先想 skip 那一跳：Nikhyl Singhal 谈怎样把产品生涯做成六十年

Nikhyl Singhal estimates he has mentored hundreds of product managers, though he resists a simple count because the definition is fuzzy. What started 15 years ago as ad hoc help was hard to scale, so he now concentrates on coaching people through transitions — deciding whether to leave a job, choosing between offers, or working through what he calls "911 calls" at work, of which he takes a few every week. He argues those dilemma moments are the most substantive times to help someone, which is why the number is closer to hundreds than to a dozen. He deliberately keeps most of those relationships private so people feel safe calling him.

The most common mistake he sees is a short-term frame: "I really dislike my boss," "this company doesn't have it anymore," "it's just too hard to ship things." Those statements may be true, but they are also true of many of the jobs someone would move into, and a lateral move is by definition not a forward move. Instead he asks people to treat their career like a product: picture what the great end state looks like, then break it into version one, two, and three. That is why his newsletter and podcast are called the Skip — he wants people thinking about the job after next, even their boss's boss's job, so that today's role can be chosen in service of it, even if that means staying put and learning to sit with ambiguity.

The biggest short-term trap in the workplace, in his view, is conflating career with promotion. When someone asks for a career conversation, they usually want to know what it takes to get promoted, which is really a question about the company's system — its levels, its process, and who decides — and it is inherently short-term and gameable: "it's two years away, how do I make it 18 months?" A career is the long arc around the job after next, and the promotion has to be understood as being in service of something. He probes that directly, because many people reveal the real goal when they say, "as soon as I get promoted, I'm going to leave."

Collecting brand logos strikes Nikhyl as shallow for builders, because most product and technical people got into the work to build things, and when he asks people when they were happiest they describe building something — often regardless of whether it worked. What he does value is a diverse set of frames: watching pre-product-market-fit, seeing smoke turn into fire, then turning fire into something great, plus inside-the-building versus outside-the-building problems and consumer versus B2B problems. Logos still matter, especially for executives hired to bring expertise for a company's next phase, since scale experience is an endorsement — though later-stage companies can substitute for it. He also argues FANG is out of date and prefers MAGMA — Meta, Apple, Google, Microsoft, Amazon — while conceding it drops Netflix, Adobe, and Salesforce.

Nikhyl traces the current situation back to a decade of blitzscaling, when distribution platforms got so good that a company could grow in 18 months what used to take ten years. The belief that any company could instantly grow once it found product-market fit birthed a generation of unicorns. Then, roughly 18 months before this conversation, "the music stopped" — zero-interest-rate money went away and growth could no longer be bought with capital. Companies retreated to their core products, laid people off, restructured, and pushed managers back into IC roles.

The pocket he worries about is the large group of growth companies that raised substantial dollars and therefore aren't going to run out of capital in 2022 or 2023. They aren't raising new rounds, they aren't laying people off, and they're still hiring and still searching for the next product — quietly "sleeping in the shadows" while their public contemporaries are worth 10 percent or less of their former value. He expects boards to pull capital back, effectively telling a scaled ocean liner crewed by 300 people to go find product-market fit, which does not square with a multi-billion-dollar valuation. Employees who joined expecting their equity to be worth something are the ones exposed.

His diagnostic is two questions. First: are we scaling a product that customers love, with what he calls a tremendous sucking pull, or are we still trying to find that pull? Second: is the valuation in the hundreds of millions? If the answer is "still searching" plus "hundreds of millions or more," you are at an ex-growth company. Employees have almost no recourse, because staying amounts to a new four-year investment while half or more of their compensation is equity that may be worthless — whereas founders can recap, reset the stock price, return money, or restart, even if they have more to lose.

The clearest signal of product-market fit is pull: how hard you have to work to get people in the door, whether customers arrive with little marketing spend, whether the sale is easy. He points to OpenAI as an extreme current example of pull that exists before massive revenue or profitability. There are legitimate reasons to stay anyway — it's the biggest role you could get, an executive team seat that is genuinely career additive, or real loyalty to a team you built, which he respects as long as you put bounds on it. Fear of not being able to find another job is the one reason he rejects outright, even though it is very common.

For early-career PMs, Nikhyl has two pieces of advice. The first is to become world-class at one of the ambiguities a product role exposes you to: craft, market ambiguity (creating something that doesn't exist), organizational ambiguity (solving complex inside-the-building problems), domain expertise such as ML or hardware, team expertise in managing managers, and growth. A PM is confronted with all of them at once, so picking a lane as early as possible is goal number one.

Goal number two is having a story to tell the next employer. In hundreds of interviews, he hears candidates describe early jobs as things that simply happened to them — "I was there, and this happened" — with no way to connect what they learned to what they did. He borrows Amazon's habit of writing the press release before building the product: think about the story and the skill first, then run your week-to-week and your performance review in service of it.

