For first-time CTOs at Series A–B
Your First 90 Days as CTO
A step-by-step plan: listen first, diagnose second, commit third. By Day 90 you have trust, a real diagnosis, and a strategy your engineers can recite from memory.
You've taken the seat. You're now The CTO. The first 90 days will set the trajectory for the next two years.
This plan is built for the situation most first-time CTOs land in: Series A or B startup, 20–80 engineers, you were either promoted from inside (VPE → CTO) or hired in from outside. You inherit a codebase you didn't build, a team you didn't hire, and expectations you didn't set.
The instinct is to act. To prove value fast. Don't.The biggest mistake first-time CTOs make is shipping changes before they understand what's actually broken.
The framework: 30 / 60 / 90
| Phase | Days | Goal | Deliverable |
|---|---|---|---|
| Listen | 1–30 | Build trust. Understand the system. | Current-state memo |
| Diagnose | 31–60 | Separate real problems from surface ones. | Diagnosis memo |
| Commit | 61–90 | Publish strategy. Ship changes that compound. | 1-pager + Q1 plan |
Days 1–30: Listen
Goal: Build trust with engineers and stakeholders. Understand the system as it actually works, not as the architecture diagram claims.
Week 1: Meet everyone close in.
- 1:1 with every direct report. Two hours each, not thirty minutes.
- 1:1 with the CEO. Ask: “What does success look like at Day 90? At month 12? What would make you regret hiring me?”
- 1:1 with founders if separate. Listen for unspoken priorities.
- 1:1 with the CFO. Understand the engineering budget, runway, and the cost story.
Weeks 2–4: Meet everyone else.
- Skip-level 1:1s with at least 50% of engineering (everyone if <30 people).
- 1:1 with peer C-level: COO, CPO, CRO. Each one believes engineering is failing them in a different way. Find out how.
- 1:1 with three customers.Critical: do this in week 2, not week 12. Customers know what's broken better than your team admits.
Throughout: Read everything.
- Last six months of postmortems.
- The runbook (if it exists).
- The previous CTO's strategy doc (if it exists) — don't dismiss it even if you disagree.
- The latest board deck.
- The on-call rotation history.
Trap to avoid
Don't reorganize.The pressure to “make a change” peaks at week 3 and the temptation is to restructure teams. Resist. You don't know enough yet. Reorgs done before Day 60 are almost always wrong.
Deliverable
A two-page current-state memo. Org chart, system map, top five risks as observed, top three strengths. This is for you.Don't share it.
The 90-Day Plan — printable + companion exercises
Get the printable checklist plus thirteen weekly companion exercises (interview questions, audit templates, retro frameworks) delivered to your inbox.
Days 31–60: Diagnose
Goal: Surface the real problems behind the visible symptoms. Make small, visible improvements that prove momentum without committing to direction.
Week 5: Audit the engineering loop.
Get the four DORA metrics for the last 90 days:
- Lead time — PR open to production.
- Change failure rate — deploys requiring rollback or hotfix.
- MTTR — time to restore service after an incident.
- Deploy frequency — how often you ship.
If you don't have these numbers, your first quick win is measuring them.
Weeks 6–7: Run a tech-debt audit.
- Ask every senior engineer: “If you could fix one thing about the codebase, what would it be?” Tally the answers.
- Pick the cheapest two from the top five. Fix them. This is your “I see you” signal to engineers.
Week 8: Diagnose the team shape.
- Are teams aligned to product or aligned to tech layers? (Team Topologies.)
- Where are the dependency bottlenecks? Who blocks whom?
- Is on-call sustainable? Who's burning out?
Trap to avoid
Don't confuse symptoms with problems.“Engineering is slow” is a symptom. The problem might be too many cross-team handoffs, unclear priorities, no testing culture, missing platform investment, or all four. Diagnose before prescribing.
Deliverable
A four-page diagnosis memo. Real problems (not symptoms), with evidence. Three to five options for each, with trade-offs. Share this with your CEO and one trusted peer. Get pushback. Refine.
Days 61–90: Commit
Goal: Publish a clear engineering strategy. Restructure where you must. Ship two or three changes that compound.
Week 9: Write the 1-pager.
A single-page engineering strategy. Headlines only:
- What we're building.
- How we work (cadence, decisions, ceremonies).
- What we're measuring (three to five metrics, no more).
- What we're not doing.
- The first quarter's three bets.
Week 10: Restructure (if needed).
Only now do you have the data and trust to reorganize. If teams are wrong-shaped, fix them. If a layer of management is missing, add it. If you have a manager-as-IC problem, name it.
Weeks 11–12: Ship the first three bets.
Pick three things from the diagnosis memo. Not five. Not eight. Three. Make them visible — track them on a wall, in Slack, in standup. Demonstrate that engineering can ship coordinated work.
Week 13: Retro.
Run a retrospective on the first 90 days. What worked? What didn't? Share with your CEO. Publish the strategy 1-pager company-wide.
Trap to avoid
Don't publish a strategy you can't walk through in two minutes.If the strategy needs 15 slides to explain, it's not a strategy — it's a wish list. Strategy is what you're choosing not to do. Make the choices explicit.
Deliverable
The 1-pager. The first-quarter plan. A scheduled monthly review with the CEO. Three live initiatives with named owners.
What Day 90 looks like
You should have
- Trust with your team and the C-suite.
- A clear view of real (not surface) problems.
- A 1-page strategy engineers can recite.
- Three concrete bets in flight.
- A measurement system that makes progress visible.
You should not have
- A 40-slide strategy deck.
- A reorg you couldn't justify with data.
- Burned-out engineers from trying to do too much.
- A list of grievances about “what I inherited.”
The first 90 days set the trajectory. The next 270 are about compounding.
Get the printable plan + 13 companion exercises
Each week for 13 weeks, one practical companion exercise — interview questions, audit templates, retro frameworks — to apply that week's work.
Next steps
- ◆ CTO Rating Framework — score yourself across the dimensions that matter in the first 90 days.
- ◆ Engineering Metrics Dashboard — capture the four DORA metrics for week 5's engineering-loop audit.
- ◆ Tech Debt Prioritizer — turn week 6's tally into ranked, actionable bets.
- ◆ Team Topology Diagnostic — diagnose team shape and dependency bottlenecks in week 8.
- ◆ Frameworks library — Team Topologies, DORA, Wardley mapping, and more.