Adam Li has watched the same mistake happen quite a bit in the past few years, and it isn’t usually a bad hire. It’s a good engineer put in the wrong seat. A founder promotes their strongest coder into a leadership role that asks them to do two incompatible jobs at once, and six months later the codebase is fine but the team is bleeding people, or the team is happy but nobody can explain why the architecture doesn’t hold together anymore.
Adam leads engineering search at Salient, and I sat down with him to get his perspective on a common scaleup question: engineering manager or tech lead, and which one first?
“People treat it like a seniority question,” he told me. “It’s a specialisation question. An EM and a TL aren’t two rungs on the same ladder. They’re two different jobs that happen to both have the word ‘lead.'”
Two jobs, not one ladder
The Engineering Manager, according to Adam, owns the humans and the machinery around them: 1:1s, performance reviews, career paths, headcount planning, the recruiting pipeline, the friction that builds up between engineering and product. It’s a role with formal authority attached to it. An EM can hire, fire, and set compensation.
The Tech Lead owns the system. Architecture decisions, coding standards, the pull requests nobody else on the team feels qualified to approve, the slow accumulation of technical debt that either gets managed or becomes a rewrite. A TL’s authority is earned rather than granted, built on the kind of credibility that comes from being right about hard technical calls often enough that people stop second-guessing you.
Adam estimates most EMs he places spend less than 10 percent of their week actually writing code. Most Tech Leads he places spend somewhere between 30 and 60 percent of theirs still shipping. “That number surprises founders every time,” he said. “They think a Tech Lead is basically a manager who codes on the side. It’s closer to the reverse — a strong engineer who’s picked up just enough leadership to keep the system coherent.”
I asked him whether the split ever gets fuzzy in practice. It does, he said, especially at companies under 50 people, where job titles tend to lag behind what people actually do all day.
The big question
Adam’s answer to “which one first” isn’t really about title preference. It’s about where the pain is currently showing up.
If a founder or CTO has eight or more engineers reporting to them directly, he said, that person is probably burning 20 hours a week on 1:1s and status syncs. That’s an Engineering Manager problem, and it tends to announce itself through attrition before anyone calls it out: good engineers leaving not because the work is bad, but because nobody sat with them long enough to build a case for promotion.
A different kind of pain shows up when the team is still small, under five engineers. Outages start clustering. Cloud costs creep up without an obvious cause. Junior developers sit blocked for a day and a half waiting on a design decision nobody has the standing to make. That’s a Tech Lead problem, and Adam said founders often misdiagnose it as a hiring-speed issue when it’s really an architecture-ownership issue.
“I ask clients one question before we even open a search,” he said. “Is the pain about people leaving, or about the system breaking? Almost every founder can answer that in under thirty seconds once I frame it that way. They just haven’t been asked.”
The hybrid that works right up until it doesn’t
Plenty of early teams, Li said, solve this by not solving it. Somewhere between three and six engineers, one senior person ends up doing both jobs. Salient’s internal shorthand for this is the Tech Lead Manager, or TLM, and Adam is candid that it works fine for a while.
“It works because at that size there isn’t enough of either job to fill a person’s week,” he said. “The trouble starts once the team crosses six or seven people. Now there’s a full-time people job and a full-time architecture job, and you’re asking one person to context-switch between an emotionally demanding performance conversation and a genuinely hard system design problem, sometimes in the same afternoon.”
Adam asserts that what tends to give first, is the people side. Faced with a sprint deadline and an overdue performance review, the TLM almost always ships the code and lets the review slide. It’s not a character flaw. Coding has a deadline attached to it and people management usually doesn’t, at least not one anyone enforces.
“I think the TLM model gets romanticised in startup culture as scrappy and efficient, and mostly it’s just deferred cost. Someone pays for it later, usually in the form of an engineer who starts interviewing elsewhere”.
Adam’s rule of thumb is: split the role once the pod hits six to eight engineers. Below that, a hybrid can hold. Above it, something is quietly failing, even if it isn’t visible yet.
What it costs to hire either one
We got into numbers, because Li’s clients always ask about numbers before anything else, and here’s where I’d flag that the range is wider than the averages people quote suggest. Compensation for both roles has moved noticeably across Sydney, Melbourne, and Singapore over the past two cycles, and the bands below reflect base salary only.
| Role | Sydney (AUD) | Melbourne (AUD) | Singapore (SGD) | Typical equity |
| Tech Lead (Senior/Lead) | $175,000–$215,000 | $165,000–$200,000 | $150,000–$190,000 | 0.15%–0.35% |
| Engineering Manager | $190,000–$240,000 | $180,000–$225,000 | $170,000–$220,000 | 0.25%–0.50% |
| Senior Engineering Manager | $240,000–$290,000+ | $220,000–$270,000+ | $210,000–$260,000+ | 0.40%–0.80% |
Note these exclude Australia’s mandatory 12 percent superannuation on top of base. Adam was upfront that the top of these bands moves fast in Sydney and Melbourne specifically, because the pool of engineers who’ve actually run a team through a Series B or C is still thin relative to demand. Singapore’s numbers, he said, have been more stable, though the sample he’s drawing from is smaller.
Final Thoughts
By the end of the conversation Adam kept coming back to the same idea: hire for the pain you can see, not the title you think you’re supposed to have at your stage. A company with ten engineers and no attrition problem might not need an EM yet. A company with four engineers and a system crumbling under them needs a Tech Lead more than it needs anyone with the word “manager” in their title.
“The mistake isn’t picking the wrong role,” he told me. “It’s picking a role before you’ve actually named the problem.”