When someone isn't getting promoted and asking their manager doesn't reveal why, Nikhyl offers four common causes. The first is missing advocacy: you need someone who sees the magic, and it doesn't have to be your manager — if you have the magic and no one sees it, you're in a bad setting. The second is that the next role doesn't exist, which was rare in hypergrowth but is everywhere now that layers are being removed. The third is impatience: high performers set aggressive timelines, but leadership skills are softer and impact is lagging, so the delay may be working as intended. The fourth, which he puts at roughly half of cases, is a genuine development area that the manager may miss, the promotion committee sees, and the individual refuses to hear because "this is who I am."

The way out is ground truth rather than formal feedback. He has run experiments where teams ignored the official feedback process and held real conversations with peers, and the signal that emerged was dramatically different. What he looks for is strong listening: triangulating input from people who see you often and people who don't, creating safety so no one fears retaliation, and resisting the urge to justify why the feedback is wrong. He cites Jules Walter's technique of repeating feedback back better than it was delivered, and adds a manager's version — naming the person who gave you feedback in your staff meeting, so everyone learns that giving you input gets recognized.

Nikhyl predicts that in a hundred years, archaeologists studying early tech will be most surprised by how acceptable it was not to train managers. The military corrected that mistake early, and most mature industries train their managers, but tech assumes that if you're a good coder you're ready to manage. He calls the coupling between building something well and teaching others to build it a massive epidemic, the source of a thousand downstream problems around bias, coaching, and unsolved development areas. He suggests podcasts may currently do more for management education than most corporate L&D departments.

His mental model for a first-time manager is not divide-and-conquer and not teaching someone to ride a bike by holding on and letting go. It's the sidecar on a motorcycle: the other person is driving, you're attached whether you like it or not, and your role is relatively specific counsel. The second trap is assuming that organizational power is earned power — he describes a manager as a vampire who has to be invited in. He asks new reports "what can I help you with," and sometimes the honest answer is "I don't need you, I don't need another layer between me and the CEO"; saying anything after that backfires, so you wait in the sidecar, and when you're finally invited in, you pick one or two areas and partner deeply.

The senior IC track is rising partly because of the growth backlash: management used to be the way to be in charge of expansion, and now there is far less expansion to be in charge of. Nikhyl thinks it's one of the best things to happen to the industry, because a decade of early-promoted, untrained managers produced a lot of average managers and a weak story to tell candidates. In hiring, he compares someone who says "I managed two people but the team did the details, and before that I was an IC getting something out the door" with someone who stayed an IC, went from learning to demonstrating to owning one ambiguity — domain expertise or organizational complexity. For cracking a hard market or a messy organization, it's the second person he wants.

The tracks are real but not yet set in cement. He calls them tender and expects promotion systems and industry perception to solidify over the next six to nine months, after which they become a genuine path for builders. Engineering and design already work this way, with technical and craft tracks alongside management. In product, he thinks it's a long-standing bug that "manager" is in the title, which pushed nearly everyone into managing rather than building.

Nikhyl frames learning as three channels: self-service tools that keep improving, structured teaching that he considers weak, and community, which he thinks is the answer. Community creates safety, enables authentic conversation, shows people they aren't alone, and spreads best practices. Social software and Slack, Discord, and Zoom have made it feasible, and he believes everyone should belong to at least one active community where they can be fully honest — which is often hard to do with coworkers.

His own community started because coaching didn't scale. As a head of product, he kept having the same conversation on Tuesdays with different people, and he noticed how lonely the top job is: everyone is busy, everything is single-player. He took the half-dozen CPOs he'd talked to that month, put them in a WhatsApp channel, added a monthly Zoom call during the pandemic, and watched the group bond.

What made it work was the realization that a problem isn't yours alone — one member describes a difficult CEO, another says theirs is worse, and a third says they hit it and found a way out. Members try the advice the next day and report back that it worked. The group grew from six to twelve to fifteen to twenty-eight members, and he is now cautious about scaling further because scale is the enemy of trust and authenticity, calling community building a tree-balancing act. Joining requires being the top product leader at a mid-to-late-stage company, since those problems look most alike, and access is through members or the Skip CPO community on LinkedIn.

For managers who have just been pushed back into an IC role during a bad hiring market, Nikhyl argues community beats the alternative of being attached to one random project that may never ship. Lenny adds that the paid-subscriber Slack around his newsletter has roughly twelve to thirteen thousand members with monthly meetups worldwide, open to any level of product manager and adjacent functions.

Nikhyl picked up the phrase "shadows of superpowers" from a therapist and considers it the story of many executives. It pairs with the old adage that what got you here isn't what gets you there: the tools that made you successful have to be rebuilt for the next phase. The typical scene is a performance review saying you struggle with collaboration, while the individual protests that their last five reviews called them one of the best collaborators in the company — both are true.

