Slack founder Stewart Butterfield talks about a key trap for founders: they overestimate how much users care (he calls it 'owner's delusion'). Instead of just selling the product, he says, help users realize they have a problem they didn't know they had—like poor team communication. Key mentions: Slack (challenging but behavior-changing), Flickr (open API led to 2,000 third-party apps), and Uber/Lyft (same function as taxis, but experience so good people never go back).
This episode of Invest Like the Best invites Slack founder Stewart Butterfield to discuss his 2014 article, "We Don’t Sell Saddles Here." The core themes include founder's delusion, defining the boundary between product and market, and how innovative products reshape users' mental models and behavio
Stewart Butterfield (Slack founder and CEO) engages in a conversation with Patrick O'Shaughnessy around his 2014 article "We Don’t Sell Saddles Here," focusing on founder's delusion, the boundary between product and market, and how innovative products reshape user mental models and behaviors. Butterfield argues that the essence of innovation is not the product itself, but the change in human behavior—any truly meaningful innovation inevitably leads to a fundamental shift in how people act, and the ultimate measure of innovation is whether it changes human behavior.
Butterfield introduces the concept of "Owner's Delusion," referring to the natural tendency of founders/product owners to overestimate the importance of their product in others' eyes. He uses the example of restaurant websites—where users only want to find the address, phone number, and menu, yet are forced to endure slow-loading food photos and background music—to illustrate this cognitive bias.
Core mechanism: Because founders invest significant time and emotion, they unconsciously assume that users approach the product with the same enthusiasm and attention. In reality, users arrive at the website "with only a threshold of intent that is just barely sufficient," their minds still occupied with unrelated matters such as arguments with their boss, health concerns, family conflicts, and more.
Butterfield emphasizes that Slack's core challenge is: "Slack is something you don't know you want, but once you have it, you can't live without it." The problem is that people simply do not know they want it, and explaining why something they have no concept of is worth trying is extremely difficult.
Key data: When users visit Slack's website, their level of intent is far lower than that of users searching for "Justin Bieber concert tickets"—the latter will persist through a poor purchasing experience to complete the transaction, while Slack users will leave at the slightest confusion.
Butterfield proposes a framework of "marketing from both ends," using a yin-yang diagram as a metaphor: one side is the market, the other is the product, and what truly matters is not either end, but the boundary between the product and the market—that is, how the product captures human attention and creates utility.
Core argument: Most founders focus excessively on the product itself (because building software is fun), while marketers reduce marketing to promotion or advertising. However, the real challenge lies in changing users' mental models and behaviors—this requires leveraging people's existing conceptual frameworks, as no one has the time or energy to learn a large amount of new knowledge to understand your product.
Slack's most successful market information strategy: Instead of selling Slack's features, it helped users recognize that they had a "problem not perceived as a problem"—namely, the issue of information asymmetry and coordination inefficiency within organizations. Butterfield notes that leaders often talk about "alignment," ensuring teams share a common understanding of goals, status, and roles, but few realize that the tools themselves are hindering this alignment.
Specific argument: In Slack, a new employee on their first day can access the organization's historical record of tens of millions of messages (via channels), whereas in an email system, a new employee only has an empty inbox with no knowledge of the organization's history. This difference results in a "three to four orders of magnitude gap in information access," leading to a fundamental change in behavior.
Butterfield recalls the "5K Contest" he launched in 1999—creating the best webpage using no more than 5KB of data. The contest unexpectedly became a global phenomenon, covered by media outlets such as Playboy, a Brazilian newspaper, and Russia's Pravda.
Core Insight: There is an ancient and profound connection between constraints and creativity. Butterfield draws analogies from music (modes, time signatures), poetry (rhyme, meter), and visual arts (20th-century modernism's breakthroughs through constraints) to argue that constraints are not the enemy of creativity but its catalyst.
Business Application: Butterfield introduces the concept of the "utility curve"—a two-dimensional coordinate system where the horizontal axis represents effort invested and the vertical axis represents value generated. The curve takes an S-shape: flat at the start, steeply rising in the middle, and flattening again at the end. The key challenge lies in identifying where the "steep section" is, and constraints help teams filter out worthwhile directions from an infinite set of possibilities.
Specific Case: Uber/Lyft and taxis are "equivalent" in terms of functional requirements (specify a location, wait for arrival, complete payment), yet the difference in experience is so vast that people's behavior has fundamentally changed—once the threshold of "easy enough" is crossed, people never return to the old way.
Butterfield’s two attempts at gaming (Game Never Ending and another unnamed game) both failed, yet each gave rise to successful consumer software—Flickr and Slack. He argues that these four threads are essentially the same thing: using computing technology to facilitate human interaction.
Historical trajectory:
Key turning point: During the development of the second game, the team used IRC as an internal communication tool. However, IRC was a 1989 protocol lacking store-and-forward functionality. They gradually built a log bot, search functionality, database alert integration, and more—these “incremental improvements” eventually became the precursor to Slack.
