低效产品公司都长一个样，高效的却各不相同

John Cutler spent about four years at Amplitude with the unusual title of product evangelist, a job that Sandhya from the company floated to him as customers beyond classic growth-stage startups began to arrive. Each day he talked to current or future customers, ran workshops and coaching sessions, and held what he called product therapy sessions in which he simply absorbed whatever problems people brought. He estimates roughly 800 one-on-ones with leaders and, at the peak, around 150 workshops in a single year, plus talks for tens of thousands.

Cutler stresses that the well-known Amplitude artifacts were team efforts rather than marketing flashes. The North Star material grew out of a workshop Sandhya improvised for a customer, and the retention and engagement playbooks started as a 110-page body of research from the customer success team. He also insists the company never invented the North Star framework and says he always corrects people on LinkedIn who suggest otherwise.

Cutler describes the access as once in a lifetime: one day Amazon or Ikea, the next a two-person startup or Figma. He notes that, unlike a paid consultant, he did not choose who came to him, so he saw teams reach similar results in very different ways. That experience, he says, taught him the power of context, and it led him to join Toast to enable product teams from the inside in a vertical SaaS business.

Lenny relays a spicy listener question, raised by Jason Cohen, that Cutler's newsletter often lacks a clear recommendation. Cutler welcomes the criticism and jokes that The Beautiful Mess sounds like an emo band name. He explains that four years earlier he had scanned the product-advice landscape and noticed three dominant themes: success as tools, skills and mindset; a meritocratic, highly individualistic worldview; and advice stripped of context so it can be actionable.

Cutler says he does not judge those perspectives, and he admires people who produce crisp, actionable content. His own angle was to explore complex adaptive systems, where teams are nested inside larger systems and effects are non-linear, and to dig into counterintuitive dynamics. He gives two examples: teams loading themselves with too much work in progress despite knowing better, and executive strategies of three neat pillars that leave everyone unsure what to do.

The tension, he says, is that getting anything done requires reducing the world a little. The difference between oversimplifying and focusing is deciding which variables to hold constant. Lenny admits he feels the same pull, noting that he added the word mostly to a post about virality being a myth, and Cutler jokes that his own posts always say mostly, maybe or it depends.

Asked what differentiates the best product teams, Cutler starts with a friend Josh Arnold's reverse Anna Karenina principle. In Tolstoy's novel unhappy families differ and happy ones are alike, but with product companies the dysfunctional ones share easy-to-spot anti-patterns, while the high performers get there in vastly different ways. As examples, great product leaders can be humble servant leaders or dominant sparring partners, and good decisions can come from rigorous red-team processes, serendipitous collaboration or a top-down CEO.

Within that diversity he names a few recurring patterns. High performers have coherence between their structure, incentives, technical architecture and current strategy, and even brilliant hires cannot escape a strategy-structure mismatch. They hold strong opinions loosely, with a stubborn belief in things like customer contact or craft that survives short-term doubts. They also share a slightly irrational faith in what separates a nine-level product from a seven, citing the Bezos idea that today's success was set in motion years ago.

Cutler adds that leadership must be coherent, so that words and actions match, and that skills need to fit the context. Lenny summarizes five attributes: coherence with strategy, strong opinions loosely held, belief in the power of product, coherent leadership, and the necessary skills. Cutler's point is that these sound like common sense, yet common sense is hard to put in motion, and he illustrates it with Amazon's bar-raiser hiring and the founders of AppFolio, who refused to trade away quality or customer contact.

Lenny asks what share of success comes from the people a company hires and whether a team can become high-performing at all. Cutler calls it the trillion-dollar question and declines to give a percentage. He notes that a room of geniuses fails without coherent leadership and structure, while teams of people who have not done it before sometimes succeed with the right things in motion. Many high performers either vet everyone heavily or keep people for years and are good at growing them.

He says everyone should challenge their happy place. Years ago he believed skill barely mattered and that bad management explains everything, and he still leans toward giving people a seventh chance. He sees the opposite error in leaders who insist on only A-plus players or blame whoever was in charge when things went wrong. Using Satya Nadella as an example, he adds that heroic stories oversimplify, since a leader acts amid wealth of talent, structures, boards and market conditions.

For a founder deciding between investing in process or people, Cutler's first piece of advice is introspection: know what you and the people around you believe, and aim to be a coherent leader. You can nudge yourself away from your natural style but not move all the way from a process-driven meritocrat to a consensus-driven collective. Lenny observes that companies lifted by a hot market can look like the work of a genius until the market turns.

Asked why culture was not on his list, Cutler says it is the fabric that creates those other things, so he almost took it for granted. He points out that a value such as ownership means something different in an individualistic culture than in a collectivist one. The useful move, which he credits to Amplitude and Spencer, is to describe the behaviors that show what ownership looks like rather than just declaring the value.

Cutler declines to name a best company, preferring to recall moments when a team was clicking despite difficulty. One example is Angela, a leader at Lego, who he says has things dialed in while transforming a very large business with one of the world's most iconic brands toward digital. He also recalls a team in India with passion and curiosity and no jaded people in the room.

On differences across countries, Cutler defers to Erin Meyer's book The Culture Map but names two he noticed. One is individualism versus a communitarian view of the team; in an extreme individualistic setup, an engineering manager brokers a premier project for each engineer who wants promotion. The other is how deferential a culture is to hierarchy, which shapes how information flows.

