How to foster innovation and big thinking | Eeke de Milliano (Retool, Stripe)
Executive Summary
Eeke de Milliano joins Lenny Rachitsky on learning from failure, small internal bets, deliberate big thinking, minimum viable process and staying close to customers. Selected coverage combines the innovation and process discussions, omitting detailed Stripe history, hiring and the AI comparison. Company examples are her interview-era account, unverified here. Untimed editorial topics are not video chapters.
Key Takeaways
Account for the cost of innovation
De Milliano says leaders can want the upside of innovation without accepting that larger swings sometimes fail. Normalize learning from failure, but distinguish it from repeated failure without learning. Her account of Retool and Stripe is historical experience, not independently verified current policy.
Lower the stakes of an early bet
Small teams and early customer feedback make a new idea less costly to test. Waiting a year to ship without feedback raises the perceived stakes. She recommends more chances to learn rather than putting too many resources behind an unproven swing.
Give people permission to think bigger
Urgent incidents and product debt can crowd out creative work. Diagnose that burden and create deliberate moments to consider larger ideas. Her examples include a Think Bigger prompt in team planning and a separate request for unlikely ideas with potentially large upside.
Treat internal bets like startups
She recounts starting initiatives with one or two people and expanding when there was evidence of something useful. Teams needed to demonstrate engagement or eventual revenue return. Keeping early bets separate helped speed, but created later integration work, so separation has costs too.
Inspect what process does to the best work
De Milliano describes process as reducing variance: it helps meet a standard but can pull high performers and unusual ideas toward the average. Before adding a rule or template, ask what quality it protects and what creativity it may suppress. Her criticism is not a claim that all process is harmful.
Use minimum viable process with escape hatches
She offers a template as a starting bar, while explicitly allowing people to depart from it when it does not explain their work. This preserves a shared minimum without treating one format as the only acceptable answer. The selected summary omits the AI comparison in this passage.
Keep product work close to customers and use
De Milliano describes Retool's shared customer Slack channels and PMs using Retool apps for their own roadmap and feature flags. Those are her interview-era examples of staying close to customer questions and daily product use, not a required tool choice for other teams. Selected coverage also omits the detailed Stripe history and hiring discussion.
The Conversation Breakdown
Account for the cost of innovation
⏱ Untimed topicDe Milliano says leaders can want the upside of innovation without accepting that larger swings sometimes fail. Normalize learning from failure, but distinguish it from repeated failure without learning. Her account of Retool and Stripe is historical experience, not independently verified current policy.
Lower the stakes of an early bet
⏱ Untimed topicSmall teams and early customer feedback make a new idea less costly to test. Waiting a year to ship without feedback raises the perceived stakes. She recommends more chances to learn rather than putting too many resources behind an unproven swing.
Give people permission to think bigger
⏱ Untimed topicUrgent incidents and product debt can crowd out creative work. Diagnose that burden and create deliberate moments to consider larger ideas. Her examples include a Think Bigger prompt in team planning and a separate request for unlikely ideas with potentially large upside.
Treat internal bets like startups
⏱ Untimed topicShe recounts starting initiatives with one or two people and expanding when there was evidence of something useful. Teams needed to demonstrate engagement or eventual revenue return. Keeping early bets separate helped speed, but created later integration work, so separation has costs too.
Inspect what process does to the best work
⏱ Untimed topicDe Milliano describes process as reducing variance: it helps meet a standard but can pull high performers and unusual ideas toward the average. Before adding a rule or template, ask what quality it protects and what creativity it may suppress. Her criticism is not a claim that all process is harmful.
Use minimum viable process with escape hatches
⏱ Untimed topicShe offers a template as a starting bar, while explicitly allowing people to depart from it when it does not explain their work. This preserves a shared minimum without treating one format as the only acceptable answer. The selected summary omits the AI comparison in this passage.
Keep product work close to customers and use
⏱ Untimed topicDe Milliano describes Retool's shared customer Slack channels and PMs using Retool apps for their own roadmap and feature flags. Those are her interview-era examples of staying close to customer questions and daily product use, not a required tool choice for other teams. Selected coverage also omits the detailed Stripe history and hiring discussion.
Core Frameworks & Ideas
Lower the stakes
Use small teams, early feedback and retrospectives to learn without a large unvalidated commitment.
Internal startups
Expand new bets when evidence of engagement or revenue potential justifies more resources.
Minimum viable process
Set a shared starting bar and explicitly allow a template to be replaced when it does not fit.
Actionable Next Steps
- ✓Review whether urgent work prevents the team from considering larger ideas.
- ✓Run an early bet with a small team and quick customer feedback.
- ✓Use a retrospective to learn from a failed experiment.
- ✓Add an explicit escape hatch to a template and check its cost to your best work.
Notable Quotes
“I think leaders want the upside of innovation, but they're not really willing to deal with the cost of innovation.”
“Process, by definition, is variance reducing.”
“So I've started referring to this as the MVP, the minimum viable process.”
The One Thing to Remember
Protect room for big thinking by lowering the stakes of early bets and keeping process to a useful minimum.