Butterfield’s assessment: Nearly all the value in consumer internet comes from “facilitating human interaction in different ways” (excluding Google’s search business). Slack succeeded not because it was a “better way to input text,” but because channel-based messaging is a new category—it puts the team/organization first, not the individual.
Butterfield argues that knowledge workers spend over 50% of their energy on "basic communication and coordination"—daily stand-ups, quarterly business reviews, approval meetings, progress syncs, and the like. For a company with 10,000 employees and a $1 billion payroll, this means $500 million is invested in activities with almost zero efficiency improvement.
Core thesis: If any improvement can be made in this area, its impact would be "disproportionately greater than any other change the company could make." Butterfield cites Dustin Moskowitz's term "work about work," emphasizing that these seemingly trivial coordination activities are actually critical—without context on goals, status, and roles, teams "drift," and people end up doing the wrong things.
Critique of "hybrid work": Butterfield believes that in 10–20 years, the term "hybrid work" will seem as odd as "paperless office." He proposes a key framework: Don't start from February 2020 (pre-pandemic) and ask "what's missing"; start from today and ask "what's added"—remote work does not have to be done at home; employees can go to coffee shops or co-working spaces; companies can recruit more broadly; employees have more choices.
The true value of the office: Butterfield breaks down the functions of the office as: ① Power projection (flagship buildings) ② Recruitment showcase ③ Cultural landmark ④ Client reception ⑤ All-hands meetings/training/inspiration—while "having people sit at desks working alone on laptops" is the least valuable function, yet occupies the most space. He admits that although he has recognized this for 9–12 months, Slack still has no new office layout plans, illustrating how difficult change is.
Butterfield draws on the book Leadership and Self-Deception, whose core argument is: when a person acts against what they believe is right, their immediate response is to create a narrative that "someone else is wrong."
Specific mechanism: The book tells a story—a baby cries in the middle of the night, the husband wakes up thinking "I hope my wife wakes up," then begins blaming her for being "too lazy." Butterfield maps this to the workplace: when results fall short of expectations, leaders instinctively blame others. But "if someone was placed in the wrong role, who placed them there?"
Butterfield's self-reflection: He admits, "This is my instinct every time"—when things don't go as planned, his first reaction is to blame others. Yet in nearly every case, the problem lies with himself (or other leaders) failing to identify the incentive structures that cause people to behave in undesired ways.
Specific example: For years, Butterfield worked to get product managers and designers to accept that "making mistakes is normal"—much like writing, where you draft the first sentence, then the second, realize the first two are too complex, and revise—this is the only path to the right answer. But people feel they "must get it right the first time," so they spend endless hours in meeting rooms debating "recipes" (static interface mockups) instead of actually "cooking" (building and iterating). Butterfield believes that as CEO, his responsibility is to create the right incentives and communication methods, not to blame the team.
Butterfield links the ethos of open APIs and open-source software to "hippie" culture—this approach of "providing tools for free" may seem to sacrifice short-term profits, but in the long run, it is a significant advantage.
Flickr's data: The open API spawned approximately 2,000 third-party applications and around 900,000 customer-built integrations. Butterfield argues that if it had remained closed, its utility would have been greatly diminished.
Implications for investors: The software industry is far from a zero-sum game. Butterfield notes that the revenue and market value of software companies that went public in the past 24–30 months "simply did not exist 5–7 years ago." Slack itself (with 2,800 employees) procures software from 450 different vendors, while large enterprises use an average of 1,000 cloud services—a number that "has been monotonically increasing for decades and will continue to do so for decades to come."
Core judgment: In the software industry, an "abundance mindset" is far more effective than a zero-sum mentality. Because the overall pie is growing rapidly, open and generous approaches create more value over the long term.
Butterfield uses the analogy of "shovels and excavators" to illustrate the moral neutrality of technology and its risks: an excavator is 100 times more efficient than a shovel, but "accidentally" killing someone with an excavator is far easier than with a shovel—killing with a shovel requires "real determination."
Historical analogy: In the mid-19th century, the Thames River and Boston's Charles River once "caught fire" due to industrial pollution, but humanity eventually learned to reduce black lung disease and dangerous factory conditions while reaping the benefits of the Industrial Revolution. Butterfield argues that work software, though not as dangerous as early coal-fired power plants, similarly requires time to learn how to "integrate new forces into existing habits and practices to gain benefits without causing side effects."
Most Interesting Open Questions:
1. Can humanity achieve sufficient progress in material and energy efficiency to address climate change and bring billions of people to the living standards of developed countries?
2. Can we realize a "15-hour work week"? Or will humans continuously invent new work challenges, reducing the space for creative activities?
Butterfield's Personal Philosophy: He believes that the purpose of life is, to some extent, "creative expression"—whether through sculpture, management, or technological innovation. He recently resumed playing jazz guitar, emphasizing the importance of "mastery"—which encompasses both theoretical and creative components.