The explanation is that you were an excellent collaborator as long as people agreed with you. As a leader you are now asked to be opinionated, so you apply the same toolset, but senior peers who may not be in product or tech recoil from it. Because the behavior is your superpower, you move too fast to notice that it needs rebuilding.

He has a catalogue of these shadow pairs: great collaborators who are reluctant to present their own opinions because they assimilate others so well; growth experts who struggle to be innovative; world-class storytellers who struggle with details; taste-makers who are always first to have a point of view but rarely introduce change; politically strong leaders whose decisions are unprincipled; structured thinkers who can't handle blue-sky innovation; amazing listeners who can't be decisive. From a 30-minute conversation, knowing only someone's world-class strength, he can often predict three development areas they haven't noticed, because they experience the feedback as an attack on their identity.

The intervention is not to dismiss contradictory feedback but to audit the discard pile — the anomalies your managers explained away as a weird peer or a bad quarter — because perception is reality. The same pattern afflicts companies: Meta's move fast and break things, Uber, Airbnb. His own shadow came from being an opinionated entrepreneur who was right a lot using small amounts of information; as an executive, opinions have to be collected and edited for the collective, which felt like becoming a weaker leader. He dismissed the feedback until someone he trusted delivered it in language he could accept, and the unlocking question was simply: what if they're right?

Nikhyl argues that the mental health struggles of people who have "made it" go underdiscussed because they look like a luxury problem next to layoffs and basic needs. He frames a career in three acts: act one is learning and building, act two is building the factory, and act three is whatever comes after. Act three is far longer now than it used to be — desk work means people don't need to retire at 60, health is improving, and careers may run 60 years, so someone 20 or 30 years in may only be halfway.

He illustrates the risk with greyhound racing from his Midwest childhood. The dogs chased a mechanical rabbit that always outran them; when the machine broke and a dog actually caught it, the dog never ran again — there was nothing left to chase. People wire their entire career around a North Star like becoming a VP, making more money, or founding a company, and never plan what happens after they reach it.

If you arrive at that North Star 30 years into a 60-year career, the second one has to be under construction already. Otherwise, he warns, people keep working because there's nothing else to do but end up sadder, or they create war when peace is needed, or spend money trying to feel more, or develop bad habits. Lenny connects this to his own experience: his single goal was to start and sell a company, he sold it to Airbnb, and then found himself with no other goals and no idea what to do next.

Nikhyl sees two arcs for act three. One is scaled economics — from millions to tens of millions to hundreds, through entrepreneurship, investing, or private equity. The other is giving, which Eastern philosophies have described for thousands of years as the end state of happiness. He is deliberately provocative in saying it's fine to take and create in acts one and two, and that if you can take money off the table, 30 years of giving is likely to be more fulfilling than either earlier act and is tremendously valuable to society. His own plan is to spend act three coaching and lifting people up at scale, which he sees as a product problem: the hard part is doing it authentically rather than just big.

Nikhyl's two most-recommended books are Crossing the Chasm by Geoffrey Moore, which he values for the beachhead concept and the reminder that marketing is under-discussed in product, and Leadership and Self-Deception, a six-hour audio narrative about a person who hits a wall with feedback they can't make sense of and is stuck in a box. He listened to the second in his late twenties and calls it unusually powerful for something almost nobody has heard of.

His favorite interview question is "what's something everyone takes for granted that you think is essentially hogwash," or for managers, "what conventional wisdom have you bet against?" The point is to break the interview mindset in which candidates spend the whole conversation predicting what you want to hear, which causes strong people to be dismissed for not being genuine. His favorite recent watch is the Disney+ documentary about Giannis Antetokounmpo's rise from obscurity, a family-friendly zero-to-hero story.

His favorite product is the Arc browser, both because it handles hundreds of tabs and because it proves that innovation is still possible on a mature surface like Chrome with a much smaller team. Lenny notes the 24-hour tab archive and the onboarding experience. Nikhyl's minor change with outsized impact is treating the meeting operating system as a product: his teams are on version seven, re-cut every 90 days with defined attendees, decisions, and weekly cadence, plus feedback at the two-month mark, because meeting time is the most expensive time in a company. New leaders with low context should reboot meetings first — process, then people, then product, then strategy.

His closing advice is to write the story you'll tell about today's role at six, twelve, and twenty-four months: the problem you solved, the skill you built, the headwind you faced, what you did about it — using "I," never "we." If the story you write isn't compelling, change the work until it is. He points listeners to the Skip podcast, which is moving toward recorded coaching calls, theskip.substack.com, and LinkedIn or Twitter, and asks for feedback on which topics he should spend more time on.