded, while engagement among all employees slid to 20%, the lowest reading since 2020.³ For a director that arithmetic is the job itself. The engineers are no longer yours to manage; the managers are yours to pick and to grow. Google's Project Oxygen, its own study of what separates its best managers, ranks "Be a good coach" first among ten behaviors, before empowering without micromanaging, communicating effectively, supporting career development, setting a clear vision and strategy, and keeping key technical skills, and reports that teams under highly effective managers performed better, were happier and lost fewer people.⁴

AI is the newest line on the same desk. Asked in late 2024, 77% of software engineering leaders rated building AI capabilities into their applications a significant or moderate pain point, and 71% rated the use of AI tools in the engineering workflow the same way.⁵ A separate 2025 survey of more than 640 CTOs, engineering managers and engineers found nine teams in ten using AI coding tools, an average reported velocity gain of 25%, and 61% of engineering budgets growing as a share of company revenue.⁶ The mandate to lead that change arrives from above and from below at once, and it rarely arrives with a plan.

Two more numbers describe the seat itself. Engineering executives change roles roughly every three years, and when the seat above yours opens, 68% of the time it is filled from outside rather than by the director already in the building, 77% in North America; the leaders who move say they are looking for autonomy, full ownership of a product, or a strategic seat at the leadership table.¹² And the step onto the ladder is where most of the attrition happens: a VP of Engineering at Figma, previously Slack, puts the share of newly promoted engineering managers who leave the post and go back to being an individual contributor at about half, an estimate rather than a study, but from someone who has run the promotion many times.¹³

The last number explains why guides like this one exist. When Stanford and The Miles Group surveyed more than 200 CEOs, board directors and senior executives, close to two-thirds of the CEOs were getting no coaching or leadership advice from outside their company and nearly half of the senior executives were getting none, yet almost all of them said they would take it. The capability they most wanted to build was conflict management, followed by mentoring and developing talent, and delegating and sharing leadership.⁷ Read that list again with a director of engineering in mind: it is the whole job.

What a coach can and cannot do, according to the evidence

Two meta-analyses carry most of the weight here, and both land in the same place: coaching works, and the effect is real but moderate. Pooling the organizational studies available in 2014, researchers found positive effects on every outcome they tested, ranging from 0.43 on coping up to 0.74 on goal-directed self-regulation, with performance and skills at 0.60 and an overall figure of 0.66.⁸ Two years later a second team pooled 17 workplace-coaching studies and reported an overall effect of 0.36, stronger for individual-level outcomes and for how people felt about their work than for measured skills.⁹ Translated: expect better goal-setting, self-management, coping and attitude, and a moderate lift in performance. Do not expect a number with a dollar sign on it. Neither analysis covered engineering leaders, the studies are few, and no one has run a controlled trial on directors of engineering; when a coach quotes a return on investment, ask which study produced it.

What a coach cannot do is run your organization, write your roadmap, or replace the other kinds of help. In engineering the categories blur more than in most functions, so it is worth drawing them. A mentor has held the seat and tells you what they did; the value is pattern recognition, the risk is that their company was not yours. A consultant, including the fractional CTO, fixes the problem for you and leaves; the value is speed, the risk is that you did not learn to do it. Technical mentoring works on the code and the architecture. Leadership coaching works on you: how you decide, how you lead through other people, and how you show up in the executive room, with someone who has no stake in the outcome and enough context to know what the decision feels like. Interview coaching, which crowds the search results for engineering manager and director roles, is a fourth thing again: it gets you the seat, it does not help you hold it.

The three kinds of coach you will meet in engineering

The coaches who rank for engineering leaders sort themselves into three kinds, and most say so on their own pages. The former operator: a past engineering manager, VP of Engineering or CTO who now coaches, and who will tell you plainly that sometimes clients want answers, not questions; the strength is credibility and speed, the weakness is the square peg, the operating playbook that worked at their company applied to yours. The trained coach without an engineering past: often an ICF-credentialed professional who is excellent at the listening, the questions and the psychology; the strength is the method, the weakness is that they cannot read your org chart or your architecture debate, and generic management training rarely survives contact with an engineering team. The marketplace mentor: an engineering leader still doing the job, booked by the month through a platform, at a fraction of the price; the strength is currency and cost, the weakness is that it is advice, not development, and the platform vets the profile rather than the outcome. Credentials remain table stakes: in the profession's own 2023 study, 80% of coaches agreed that clients expect a coach to be certified or credentialed.¹⁰ Know which kind you are buying, and which you need this quarter.

Seven tests, one call each

Seven jobs, seven tests. For each one: the capability it checks, the question that checks it, and the shape of an answer that passes. Run all seven on every name on your list; a recommendation gets a name onto the list, it does not get it off the hook.

Test one: have they managed managers, at a scale beyond yours?

