The first thirty days with a fractional CTO

Founders rightly fear consulting theatre: workshops that sparkle, slideware that rusts.

The honest shape of early fractional CTO work is smaller and more tactile: map risk, tighten decision paths, publish a few artefacts everyone can cite, then repeat until the team recognises the rhythm.

This is illustrative — engagements flex with runway, timezone spread, audit deadlines, AI adoption messiness, regulatory constraints. Treat it as a pattern language, not a contract template.

Week 1 — Read the truth, quietly

Goals:

System map that matches lived reality, not aspiration diagrams.
Surfacing decisions already made implicitly (timeouts, tenancy assumptions, observability posture, retention).

Typical artefacts:

Risk register — grouped by severity and reversibility (“cheap to revisit” vs “structural”).
Architecture snapshot — bounded to what affects the next ninety days.

Anti-patterns to avoid immediately:

Big-bang rewriting before constraints are enumerated.
“Best practices” detours unrelated to imminent bets.

Cadence founders feel: structured interviews with tech lead + PM + whoever holds customer truth; short readouts (written, linkable).

Week 2 — Decision hygiene and communication spine

Goals:

Single prioritisation backbone founders and engineering can cite.
Obvious escalation rules — who decides when infra and product collide.

Typical artefacts:

Lightweight decision record format (ADR-lite is enough if you hate ceremony).
Delivery cadence sketch — rituals that fit your timezone spread, not Spotify’s playbook.

Fractional CTO leverage here is facilitation with teeth: insisting that ambiguous ownership counts as structural debt.

Week 3 — Hiring, vendor, and toolchain alignment

Goals:

If you plan hires, fractional CTO involvement sharpens loops and rejects misfits early.
If AI-assisted development is fraying consistency, converge on minimal enforceable norms review can apply without heroics.

This is frequently where sibling work emerges: align the team materially (engineering workshop), then deepen architectural strategy without debating basics ad nauseam forever.

Avoid:

Tool religion debates — pick defaults with reversibility timelines.

Week 4 — Stabilisation and ninety-day storyline

Goals:

A forward-looking ninety-day sketch — sequencing, exits, deliberate tech debt compartments.
Durable artefacts that replace oral tradition for onboarding new engineers.

Healthy exit criteria for thirty days sound boring:

Risks articulated and owned.
A weekly ritual exists that survives one vacation without hallucinated priorities.
At least one previously implicit architectural choice is documented enough that a newcomer would not reinvent it blindly.

Related: