How to Foster Innovation
and Big Thinking

Eeke de Milliano

Head of Product, Retool
Early PM, Stripe (Connect · Radar)

FEB 2 2023

Core Concept

Process Is Variance Reduction

"While you bring folks up to the average, you also bring other folks down. And those are usually your highest performers."

Process reduces variance in both directions, bad and exceptional
Top performers don't need process, they need escape hatches
Minimum Viable Process: set a floor, not a ceiling
Write at the top of every template: "break it if this doesn't fit"

Framework

The Innovation Operating System

70%

core product

20%

strategic bets

10%

crazy ideas

Crazy Ideas doc (Stripe & Retool)

Every January, a blank doc goes to the whole org: "90% chance these make no sense, but the 10% chance could be 10–100x for the business." Result: 3–8 crazy ideas actually ship each year. Retool Workflows started on this list.

Think Bigger in every team charter

"With 20% more time, what would you do that isn't on this list?" and "If you doubled the team today, what would you build?", baked into every planning cycle to break people out of planning to the team they have.

Stripe's 3 operating principles

Think rigorously, no best practices without asking why. Move with urgency and focus, time is your biggest enemy. Micro pessimist, macro optimist, critical on the day-to-day, bullish on the long arc.

Playbook

How Retool Launched 3 Products in One Year

Start tiny: 1–2 people per bet for the first 6 months, no funding until there's real signal
Treat them like startups: Retool is the VC; teams must prove ROI in engagement or revenue
Keep them separate: Isolate new bets from core product org to protect speed and independent thinking
Sequence your launches: In hindsight, staggered ship dates reduce org-wide cognitive load
Build the scooter, not the axle: Don't ship half a car, ship a complete, smaller thing first, then bicycle, then motorcycle, then car

"We really treated them like startups. Retool is the VC, funding with resources and our existing customer base, but the team had to prove the ROI."

The 3-doc minimum viable process

Every team at every stage needs: Charter (mission + vision + strategy), Goals (what success looks like and how it's measured), Roadmap (what ships). Build top-down, don't start from a list of features.

Time horizon shifts with maturity

Pre-PMF: charter horizon = 3 months. Growing: 1–2 years. Mature, humming company: a decade. Don't look further out than you can actually see, it produces false precision.

Build for your best user

Most teams design onboarding for the confused user. Design instead for the one who instantly gets it. Worst-case users are a fraction, don't let them shape the product for everyone else.

Tactics

The Talent Portfolio Model

Don't hire PMs in your own image, hire for your team's gaps, not your own strengths
Balance homegrown PMs (culture, product depth) with external PMs (craft, outside perspective)
Balance execution machines with visionaries, you need both in every product pillar
Every 6 months: chart your team's strengths and weaknesses, then hire specifically for the weaknesses

Customer closeness at Retool Hundreds of dedicated Slack channels with customers. PMs from sales and CS backgrounds. Everyone has a CS background. And they use Retool to build Retool, their own roadmap and feature flags run inside Retool apps.

On high performers + process "Are you willing to break the org for this person?", Claire, COO of Stripe. Some people are that good. Give them air cover to bypass the process. Managers are the unlock, they detect the talent, then protect it.

Contrarian

Innovation Myths, Busted

✗ Process helps your best people perform better INSTEAD → ✓ Process drags top performers toward the average. They don't need it. The fix: Minimum Viable Process, set a floor, give escape hatches, and let greatness exceed the template.

✗ Big bets need big teams from day one INSTEAD → ✓ Retool launched 3 products in a year with 1–2 people each. No funding until signal is real. Small bets move faster, think independently, and discover the right customer surprisingly often.

✗ Idea docs and hackathons are feel-good theater INSTEAD → ✓ Retool's Crazy Ideas doc ships 3–8 real products per year. Retool Workflows, a major product line, started as a "crazy idea." Structure the permission; results follow on their own.

✗ Hire PMs like yourself, you know what good looks like INSTEAD → ✓ The best PM teams are complementary portfolios, homegrown + external, visionary + executor. Map your gaps every 6 months and hire precisely for the weaknesses you find.