CTO Handbook
How I build and scale exceptional technical teams in the age of GenAI.
By Tim Vasil • Last updated
Legend
Do Don’t Figure Table Interactive tool
I’ve written this guide as a model I can use as a CTO or consultant, codifying hard-learned best practices that, admittedly, I continue to refine! I hope this guide also provides value to my peer CTOs out there.
This handbook is based on my experience over 20 years scaling technical teams from 0 to 150 people in VC- and PE-backed SaaS companies working towards an exit. I’ve maintained largely the same operating principles over the years, which reflect my core values of how to treat people and maintain healthy teams.
Operating principles
- People first. Be flexible wherever possible to meet people’s needs. Create meaningful opportunities to build relationships and work together in person. Treat everyone equitably with empathy, respect, and genuine care.§ 1.7§ 2.4§ 2.6
- Customer-proximate. Understand the customers’ and purchasers’ needs and motivations, primarily through firsthand experience. Continuously test whether the problems we are solving are the ones that matter most.§ 2.1§ 3.3§ 8.2
- Tight feedback loops. Unify engineering, product, design, and data science disciplines as one team, regardless of departments or reporting lines. Delegate meaningful ownership. Use process to enable good decisions, not substitute for them. Optimize for learning speed to “fail fast.”§ 1.1§ 3.5§ 5.5
- High standards. Hold a high bar for outcomes, quality, and integrity. As recommended by Kim Scott, give radically candid, timely feedback; address problems directly while caring personally. Build quality in through thoughtful design, automated testing, code review, and observability.§ 2.6§ 3.4§ 3.6
- Multiple hats, not pigeon-holes. Allow any combination of individual contributor, tech lead, and manager roles as complementary paths. Each of these has opportunities to increase influence, scope, and compensation. Don’t insist on functional silos either; for example, embrace a PM who also wants to code.§ 1.3§ 2.4§ 2.5
- Best tools for the job, including GenAI. Do not skimp on hardware, IDEs, or other tools that enable productivity. Use AI wherever it can eliminate mechanical work or improve human judgment across the entire PDLC. With GenAI specifically and DevEx generally, ensure the team can focus primarily on the work that requires judgment, creativity, taste, and empathy. Embed GenAI tools seamlessly where people already work.§ 3.1§ 3.4§ 3.7§ 3.8
- Collaborative & transparent. Work together without regard for titles; the best decisions come from multiple points of view coming together. Make decisions, context, and information accessible to the people who need them. Share both good and bad news openly and authentically.§ 1.7§ 4.3§ 5.5§ 8.5
- Leadership by example. Act the way you want others to act by showing care and effort. Reject “not my job” thinking. Take initiative to pitch in and help resolve obstacles that arise, not ignore them. Help someone stuck on a setup step. Offer advice. Find things to praise.§ 2.1§ 2.3§ 5.2§ 7.2
- Continuous improvement. Treat the organization and its operating model as products: measure them, experiment, learn, and change for the better. Solicit ideas for improvement from everyone.§ 3.1§ 4.1§ 8.6