Here’s a number worth sitting with.
In Lenny’s 2026 survey of how tech workers are feeling, manager effectiveness came out as the strongest driver of burnout in the entire dataset. So much that it beat role, company size, and even the hot topcic du jour ‘AI anxiety’. Workers with a highly effective manager on the other hand reported roughly 65% higher job enjoyment and dramatically lower burnout.
And yet, only about a quarter of tech workers rate their manager as highly effective. More than a third rate theirs as outright ineffective. Numbers that have barely moved in a year.
Just to bring that point home: the most powerful retention lever in tech is the one that’s mostly ignored. In a 2x2 of importance and focus, manager effectiveness sits in the “ignore at your own peril” quadrant.
But before we get to addressing that, a harder question hides inside the survey: effective at what exactly?
What does an effective manager really look like? Link to heading
In my book an effective manager is a combination of three factors, sadly they don’t usually live naturally in the same person. The good news is that they can be learned over time in the right environment.
The first is human skill: the ability to motivate, to relate, to be compassionate, and to articulate the reasoning behind a decision at a level the audience can understand. People don’t have to agree with you, but the ability to articulate (or better relate) the rationale behind a decision goes a long way in my experience.
The second is earned opinion: being decisive and mature enough to make a call and then own the outcomes. This is where “nice” managers often fall down, because they conflate niceness with agreement. A good manager can be perfectly warm, human and nice and still tell you no. “Nice” does not mean “agree at all times” or “avoid any confrontation”. In here is also just raw competence. It’s not “have an opinion”, it’s deliberately phrased “earned opinion”. It’s competence and good decision making that earns the right to an opinion, not just being shouty.
And then there’s the third property that matters more than most people expect: consistency. Working with the same group of people gives them a longitudinal view of your decisions. So perception will shift far less on any single decision and far more on whether you’re the same person from one week to the next. This carries a lot of weight, particularly with engineers whose working world runs on deterministic, consistent systems: the same input produces the same output, every time. A manager whose decisions are unpredictable feels, to them, like a flaky dependency and someone they will quietly start routing around.
With that we have a working frame of mind for “effective manager”.
The most neglected lever Link to heading
The number in Lenny’s report never budges over the years in part because good management is slow to grow and easy to defer. There’s always something more important that growing and supporting new managers. In practice you promote your best engineer, hand them a team, give them no support, and call it a management layer. In steady state these new managers can just about get away with developing their way to effective managers over years.
In a crisis however you can’t rely on this. A departed CTO, a demoralised team, a runway that’s ticking is precisely the window where the ineffective manager does the most damage and takes the longest to learn on its own. Which is the whole reason experienced drop-ions exists: a new role is first and foremost being a good manager, on a deadline.
The reason I call it on a deadline is because that initial period is about gaining tust, bringin stability, a sense of healing and reducing uncertainty. Because no matter how clearly signaled, well intended and otherwise amicable. a new leader means new focus, new expectations and a period of settling down.
What the job actually looks like in those early days Link to heading
Let me make it concrete, since I can speak from plenty experience here.
The CTO had just left which meant the business went from 3 founders to 2. What he left behind wasn’t really a technical problem. The code still compiled, the servers still ran. What he left behind was a technical team that had lost its rudder, and a leadership team that lost a leg of the stool. And the reflexive move to “find a new CTO, fast” is an obvious a trap, because you cannot hire a great CTO in a panic, and the team cannot sit rudderless for the months a proper search takes. Classic catch 22.
So I came in to hold the fort, and the job fell into three areas of focus. Notably, the least important one was technical.
Stabilise the team first. This is not all cold, hard business. A team that’s just lost its leader is anxious, and anxious people do their safest work, or they leave. Safe cozy work however is not what you need when you’re in scale-up mode. Before anything else, steady the team: be visible, be honest about what you know and don’t, and give them a reason to believe the ship still has a course. Get this wrong and every other fix is moot because the people who understand the system walk out the door.
Build a baseline of process. A thin line between bureaucracy and structure. A rudderless team isn’t suffering from too few rules, it’s suffering from no shared sense of how decisions get made. A lightweight, dependable cadence gives people something solid to stand on while the bigger questions get answered. Structure, for pepole, is a form of reassurance.
Help the founder see the shape of the next CTO. This is the area of focus that makes interims unique. You start by designing yourself out of job. The polar opposite of permanent employees. The founder didn’t need me to be the next CTO, they needed help understanding what the next CTO should look like for the company they were becoming, not the one they’d been. That profile is the most valuable thing I left behind. (this is the same idea as The 4 Stages of a CTO — the leader a company needs changes with its phase, and hiring for the last phase is how you end up back in this room in two years with bigger problems.)
The exit is the point, and you plan it on day one Link to heading
Three months later, I was out. The founder had a clear plan: the VP Eng could step up while they raised their next round and ran a proper search for the CTO the company actually needed.
That clean exit was the design from the start.
While I’m very open to long term permanent roles, even those haves an end built into it. Sometimes that’s 12 months, sometimes eighteen, sometimes a few years to deliver on the vlaue creation plan. Whatever it is, tehre is always an end goal.
Reaching that goal doesn’t mean an exit, it might just be the role naturally redefines as the business grows. The concern is inertia.
Same craft, different weather Link to heading
It’s tempting to hear all this and conclude that there are two different tiers of manager. They’re not. A short termer has to clear exactly the same bar I set out earlier. I’ts about warmth, earned opinion, consistency. If anything, the bar has to be cleared more, because a team in crisis needs effective management more, not less. A fixer who isn’t a genuinely good manager isn’t a fixer; they’re an arsonist with a job title. This is not BAU-manager versus fixer, as if one were the lesser version. It’s effectively the same craft.
The difference is the situation in which each one naturally thrives.