JM

Justin McKelvey

Fractional CTO · 15 years, 50+ products shipped

Fractional CTO 12 min read

What Does a Fractional CTO Do? (Day-to-Day, Hours, Real Examples)

A fractional CTO does for your startup what a full-time CTO would, but on a 5-15 hour/week basis: technical strategy, hiring engineers, code architecture review, vendor selection, security baseline, and translation between your engineering team and the business. As of June 2026, typical engagements run $8K-$15K/month with the CTO serving 2-4 concurrent clients.

I'm Justin McKelvey — I work as a fractional CTO with 3-4 companies at a time out of Austin, TX. This post is the day-to-day reality of the job, not the LinkedIn version. If you're trying to figure out whether you actually need one, read the "When you do NOT need a fractional CTO" section before booking anyone's calendar.

The 8 things a fractional CTO actually does

Most founders hire a fractional CTO expecting one specific thing — usually "review my code" or "help me hire engineers" — and then realize the job is broader than that. Here's what the work actually looks like across an embedded engagement.

1. Set technical strategy

This is the headline job. A fractional CTO owns the technical roadmap so the founder doesn't have to make architecture decisions from a Google search at 11pm. That means picking the stack, deciding what to build vs buy, sequencing engineering work against business milestones, and saying "no" when a feature request would torch the next six months of velocity.

  • Define the 6-12 month technical roadmap tied to fundraising or revenue milestones
  • Make stack decisions (Rails vs Next.js, Postgres vs SQLite, Vercel vs Railway)
  • Run build-vs-buy analysis on every meaningful component (auth, billing, search, AI)
  • Set engineering principles the team will follow when the CTO isn't in the room

2. Hire and evaluate engineers

The single highest-leverage thing a fractional CTO does is help you not hire the wrong person. A bad senior engineering hire at a 10-person startup is a 6-month setback. I sit in on technical interviews, write job descriptions that actually filter, and tell founders the truth about whether a candidate can do the job — even when the founder really wants them to.

  • Write role descriptions and leveling rubrics
  • Run technical screens and architecture interviews
  • Decide contractor vs full-time vs agency for specific work
  • Negotiate offers and equity ranges that match the market
  • Run 30/60/90-day check-ins on new hires

3. Review code architecture and tech debt

Not line-by-line code review — that's an engineering manager's job. A fractional CTO reviews the shape of the codebase: where the load-bearing walls are, where the duct tape is, and which pieces of tech debt are going to bite when you 5x users. Usually I do this in the first 30 days of an engagement and then quarterly after that.

  • Map system architecture and identify single points of failure
  • Audit tech debt and rank it by business impact
  • Recommend refactor projects with effort/payoff estimates
  • Review PRs that touch core architecture or new services

4. Manage vendor and tool selection

The average seed-stage startup is bleeding $1,500-$4,000/month on overlapping SaaS tools nobody chose deliberately. A fractional CTO owns the tooling stack — analytics, monitoring, hosting, CI/CD, AI APIs, feature flags, error tracking. The goal isn't to be cheap, it's to be deliberate.

  • Audit existing SaaS spend and consolidate overlapping tools
  • Negotiate enterprise contracts (often saves more than the CTO costs)
  • Pick infrastructure providers and exit strategies
  • Choose AI/LLM providers based on actual cost-per-token at your volume

5. Set security and compliance baseline

Most early startups are one breach away from a very bad week. A fractional CTO sets the minimum viable security posture — not SOC 2 on day one, but the basics that prevent the obvious failures. If you're heading toward SOC 2 or HIPAA, the CTO scopes the work and picks the compliance vendor.

  • Set up secrets management, MFA, and access controls
  • Implement backup and disaster recovery procedures
  • Run a basic security audit and threat model
  • Scope SOC 2 / HIPAA / GDPR work if it's coming
  • Set up incident response runbooks

6. Translate between engineers and business stakeholders

This is the underrated job. Founders pitch features in business language. Engineers respond in technical language. The translation layer is where most early teams break. A fractional CTO sits in product meetings and re-frames "we need to ship the AI thing" into "here are the three engineering questions we need to answer before we commit to a date."

