Mental models for building products people love ft. Stewart Butterfield
Executive Summary
Stewart Butterfield joins Lenny Rachitsky to discuss utility curves, product taste, attention to users and the difference between comprehension and friction. This selected coverage includes the opening Parkinson's law discussion but omits politics/bureaucracy, shutdown/Glitch, later work-like activities and the owner's delusion. Slack anecdotes are his account, unverified here. Untimed topics are editorial, not video chapters.
Key Takeaways
Think about the utility curve of effort
Butterfield uses an S-curve to ask where more work adds value. Early effort may not make a feature useful, a threshold can create a large gain, and further refinement may reach diminishing returns. This is a mental model for judgment, not a measured formula.
A feature is more than present or absent
He warns that shipping a weak version adds complexity without enough usefulness, while polishing past the useful point can waste effort. Ask whether the feature needs more investment or already delivers most of its value. Its existence on a checklist does not answer that question.
The quality bar moves
He invokes the idea of divine discontent, crediting Bezos, to describe rising expectations as software improves around users. Revisit essential flows that were built once and rarely improved. His examples are his perspective, not a current audit of another company's product.
Practice product taste
Butterfield believes taste can be developed through practice even if some people find it easier. He links careful attention to the experience with product advantage. That is his argument, not evidence that taste alone causes a company to succeed.
Notice the impact on the person using it
His tilt-your-umbrella story asks whether you notice when a small action inconveniences someone else. He applies that attention to product details that make people's work easier. Slack's growth and recommendation stories are his account, unverified here.
Separate comprehension from friction
Someone with a clear, strong intent may persist through a difficult flow, while a new user may not understand what a product is or what to do next. Removing clicks cannot solve that uncertainty by itself. Explain the purpose and next step before assuming the shortest route is best.
Time available can shape how long work takes
He discusses Parkinson's law as the observation that work expands to fill the time available. A deadline can help focus a bounded task, but shortening time is not a reason to skip needed checks. Use the idea to examine open-ended work, not as a universal productivity law.
The Conversation Breakdown
Think about the utility curve of effort
⏱ Untimed topicButterfield uses an S-curve to ask where more work adds value. Early effort may not make a feature useful, a threshold can create a large gain, and further refinement may reach diminishing returns. This is a mental model for judgment, not a measured formula.
A feature is more than present or absent
⏱ Untimed topicHe warns that shipping a weak version adds complexity without enough usefulness, while polishing past the useful point can waste effort. Ask whether the feature needs more investment or already delivers most of its value. Its existence on a checklist does not answer that question.
The quality bar moves
⏱ Untimed topicHe invokes the idea of divine discontent, crediting Bezos, to describe rising expectations as software improves around users. Revisit essential flows that were built once and rarely improved. His examples are his perspective, not a current audit of another company's product.
Practice product taste
⏱ Untimed topicButterfield believes taste can be developed through practice even if some people find it easier. He links careful attention to the experience with product advantage. That is his argument, not evidence that taste alone causes a company to succeed.
Notice the impact on the person using it
⏱ Untimed topicHis tilt-your-umbrella story asks whether you notice when a small action inconveniences someone else. He applies that attention to product details that make people's work easier. Slack's growth and recommendation stories are his account, unverified here.
Separate comprehension from friction
⏱ Untimed topicSomeone with a clear, strong intent may persist through a difficult flow, while a new user may not understand what a product is or what to do next. Removing clicks cannot solve that uncertainty by itself. Explain the purpose and next step before assuming the shortest route is best.
Time available can shape how long work takes
⏱ Untimed topicHe discusses Parkinson's law as the observation that work expands to fill the time available. A deadline can help focus a bounded task, but shortening time is not a reason to skip needed checks. Use the idea to examine open-ended work, not as a universal productivity law.
Core Frameworks & Ideas
Utility curve
Judge whether additional effort crosses a usefulness threshold or has reached diminishing returns.
Comprehension and intent
Distinguish a user who knows what they want from one who first needs to understand the product and next step.
Parkinson's law
Examine how the time allowed shapes a task, without sacrificing necessary quality checks.
Actionable Next Steps
- ✓Review one feature for insufficient investment or diminishing returns.
- ✓Watch a new user and ask what they think the product does.
- ✓Inspect one neglected essential flow for a small unnecessary inconvenience.
- ✓Set a realistic boundary for one open-ended task and keep the checks it needs.
Notable Quotes
“But you can definitely practice and you can definitely get better.”
“And then there's some magic threshold where it produces an enormous amount of value and then continued investment doesn't really pay off.”
“And so the problem there is comprehension, and this will come up an enormous amount.”
The One Thing to Remember
Ask whether a product detail helps users understand and accomplish their goal, rather than counting features or clicks.