Building Great Product Teams: Principles & Rituals
Lane Shackleton
Head of Product, Coda
DEC 2024
Core Thesis
Turn Ambiguity Into Clarity
"The core job of a product person is to turn ambiguity into clarity. Everything in product is ambiguous—roles, problems, customers, solutions."
Product ambiguity is constant: What's my role? What problem? Who's the target?
Systems beat goals: Jerry Seinfeld threw out 15 years of material and rebuilt through daily writing practice
Build recurring rhythms for customer learning, not quarterly OKR-driven bursts
Default-on systems create good instincts over time
Framework
Systems Not Goals: Why Rhythms Matter
The Problem: Goals with good intentions don't stick. Teams chase customer research OKRs, then abandon them next quarter.
The Solution: Default-on systems. Calendar slot every Friday. Customer visit every week, no exceptions.
The Paradox: The goal (talking to 10 customers) matters less than the system that achieves it
Early Coda Example: Friday customer slot was mandatory. Team scrambled 3 hours before if needed. Result: exceptional customer instincts built through repetition, not planning
The Seinfeld ModelGoal = 1 hour of new material. System = write 1 hour every morning, perform every night. After 300+ repetitions, the goal happened naturally. Product teams should work the same way.
System vs. Goal
Goal-driven: Talk to 10 customers this quarter. Maybe hit it. Maybe not. Next quarter? Goal changes. Learning is inconsistent.
System-driven:
Every Friday: Customer on calendar. No escaping. 200+ customer conversations a year. Instincts become world-class.
Rituals & Architecture
Coda's Operating System: Three Ceremonies
1. Catalyst (Decision Reviews)
Three 1-hour blocks per week. Whole company assumes they're available
Four roles per topic: Driver, Maker, Braintrust, Interested
Multi-threaded: 3+ topics run simultaneously (no standing attendees)
Right people, every time. Faster throughput. More autonomy for PMs
Before: Single-threaded reviews. 3-hour meetings once a week. Impossible to get on calendar. Slow velocity.
2. Tag-Ups (Synchronous Huddles)
Quick decision-making on smaller topics
Stakeholders signal: "Keep going," "I have opinions, proceed," or "Heavily involved"
Nathan's table system: All upcoming decisions visible, people react on bandwidth
Result: PMs know exactly when they need wider input vs. can decide independently
3. Bullpen (Continuous Feedback)
Always-on feedback channel separate from formal reviews
Reduces bottlenecks and prevents surprises in Catalyst
Career Clarity
Five Role Stages (Not Titles)
Apprentice Learns about rope
Practitioner Ties basic knots. Shown complex knots
Capable Can solve problems independently
Expert Knows all knots + rope strength
Principal Invented nylon. Bar: Aspirational high
Coda's approachSame 5 levels across all roles (design, PM, eng). Roles don't appear publicly. Comp committee sets salary (not managers). Decentralized, transparent, orientation toward company > title chasing.
Counterintuitive
Product Team Myths Lane Rejects
✗OKRs drive behavior you wantINSTEAD →✓ Systems drive behavior. Quarterly goals vanish. Weekly practices compound into excellence.
✗Career ladders need 10+ levels to be usefulINSTEAD →✓ Five clear role stages beat 15-level ladders. Transparency isn't always good. Hide levels to keep people company-focused.
✗Review meetings need standing attendeesINSTEAD →✓ Dynamic attendees (Catalyst model) = right people every time + 3x throughput. Multi-threading changes everything.
✗Growing means constantly redesigning processesINSTEAD →✓ Build a stable backbone (Catalyst, tag-ups, bullpen). Iterate on top. Avoid process whiplash. Stability enables velocity.