What it tests: whether they know the job you are actually doing. Managing managers is the step every guide names and few coaches have taken; it is also where the engagement variance gets decided.² What to ask: how many people did you lead through other managers, at how many layers, and what changed for you at that step? What a good answer sounds like: a number, a company, and one specific thing they got wrong the first time they delegated an outcome instead of a task.

Test two: can they fix your operating rhythm without slowing delivery?

What it tests: the move from firefighting to strategy, which is the sentence on almost every ranking page for this seat. What to ask: describe my week, then ask what they would change first: the planning cadence, the escalation path, or the meeting where the roadmap gets renegotiated. What a good answer sounds like: they ask how capacity is allocated between the roadmap, reliability and the sales asks before they ask about your calendar, and they can name a rhythm they installed that let the teams keep shipping while the director stopped being the referee.

Test three: can they translate your work upward?

What it tests: executive communication, the skill the VP and CTO evaluate you on whether or not they say so. Strong delivery does not travel upward by itself. What to ask: here is the platform decision I have to defend next month; how would you frame it for the executive team? What a good answer sounds like: they turn architecture and reliability into outcomes, risk and a trade-off in under a minute, and they have presented to an executive team or a board themselves rather than coached other people to.

Test four: do they know the seam between product and engineering?

What it tests: whether they have lived the most strained relationship in a growth company from either side. What to ask: our product lead and I disagree about what ships this quarter; what do you do with that? What a good answer sounds like: one owner for the business metrics, one scoreboard both of you are measured on, and a story about a time they made that pairing work, with the numbers it produced.

Test five: can they help you hire and grow managers, and have the hard conversations?

What it tests: manager quality, promotions, feedback and performance, the work most of the hours in this seat go to and the area executives most want help with.⁷ What to ask: I have a manager who is loyal, liked and not good enough for the next size; walk me through the next ninety days. What a good answer sounds like: a sequence, not a slogan, and a clear view on when to develop, when to move and when to part, from someone who has done all three.

Test six: can they lead AI adoption without stalling the team?

What it tests: whether they treat AI as a people problem before a tooling problem. Nine in ten teams already use the tools;⁶ the difference between the teams that resist and the teams that embrace shows up in productivity, and the pressure to decide sits with you.⁵ What to ask: what should I say to the executive who wants an AI plan by Friday and to the senior engineer who thinks it is a fad? What a good answer sounds like: a bet worth making, a way to measure developer productivity you could defend upstairs, and no vendor names in the first five minutes.

Test seven: will they help you decide the VP or CTO question, on purpose?

What it tests: whether they have held the seat above yours, and whether they will tell you the truth about it. "Is VP of Engineering higher than director" is the question people type; the real one is whether you want the job that owns strategy, hiring the leaders and the plan the CEO and board can believe. What to ask: what does a VP of Engineering do that I do not, and how would you know if I were ready? What a good answer sounds like: a plain description of the seat above yours from the inside, and a plain answer about what they would say if you were not ready yet. While you are at it, settle confidentiality: if your employer is paying, ask what, if anything, gets reported back. A coach worth hiring will say that the sessions belong to you, that the company learns whether you showed up and whether the goals are moving, and that this goes in the agreement before session one.

What it costs, who pays, and the terms to insist on

The profession's own census puts the mean price of a one-hour session at $244 in 2022, a 9% rise over three years, across roughly 109,000 practising coaches.¹⁰ For engineering leaders the spread is much wider than that average suggests. At the low end, the marketplaces list engineering-leadership mentors at monthly rates of about $150 to $300.¹¹ At the other end, the solo coaches who rank for the director and VP seats publish anything from about $2,400 a month on a six-month program to $2,500 to $10,000 a month for CTO-level engagements. How the price moves with seniority, who normally pays, and my own pricing are laid out in my guide to what executive coaching costs. From director upward the bill usually goes to the employer's learning or leadership-development budget, a fact the coaches in this field state on their own pages. Frame the request as a development plan: one goal the business owns, one goal you own, a checkpoint at six months.

On terms, the free thirty-minute intro is the norm in this field, and it doubles as the fit test, so take one with every candidate. The sound structures are month to month with no lock-in, or a fixed block of three to six months with a review at the end; a contract that needs a year to demonstrate value is selling the contract. Weekly or fortnightly is the usual cadence, with a way to reach the coach between sessions when the outage or the resignation will not wait.

Red flags

A coach who has never managed managers, coaching a director on how to.
Interview coaching sold as leadership coaching: it gets you the seat and does not help you hold it.
Generic management training with the word "engineering" added: frameworks built for a stable corporate function rarely survive an engineering team, and the ranking coaches say so themselves.
A fractional CTO engagement dressed as coaching: if they are doing the work, you are not learning it.
A return-on-investment percentage with no study behind it, or a promise about the promotion.
No line between coaching, mentoring and consulting: the good ones name which they are doing and when.
Unreachable between sessions in a seat where the outage, the resignation and the reorg do not wait for the calendar.

