CTO responsibilities change so much between a two-person founding team and a company with four engineering directors that a single checklist can't cover both. Most guides try anyway: they list architecture, hiring, and infrastructure as if those things carry the same weight and the same hours regardless of headcount. They don't. What a CTO owns at pre-seed and what a CTO owns at Series B are close to two different jobs wearing the same title.
Most published breakdowns solve this the wrong way. Toptal names four core CTO engagement models, fractional, virtual, on-demand, and part-time, but never says which one fits a five-person startup versus a fifty-engineer one. What's missing is a map from company stage to what actually needs owning right now, not another inventory of responsibilities.
What does a CTO actually do?
Strip away the stage variation and a CTO owns four things at every company: the technical architecture, the infrastructure the product runs on, the engineering team's structure, and how engineering spends its budget. That's the company-agnostic version, and it's also the least useful one, because none of those four things stay fixed. Architecture decisions that take an afternoon at five people take a quarter-long migration project at fifty. A hiring plan that means "find the next two engineers" at seed round means "build a leveling framework and a management layer" at Series B. The real answer depends on which of those two companies you're asking about.
Pre-seed and founding stage: hands-on ownership
At pre-seed, the CTO is usually still writing code most days. They're picking the stack, standing up the cloud account, wiring up CI, and making MVP architecture calls under real time pressure, decisions that are cheap to make now and expensive to unwind once customers depend on them. Hiring, if it happens at all, means a handful of engineers, interviewed and onboarded personally, because there's no one else to delegate that to.
A lot of companies at this stage don't have a CTO in the formal sense. They have a technical co-founder or a founding engineer covering the same ground without the title, and for a team building toward a first paying customer, that's frequently the right call rather than a compromise. Hiring developers for your startup at this point means weighing a founding engineer against a fractional CTO before you write a single job post, since the stack and seniority calls made here are the hardest ones to reverse later.
Seed to Series A: building the team
Funding changes the job. Headcount climbs fast, and the CTO starts handing off hands-on coding to make room for hiring, onboarding, and the first real management structure. Interviews stop being one-on-one conversations and become a process with a scorecard.
This is also when the sourcing question gets real. Do you build the team in-house, bring in staff augmentation to move faster than direct hiring allows, or lean on an agency for a defined project? Which hiring model fits your stage is a different answer at ten engineers than it is at three, and getting it wrong here is an expensive way to learn the difference. The CTO also starts setting rituals that didn't exist before: sprint cadence, code review standards, an on-call rotation, especially once the team includes contractors or an agency working alongside full-time staff. How to structure a blended engineering team covers exactly this transition, where the roles stay the same but who fills them starts to vary.
Series B and beyond: strategic ownership
By Series B, the CTO's calendar looks nothing like it did at seed. Individual code review gives way to org design: adding engineering managers, deciding how many directors report directly, and figuring out where the technical org chart needs to split by product line versus stay unified. Budget ownership stops being "can we afford this tool" and becomes vendor negotiation across a real spend line. Board and investor reporting becomes a recurring part of the job, not an occasional favor.
Compliance ownership also lands squarely on the CTO at this stage, even though most responsibility lists skip it entirely. Data residency, security certifications, and, for any company processing EU customer data, GDPR, all become the CTO's problem to own. They stop being abstract once the company has enough customers and enough data that a compliance failure turns into a board-level conversation instead of a support ticket.
Do you need a CTO yet, or something else?
Not every company at this stage needs a full-time CTO. The real decision is between three options: a full-time hire, a fractional CTO, or continuing with a founding engineer or technical co-founder a while longer. What a fractional CTO actually covers walks through the cost and timing of that middle option in detail. The short version is that a fractional engagement picks up the same categories of ownership described above, scoped to part of the week instead of all of it. It's usually the right call once a company has outgrown a founding engineer's bandwidth but doesn't yet have the budget or headcount to justify a full-time seat.
Frequently asked questions
What is the difference between a CTO and a VP of engineering?
A CTO owns technical strategy and direction, architecture decisions, the technology roadmap, and increasingly, board and investor communication on the technical side. A VP of engineering owns execution: shipping velocity, team structure, and the day-to-day management of engineering managers and their reports. Smaller companies often combine both into one CTO role; the split usually shows up once the engineering org is big enough that strategy and execution genuinely need separate owners.
Does a startup need a CTO before it has funding?
Not necessarily. Plenty of pre-seed companies run on a technical co-founder or a founding engineer instead, and that's often the right structure for a small team building toward a first customer. A formal CTO hire, full-time or fractional, tends to make sense once there's a team to manage and technical decisions carry real consequences beyond the founders themselves.
Can one person be both CTO and lead engineer?
Yes, and at pre-seed and early seed stage, that's the norm rather than the exception. The combination gets harder to sustain as the team grows, because management and hands-on coding compete for the same hours. Most CTOs who also code full-time start delegating the coding first, well before they hand off any strategic ownership.
When does a company typically hire its first full-time CTO?
There's no fixed headcount trigger, but it tends to come once a company is hiring engineering managers, which for many startups lands around Series A or Series B, and once architecture, hiring, and vendor decisions need a dedicated full-time owner rather than someone splitting attention with a co-founder's other responsibilities. Companies that raise a large seed round with an aggressive hiring plan sometimes hit this point earlier.