  • Translate roadmap commitments into engineering tickets
  • Translate engineering risk into business decisions for the founder
  • Run sprint planning and retros if there's no EM yet
  • Communicate technical decisions to investors and board members

7. Lead technical due diligence

When you're raising a Series A or selling the company, an acquirer or VC will run technical due diligence — code review, architecture review, security review, team review. A fractional CTO either runs your side of that process or runs the buyer's process on the company you're acquiring. This single line item is often why founders hire a fractional CTO three months before a raise.

  • Prepare technical due diligence packets for fundraising
  • Run reverse due diligence on acquisition targets
  • Sit in on investor technical deep-dives
  • Write the technical sections of the pitch deck

8. Be the engineering escalation point

When production is on fire at 2am, someone needs to make the call about whether to roll back, page the team, or notify customers. When a senior engineer threatens to quit, someone needs to handle that conversation. A fractional CTO is the escalation surface — not the first responder, but the person the team escalates to when the founder isn't equipped to decide.

  • Decision authority during outages and security incidents
  • Handle escalations from engineering team conflicts
  • Final approval on major architecture or vendor changes
  • Mediate between engineering and product/sales when priorities collide

What a fractional CTO does NOT do

This is where most engagements go sideways — the founder hired a strategist and expected a senior engineer. Here's what's outside the role.

  • Doesn't write production code. Occasionally I'll prototype something to prove an architecture decision is feasible, but if you need someone shipping features, you need an engineer, not a CTO. A $10K/month fractional CTO writing CRUD endpoints is the most expensive engineer in the world.
  • Doesn't replace a senior engineer. A fractional CTO at 10 hours a week cannot do the work of a full-time senior engineer at 40 hours a week. If you don't have at least one engineer (contractor or FT), a fractional CTO will mostly tell you "go hire an engineer."
  • Doesn't do project management. Scrum mastering, ticket grooming, daily standups, sprint reports — these are EM or PM jobs. A fractional CTO can set up the system once, but they shouldn't be running it weekly.
  • Doesn't take operational on-call. If your pager goes off at 3am, the fractional CTO is not the person who responds. They're the person you call after the incident to decide what to change.
  • Doesn't replace a head of product. Some fractional CTOs have product chops, but the role is engineering leadership. If your bigger gap is "what should we build" rather than "how should we build it," you might need a fractional product manager first.
  • Doesn't manage your roadmap day-to-day. They set it, but the team executes it. If nobody on the team can drive a sprint, you have an engineering management gap, not a CTO gap.

A typical week as a fractional CTO

Here's what last week actually looked like for me across three active clients. Hours are rough — the real week is more interrupt-driven than this implies.

Monday morning — Slack triage (2 hours). I start the week by going through each client's engineering Slack channels. I'm reading PR discussions from the weekend, picking up on tension between engineers, spotting questions where the team is going in circles. I'll drop into 3-4 threads with a decision or a pointer to someone who already solved this. No meetings — just async reading and writing.

Tuesday — Tech debt review with Client A (2 hours). Their lead engineer and I walk through their service map. We've been tracking 7 tech debt items for a quarter and we re-rank them based on what's actually impacting velocity. Decision: rebuild their background job system this quarter, defer the auth refactor to Q4. I write up the decision in their engineering doc so the rest of the team understands why.

Wednesday — Engineering hiring call with Client B (1.5 hours). They're hiring a senior backend engineer. I run a 60-minute system design interview, then debrief with the founder for 30 minutes. Verdict: strong technically, but probably under-leveled — would be a great hire at staff, will get bored at senior. We pass, founder is relieved because she had the same feeling but couldn't articulate why.

Thursday — Architecture review with Client C (3 hours). They're adding a new AI feature that needs vector search, async job processing, and a webhook layer. I sit with their team and we whiteboard three approaches. We pick the boring one — Postgres + pgvector + Solid Queue — because it adds zero new infrastructure. I write up the decision and the alternatives we rejected so we have a paper trail for the next engineer who joins.

Friday — Strategy doc and founder 1:1s (3 hours). Each client gets a 30-45 minute founder 1:1 on Fridays. We review the week, set priorities for next week, and surface anything I'm seeing that the founder isn't. After 1:1s I update the rolling strategy doc for each client — usually a 1-page Notion page with current priorities, blockers, and the next 2-3 decisions coming up.

Total: ~11.5 hours of direct work plus another 3-4 hours of async Slack and document review. That's a normal week. Crisis weeks (production incident, key engineer quits, surprise due diligence) can hit 20+ hours and the retainer absorbs it.