Advice for Builders and Investors: Anything worth doing is worth doing well. Butterfield stresses the importance of the "degree of craft"—because the owner's illusion makes it too easy to assume that others care about and understand the product as much as oneself. Successful products/services have all found the point where "things are made easy enough for people to change their behavior."
Most Profound Philosophical Question: Butterfield considers "Why is there something rather than nothing?" to be the most profound question. He recommends that everyone read an introductory philosophy 101 book, because "all progress, all journeys along an upward arc, come from questioning things we thought we had answers to but never questioned enough."
| Position | Guest Sentiment | Key Data |
|---|---|---|
| Slack | Bullish (founder's perspective, but acknowledges challenges) | A 2,800-person company procures software from 450 vendors; 900,000 customers have built custom integrations |
| Flickr | Retrospective analysis (success case) | Open API spawned 2,000 third-party applications |
| Uber/Lyft | Neutral (as a case of behavioral change) | Functionally "equivalent" to taxis, but differences in experience led to fundamental behavioral shifts |
| Amazon | Neutral (as a case of communication innovation) | Six-page memo format |
| Discord | Neutral (analogy) | Also a communication tool born from a failed game venture |
1. “Owner's Delusion” Law (Butterfield): Founders naturally overestimate how important their product is in others’ eyes—when users visit a website, their minds are occupied with arguments with their boss, health worries, and family conflicts, yet founders assume they bring the same enthusiasm as themselves. Evidence: Restaurant website case—users just want to find the address and phone number, but are forced to watch slow-loading food photos.
2. The essence of innovation is behavior change (Butterfield): “No great innovation results in people’s behavior remaining exactly the same afterward”—the ultimate measure of innovation is not how cool the technology is, but how many people’s behavior it changes. Evidence: Slack allows new employees to access the organization’s history on day one (something email systems cannot do), creating “a three-to-four order-of-magnitude gap in information access.”
3. The boundary between product and market matters more than either side (Butterfield): Using the yin-yang metaphor—one side is the market, the other is the product, and what truly matters is “the boundary where the product touches human attention or creates utility.” Evidence: Steve Jobs’ genius was more on the marketing side (understanding what people would accept) than on the product side.
4. Constraints are catalysts for creativity, not enemies (Butterfield): The 5K contest (creating a webpage within 5KB) became a global phenomenon because “the relationship between constraints and creativity is ancient and profound”—music has keys and time signatures, poetry has rhyme and meter, visual arts have formal limits. Evidence: The S-shaped utility curve—constraints help teams filter out the “steep part” from infinite possibilities.
5. “Work about work” consumes 50%+ of knowledge workers’ energy, yet receives zero investment in improvement (Butterfield): Basic coordination activities like daily stand-ups, quarterly reviews, and approval meetings consume more than half of working time, yet companies invest almost no resources in improving their efficiency. Evidence: In a company with 10,000 people and a $1 billion salary cost, $500 million is spent on activities with zero efficiency improvement—this is “the source of the biggest difference between the best and worst teams.”
6. Don’t start from “what’s missing,” start from “what’s added” (Butterfield): When thinking about remote/hybrid work, don’t ask “what’s missing compared to February 2020,” but ask “what’s added compared to today”—remote work doesn’t have to be at home; it can be at a coffee shop; companies can recruit more broadly; employees have more choices. Evidence: The least valuable function of the office (having people sit at desks working alone on laptops) occupies the largest area.
7. Leader’s self-deception: When results fall short of expectations, instinctively blame others (Butterfield): Citing Leadership and Self-Deception—just like a husband blaming his wife for being “too lazy” when the baby cries at night, leaders almost always attribute failure to others. Evidence: Butterfield admits “this is my instinctive reaction every time,” but as a CEO, “if someone is placed in the wrong position, who placed them there?”
8. The software industry is far from zero-sum; an abundance mindset is the right posture (Butterfield): The revenue and market value of software companies that went public in the past 24-30 months “simply did not exist 5-7 years ago”; large enterprises use an average of 1,000 cloud services, and this number “has been monotonically increasing for decades.” Evidence: Flickr’s open API spawned 2,000 third-party applications and 900,000 customer-built integrations—openness brings long-term advantages.
9. The double-edged sword of technology: From shovel to excavator (Butterfield): An excavator is 100 times more efficient than a shovel, but “accidentally” killing someone with an excavator is far easier than with a shovel—the more powerful the technology, the more “mastery” is needed to avoid side effects. Evidence: In the 19th century, the Thames River once “caught fire” due to industrial pollution, but humanity eventually learned to balance.
10. Every successful product has found the threshold of “easy enough” (Butterfield): Uber/Lyft are functionally “equivalent” to taxis, but the difference in experience leads to a fundamental change in behavior—once the threshold is crossed, people never go back to the old way. Evidence: Slack’s channel-based messaging is not “better text input,” but a new category that prioritizes teams/organizations over individuals.