Lenny notes that most PMs do not work at Silicon Valley startups and that many are in transformations. Cutler says some of his most rewarding conversations were with such companies, which face huge inertia: executives who may never have shipped product, annual budgeting cycles, large IT organizations and a belief that there is only one way to work. His first suggestion is to create pods where a team can get its reps, meaning shipping and going through the full learning loop.

His second is to think about frameworks appropriately. Many companies read how Figma works and treat adopting the frameworks as the end goal, whereas he sees a framework as a job aid for staying on track, and such companies should reinvent tools when they stop working. He also reminds listeners that Silicon Valley advice is optimized for a first big arc of growth and for pure digital products, not for legacy businesses or logistics-heavy ones.

Cutler points to a fundamental attribution bias: successful companies underrate luck and inertia, while big companies overestimate their systemic drag. He rejects a simple split between big slow companies and nimble modern ones, citing an Australian plumbing company with the revenue of 50 startups, Anheuser-Busch's BEES distribution app, and Brazil's NewBank. For further reading he recommends John Smart's book Better, Sooner, Safer, Happier and notes the universal appeal of Teresa Torres's techniques.

Cutler says the internet is a firehose of amazing information, but he wishes someone would just summarize what to do when waking up, joking about Andrew Huberman-style life hacks. His reminder is that skill is knowledge times practice, mediated by environment, habits and motivation. Working with learning experience designers at Amplitude taught him that consuming knowledge is not enough; people need to go around a loop repeatedly.

He describes Amplitude's data-informed product loop: set a strategy, build qualitative models such as a North Star framework, add measurement, prioritize, design bets, measure their impact and circulate the learning back. Breaking the loop at any point stops progress, and the loop also shows where you are weak. He worries that many leaders feel beaten up by the advice industry and believes shifting focus to getting the loop going for your team, career or company is a safe bet.

Lenny suggests just-in-time learning, saving material until he needs it, such as SEO. Cutler agrees but adds that listening broadly builds spidey sense: you learn that pricing, meeting design or service design are real disciplines, so you can tell when someone oversimplifies them. He compares it to seeing a genius professor as a freshman and not understanding everything, yet learning that the subject exists.

For people at slow, waterfall companies, Cutler's first point is that they usually underestimate the loops available to them. If an executive says to build X, you can write a one-pager treating X as one of five options, ask what would be observed if it were working, or document assumptions and risks. Each small nudge counts as going through the motions, and he urges people to work with what they have.

He argues you should write your portfolio as you go, because two years later everything will feel like a blur and you will only be able to say you worked at a messed-up company. He acknowledges that some people cannot leave their jobs and that many things can be true at once. He adds that top-tier and hypergrowth companies can be equally chaotic.

Cutler uses a model of chronic versus acute challenges: some companies work down their chronic issues, which lets them face acute stressors, while others are mired in acute ones. His example is the pandemic, where 15 to 20 companies that deliberately designed their response, rather than leaving managers to cope and waiting for normal, fared an order of magnitude better in the happiness of their people. He sees that as coherent action against a clearly seen threat.

Cutler says he is trying to stop using vague words like product sense and product mindset and to unpack them into skills that can be taught, such as problem modeling, systems thinking, decision-making under uncertainty and facilitation. He cites a should-versus-can divide he discussed with Craig at Toast and says the more interesting question is what skills the should person brings, such as decoupling current tactics from optimal ones. That, he says, will be part of his role at Toast.

He also wants more diverse role models, including leaders who are not on Twitter. He tells of a woman who feared that, to get ahead in tech, she had to adopt a very individualistic, meritocratic outlook she did not share. Cutler told her about companies and leaders, including Carrie in Santa Barbara, who work with a real team vibe, and she found that reassuring.

On writing alongside a day job, Cutler admits many posts come together at 1 a.m. on a Thursday, which can explain why they stop short of an answer. He hopes to offer more mental models for dealing with the mess, such as a model that separates clear, complicated, complex and chaotic systems, Simon Wardley's mapping, and his own Pyramid of Leadership Self-Awareness. He also wants to package his 700 to 900 posts, hundreds of images and talks into a meta-guide to his content.

Answering a Twitter question about what he could not say while at Amplitude, Cutler says he rarely held back, but he would have been more adamant about an anti-pattern in analytics adoption. Companies undertake huge implementations, spending months documenting every event and metric. He would tell them to use the free plan, instrument a few events with their developers in a three-hour session and start getting value, since you do not yet know what you do not know.

In the lightning round, Cutler recommends three books. Douglas Hubbard's How to Measure Anything argues that measurement exists to reduce uncertainty enough to make the next decision. Accelerate, by Nicole Forsgren, Jez Humble and Gene Kim, models performance using survey research behind the DORA reports. Jeff Patton's book on user story mapping is, in his words, deceptively simple and still has companies buzzing a year after they learn it.

His favorite interview technique is to ask candidates to retell a behavioral story from the perspective of a colleague they mentioned, which reveals self-awareness and whether they cast themselves as the hero. He ends with a snack-focused parenting tip and asks listeners to send recommendations of people the community should hear more from, as he hopes to start a guest speaker series at Toast.