How a fractional CTO compares to alternatives

Option Cost/month Hours/week Owns roadmap? Hires engineers? Accountability
Fractional CTO $8K-$15K 5-15 Yes Yes High — embedded in the team
Full-time CTO $20K-$35K + equity 40-60 Yes Yes Total — it's their full job
Technical advisor $0-$2K (or equity only) 1-3 No Sometimes interviews Low — opinion only
Consulting agency $15K-$50K 20-80 (team) No No Project scope only
Senior engineer + EM $25K-$40K 80 (two people) Partially Yes Execution-focused

The math that surprises most founders: a fractional CTO is usually cheaper than the consulting agency you'd hire to make the same architecture decisions, and dramatically cheaper than the full-time CTO you can't yet afford. The trade-off is hours — you're getting 5-15 hours a week, not 40.

When you do NOT need a fractional CTO

I turn down maybe a third of inbound conversations because the fit isn't there. Here are the scenarios where hiring a fractional CTO is a bad use of money.

  • You don't have any engineers yet. A fractional CTO can help you hire your first engineer, but if you're pre-MVP and pre-team, what you actually need is either a technical co-founder or a development shop to build the first version. Hiring me to "advise on" an empty engineering team is a waste — I'd just spend three months telling you to hire an engineer.
  • You don't have funding or revenue. $8K-$15K/month is real money. If you're pre-revenue and self-funded, that capital should usually go toward shipping the product, not toward strategy. The exception: you're about to raise and need help getting the technical story tight for due diligence.
  • You're a technical founder with time. If you can code and you have 20+ hours a week to spend on engineering, you don't need a fractional CTO yet. You might need a sounding board (which is what a technical advisor at $500-$2K/month is for). Hire me when you stop having time, not before.
  • You have a senior engineer who could grow into the role. Sometimes the best move is promoting your senior engineer into a tech lead position and giving them coaching. A fractional CTO who acts as a coach to an internal tech lead is a legit engagement — but it's a different shape than the embedded model. Decide which you actually want.

Frequently asked questions

How many hours a week does a fractional CTO work?
Typical engagements run 5-15 hours per week per client. Most fractional CTOs work with 2-4 clients concurrently, which means a full week of 30-40 hours total across all engagements. Hours are flexible inside that range — crisis weeks (outages, hiring sprints, fundraising) absorb more time, calm weeks absorb less. The retainer model exists specifically so neither side is counting hours week to week.
Do fractional CTOs write code?
Rarely, and only when it's the highest-leverage thing they can do. A fractional CTO might prototype a proof-of-concept to validate an architecture decision, or write a short script to automate a recurring engineering task. They should not be writing production features — that's $10K/month engineering work at $200K/year engineer rates, which is a bad use of the budget. If you need someone shipping code, hire an engineer.
Can a fractional CTO hire my engineers?
Yes, and it's often the highest-ROI part of the engagement. A fractional CTO writes job descriptions, runs technical screens, sits on architecture interviews, calibrates leveling, and helps you negotiate offers. They don't typically run the full recruiting funnel (sourcing, scheduling, candidate communication) — that's a recruiter or an internal hiring manager job — but they own the technical evaluation end-to-end.
When should I hire a fractional CTO?
Three common triggers: you have 2-5 engineers and nobody owns the technical strategy, you're raising a Series A in the next 6 months and need the technical story tight, or you're a non-technical founder whose technical co-founder just left. If none of those apply and your engineering team is healthy, you probably don't need one yet.
How is a fractional CTO different from a CTO?
Same job, fewer hours, no equity. A full-time CTO works 40-60 hours/week, takes equity, has total accountability for engineering, and usually manages all hiring directly. A fractional CTO works 5-15 hours/week for cash, has high accountability for strategy and key decisions, and typically partners with an internal engineering manager or senior engineer who handles day-to-day execution.
How long do fractional CTO engagements last?
The median engagement is 6-18 months. Engagements end three ways: the company hires a full-time CTO (success), the company raises a big round and outgrows the model (success), or the engagement isn't producing results and one side ends it (the right call). Engagements under 3 months rarely produce real value — the first month is mostly the CTO learning your codebase and team.
Do fractional CTOs take equity?
Sometimes, but cash-only is the norm. Some fractional CTOs accept a small equity grant (0.1%-0.5%) for strategic engagements, especially with pre-revenue startups where cash is tight. Most engagements at companies with funding or revenue are pure cash retainer with no equity. Be skeptical of any fractional CTO who insists on equity-only arrangements — they're either gambling on you or undervaluing their time.
How do fractional CTOs juggle multiple clients?
Calendar discipline and clear hour boundaries per client. Most fractional CTOs assign specific days or time blocks to each client and use a shared engineering Slack channel for async communication. The 2-4 concurrent client cap exists because beyond that, response time degrades and clients start feeling like they're getting leftovers. Ask any fractional CTO you're evaluating how many clients they currently have — if the answer is 6+, run.