Where to look

How do you find a reputable one? Three places, in this order. First, ask the people who have bought this before: your own VP or CTO, directors you rate at other companies, and whoever owns the learning budget, who usually keeps a shortlist. Second, the curated lists that engineering newsletters and communities maintain, remembering that they are lists of names and not of outcomes. Third, the coaching marketplaces, which let you filter by title and price and are the sensible starting point for a first-time manager. Whatever the source, vet the background against your transition, and every name then goes through the seven tests above, the recommended ones included.

My own answers to the seven, for the record

Test one, managing managers: at Cox Automotive I ran engineering as a VP with more than 1,000 engineers below me and a full layer of directors and managers in between; at New Relic I was EVP over 450 people in product and engineering. Test two, the rhythm: the New Relic system, one owner for the numbers, capacity allocated each quarter, a single escalation meeting and a customer advisory board, is described on the seat page; it turned into $50M-plus of incremental ARR from new products. Test three, upward: I have presented to that company's board and given the keynote at its customer conference. Test four, the seam: I have sat on the product side of it as a CPO, at Coursedog, where the product and engineering organization was rebuilt in nine months. Test five, the hard conversations: I stepped into New Relic after my predecessor was let go, and later ran a turnaround as COO at Blameless. Test six, AI: I built and launched an AI platform as a CEO, I write about adoption as a people problem, and the tech geek who started with assembly at ten never left: two inference servers at home run a house that senses when I walk into a room. Test seven, the seat above: VP of Engineering, CTO, CPO, COO and CEO, the last of them at the company that became Web.com, which went public on Nasdaq; I can describe each from the inside. I am ICF and CEC certified. The rest is on the director of engineering coaching page, and the three-proof test from my general guide still holds: has this person held the seat above yours; are they trained and certified as well as experienced; and after one real conversation, did you hear something true you had not said out loud yet.

Questions engineering leaders ask before they choose

Who is the best engineering leadership coach?

The one who can do the seven jobs your seat needs next and prove it in one call: has managed managers at a scale beyond yours, can fix the operating rhythm without slowing delivery, can translate engineering work upward, knows the product-and-engineering seam, can help you hire and grow managers, can lead AI adoption without stalling the team, and has held the seat above yours. There is no single best; the guides that rank on the question all say it depends on the seat and on what is breaking.

Which coaching is best for engineering leaders?

The kind that matches the transition you are in. Engineering leadership coaching exists because the people-management, strategic and scaling skills the job needs are the ones an engineering education leaves out, and the right coach is the one whose own technical management background matches your transition: individual contributor to manager, manager of managers to executive, or VP to CTO. A first-time engineering manager is usually best served by a mentor or a marketplace coach who is still doing the job, plus a management fundamentals program. A director managing managers for the first time needs leadership coaching with engineering context, from someone who has held the seat above, including the call on when to stay hands-on. A VP or CTO needs executive coaching that can go to the board. Technical mentoring and interview coaching are different products and should not be bought by mistake.

Does executive coaching actually work?

Yes, by a moderate and well-documented amount. The two meta-analyses that pool the organizational research put the overall effect between 0.36 and 0.66, strongest on goal-setting, self-regulation, coping and attitude to work, and moderate on performance. Engineering leaders were not studied as a group, so treat any percentage return on investment as a marketing claim until someone shows you the study.

What is the difference between an engineering coach, a mentor and a fractional CTO?

A mentor has held the seat and tells you what they did. A fractional CTO does the work and leaves. A coach works on how you decide and lead through other people, with no stake in the outcome. The best engineering coaches can switch between the three inside one session and will tell you which one they are doing; the ones to avoid cannot tell the difference.

What is the 70/30 rule in coaching?

The informal expectation that the client does about seventy percent of the talking in a session and the coach about thirty. No body enforces it and nobody measures it, but it is worth noticing on the intro call. Engineering leaders often want more telling than that, which is fine when you asked for a mentor; if the coach filled most of the hour uninvited, you were being pitched.

How much do leadership coaches charge?

The average fee for a one-hour session across all coaching was $244 in 2022. For engineering leaders the band is wide: marketplace mentors from about $150 to $300 a month, solo coaches who rank for director and VP seats from about $2,400 a month to $2,500 to $10,000 a month at CTO level. The company usually pays at director level and above. Rates by seniority and how I price are in my guide to what executive coaching costs.

Is VP of Engineering higher than director?

In most companies, yes: the VP runs engineering for the whole company and sits with the executives, while a director runs a group of teams inside engineering. "Head of engineering" is used for either, depending on size. What separates the two is the room rather than the headcount: a VP is judged on strategy, on the leaders they hire, and on whether the CEO and the board believe the plan. The seventh test above is built around that difference.

Can my company pay for engineering leadership coaching?

At director level and above the employer usually does, from the learning or leadership-development budget, and the coaches who rank for this seat say the same on their pages. Put it forward as a development plan: one goal the business cares about, one you care about, a six-month checkpoint, and a written sentence on confidentiality so everyone knows what the company will and will not hear.

How to Find the Best Engineering Leadership Coach: Seven Tests