What a Real External CTO Actually Does in Your First 90 Days

The 90-day filter

If your fractional CTO has not delivered tangible value by day 90, they probably will not at day 180. The good ones run a specific playbook in the first three months that produces visible artifacts: a roadmap, a hiring plan, an architecture document, and a clear handoff path if you ever need it. Below is what that playbook looks like in 2026.

Days 1–14: Audit

The first two weeks are about understanding what you have. A good external CTO will:

• Read every line of your codebase or, if it is large, the parts that handle data, auth, and money.
• Sit in on two to three customer calls or sales conversations to understand the product from the user side.
• Interview every current engineer or contractor for 30 minutes each.
• Run through your deploy process end-to-end as if they were a new engineer.
• Read every commercial agreement that has technical clauses (vendor contracts, data processing agreements, SLAs).

By end of week two they should produce a 3 to 5 page document: what works, what is fragile, what is risky, and what they recommend addressing in what order. If you do not have this document by day 14, the engagement is off-track.

Days 15–45: Stabilize and plan

With the audit done, the focus shifts to fixing the most painful immediate problems and writing the longer-term plan. Typical work in this window:

• One or two production-fragility fixes (the things that wake you up at 2am).
• A 90-day technical roadmap aligned to the product roadmap.
• A hiring plan for any roles needed in the next two quarters, with role definitions and rubrics.
• A vendor and infrastructure review — what you are paying for, what you actually need, what to consolidate.
• A first-cut technical risk register for the company.

This is where you should start to feel like the engineering function is being run by someone who has done it before.

Days 46–90: Execute and prove out

The last month of the first quarter is about executing one or two specific projects to a clean conclusion. These are usually:

• A meaningful feature shipped end-to-end on a senior-led timeline.
• The first engineering hire onboarded under the new rubric.
• A measurable improvement in delivery cadence (e.g., PRs merged per week up by 30 percent without an increase in incidents).
• A documented incident response plan that has been walked through with the team.

By day 90 you should have artifacts: documents, dashboards, hires, shipped features. If the only thing you have is "I had good conversations with the CTO," the engagement is not working.

What it should cost

A fractional CTO running this playbook at a pre-seed startup typically charges $6k–$15k per month for 10 to 20 hours per week. Beneath that range you are likely getting someone who is using your company as a side project. Above that range you are likely paying for a recognized name where most of your bill is brand value, not delivered work.

The high-leverage middle is a senior with 10+ years of relevant experience, two to three concurrent clients, and a clear track record of running this 90-day cycle before.

The honest signal

The best external CTOs are the ones who tell you, by day 60, exactly what they would do if they were full time and exactly what they will not be able to do because they are not. They surface the gap rather than hide it. The ones to avoid are the ones who tell you everything is fine and the engagement should just continue indefinitely. Continuation is a result, not a strategy.