If you're evaluating a fractional CTO for your business

The decision usually breaks down into three questions: what does it cost, how do you find a good one, and how is this different from just hiring a full-time CTO. I've written deeper posts on each. Start with the fractional CTO cost breakdown — hourly rate vs retainer to understand the pricing math, then read how to hire a fractional CTO for the evaluation process I'd run if I were on the buying side.

If you're weighing the full-time vs fractional decision, the fractional vs full-time CTO breakdown covers when each makes sense — and if the bigger gap is actually product rather than engineering leadership, the fractional product manager post explains the parallel role.

I run a small embedded fractional CTO practice — 3-4 companies at a time, $8K-$12K/month, typically 12-month engagements. If that sounds like the right shape, you can read more about Justin's fractional CTO services or book a strategy call and we'll figure out in 30 minutes whether it's a fit.

Frequently Asked Questions

How many hours a week does a fractional CTO work?
Typical engagements run 5-15 hours per week per client. Most fractional CTOs work with 2-4 clients concurrently, which means a full week of 30-40 hours total across all engagements. Hours are flexible inside that range — crisis weeks (outages, hiring sprints, fundraising) absorb more time, calm weeks absorb less. The retainer model exists specifically so neither side is counting hours week to week.
Do fractional CTOs write code?
Rarely, and only when it's the highest-leverage thing they can do. A fractional CTO might prototype a proof-of-concept to validate an architecture decision, or write a short script to automate a recurring engineering task. They should not be writing production features — that's $10K/month engineering work at $200K/year engineer rates, which is a bad use of the budget. If you need someone shipping code, hire an engineer.
Can a fractional CTO hire my engineers?
Yes, and it's often the highest-ROI part of the engagement. A fractional CTO writes job descriptions, runs technical screens, sits on architecture interviews, calibrates leveling, and helps you negotiate offers. They don't typically run the full recruiting funnel (sourcing, scheduling, candidate communication) — that's a recruiter or an internal hiring manager job — but they own the technical evaluation end-to-end.
When should I hire a fractional CTO?
Three common triggers: you have 2-5 engineers and nobody owns the technical strategy, you're raising a Series A in the next 6 months and need the technical story tight, or you're a non-technical founder whose technical co-founder just left. If none of those apply and your engineering team is healthy, you probably don't need one yet.
How is a fractional CTO different from a CTO?
Same job, fewer hours, no equity. A full-time CTO works 40-60 hours/week, takes equity, has total accountability for engineering, and usually manages all hiring directly. A fractional CTO works 5-15 hours/week for cash, has high accountability for strategy and key decisions, and typically partners with an internal engineering manager or senior engineer who handles day-to-day execution.
How long do fractional CTO engagements last?
The median engagement is 6-18 months. Engagements end three ways: the company hires a full-time CTO (success), the company raises a big round and outgrows the model (success), or the engagement isn't producing results and one side ends it (the right call). Engagements under 3 months rarely produce real value — the first month is mostly the CTO learning your codebase and team.
Do fractional CTOs take equity?
Sometimes, but cash-only is the norm. Some fractional CTOs accept a small equity grant (0.1%-0.5%) for strategic engagements, especially with pre-revenue startups where cash is tight. Most engagements at companies with funding or revenue are pure cash retainer with no equity. Be skeptical of any fractional CTO who insists on equity-only arrangements — they're either gambling on you or undervaluing their time.
How do fractional CTOs juggle multiple clients?
Calendar discipline and clear hour boundaries per client. Most fractional CTOs assign specific days or time blocks to each client and use a shared engineering Slack channel for async communication. The 2-4 concurrent client cap exists because beyond that, response time degrades and clients start feeling like they're getting leftovers. Ask any fractional CTO you're evaluating how many clients they currently have — if the answer is 6+, run.

If this was useful, here are two ways I can help: