Why half of product managers are in trouble | Nikhyl Singhal (Meta, Google)
In a Nutshell
Product management is undergoing a renaissance driven by AI, obsoleting "information movers" (half of PMs) while rewarding hands-on builders who exercise judgment on rapid changes, testing, and differentiation—leading to massive staff shedding followed by rehiring of AI-first talent in the next 12-24 months. Builders thrive with record compensation, joy from direct creation (e.g., vibe coding personal tools via Claude/Cursor), and expanding roles into founding, C-level, or non-product functions. To survive, PMs must embrace reinvention now: find joy in building, obsolete rote tasks with AI, stay modern despite burnout risks, and prioritize long-term positioning over ego or past brands.
These notes were generated by AI and may contain inaccuracies.
The skills that used to be really valued in product managers are changing substantially. The industry is in stress. Nothing's constant. Everyone's in a state of alert. Three years ago, product leaders' days were largely moving information. The information mover is becoming a dinosaur.
A recent report shows the most open PM roles globally in three plus years.
This is a complete renaissance for the product industry, but with strings attached. In the next 12 to 24 months, massive shedding of staff followed by massive rehiring. A company might shed 30,000 and hire 8,000, but the 8,000 will all be AI first. Builders will have the time of their lives, but if you don't love building stuff, you're in trouble.
To thrive, increase pace and find reserve. The next 2 years requires fire in the belly.
Nikhyl Singhal is the number one best source of career advice for product managers and tech people. Longtime exec at Meta and Google, CPO at Credit Karma, four-time founder. Leads Skip community for heads of product and chief product officers, and larger Skip Coach community for tech professionals. Through communities, 30 years building consumer products at scale, and podcast, constantly gathers insights from top product leaders worldwide.
Episode for every product person. Real talk and actionable overview of what's happening, where things are heading, and what to do now.
Nikhyl thanks Lenny for version two of the podcast and excitement to share current thinking. Was one of first 20-30 launch episodes 2-3 years ago. A lot has changed in product management.
Everything has changed. When first chatted at end of COVID, ZIRP era (zero interest rate policy), free money cresting. Discussed ICs more in demand, first layoffs starting, ex-growth companies struggling.
Three years ago, product leaders unhappy. Days largely moving information: framing team's info for boss, who frames for boss's boss. Function focused on responsibility without authority—greatest workplace stress.
Now, people having fun again, especially product folks, because they can build. Don't rely on as many people for impact. Direct connection to ideas, testing, product instincts, customers.
Complete renaissance for product industry. For strongest builders in Skip group, compensation at all-time high. More offers than ever. Next job could be founder, CEO, or C-level in other function. More opportunities than ever.
Comes with strings attached. Industry exhausted, more tired than during COVID but for different reasons. Nothing constant—figure out job, fine for decade until manager, then executive. Now, stay up or in three months, "you're doing that thing we stopped three months ago." PRDs not even a thing anymore. Everyone in state of alert.
Tens of thousands shed by larger employers also hiring at triple wages.
Mid-career people in 30s: life’s cruel trick—best, most energetic power years, figured out what doing, but settling down, kids, aging parents, aches, diet, exercise. Health, family, friends hardly see, dependent parents, dependent kids, work that changes constantly and takes all time. That generation insanely stressed.
Joy but fatigue unlike before.
Product leaders: on balance, doing better than 2-3 years ago? Top people doing much better—more choice, long-term happier, more interested in jobs. More stressed, but stress is "wish more time to feed LLM at night" vs. past malaise of decisioning.
Industry in stress. Even doing well, feel more stressed—not keeping up, industry changing, roadkill risk. Change hard on humans. Best people feeling great.
Recent Skip meetup (125 heads of product, monthly in San Francisco): show-and-tell of personal productivity builds. Joy in audience, one-upping chief of staff apps.
Product decisions, prioritization, information movement completely foreign from few years ago. Using agents in enterprise, chief of staff apps for productivity, focusing on judgment, obsoleting mechanical parts with software/AI/agents.
None of that in language 12 months ago. Hard to anticipate.
In next couple years: product leaders paid to drive judgment, tip of spear obsoleting everything else via software/AI/agents. Most stuff PMs didn't like doing, AI cracking.
Within 2 years, most companies obsolete mechanical parts of building product. 10 to 100 times more changes presented to products—cost of testing/changing much lower. Judgment paramount.
Judgment: evaluating if change good/bad; product change one way vs. another (can't build 100 custom versions—affects brand, maintainability); sustainable, differentiated; customers asking; system/platform skill, not just feature.
In 2 years, no more bad software. Examples: house with 15 apps (shades, AC, garage door)—all horrid, break, unmonitored. Claude will fix, more secure.
Prompt: "How can we make product experience better?" Yields 10 ways, implement first 7.
Super skill will fix all app store software—consistent, better, less buggy, maintainable.
Apps by non-product-first companies, non-best engineers. Now all access best engineers (Claude, Codex) via English.
Still more lines of COBOL than any language; mainframe sales strong (IBM). Complicated systems, many engineers dead (code 10-20 years old), no one touches. Can change now.
Excited to improve taken-for-granted things.
Skills acceleration: ICs, hands-on, opinionated—even pre-Claude/ChatGPT.
Companies scrutinizing staffing: overhire? Doubled staff last 5 years, get twice as much? At Google 10-15 years ago (ancillary team), asked how many really needed—non-tech guess 90%, reality ~9% (500 people).
Huge overhead—hire to expand, try new things.
Judgment day: not getting value from last 5 years growth + AI needs different skillset.
Next 12-24 months: massive staff shedding, then rehiring (shed 30k, hire 8k AI first). Let go: underperformers + set lighter destination.
Challenging 1-2 years.
Most open PM roles globally in 3+ years (last high during COVID). Good news.
Builders have time of lives, comp up, Bay Area back in favor.
Why so many PM jobs amid layoffs/struggles? Depends on defining product manager. Dramatic shift in definition. Three years ago, archetypes; some more in favor.
Three years ago, discussion highlighted product manager archetypes, with IC builders in favor. Now, information movers will become dinosaurs. About half of product people who grew into the industry have that skill as their superpower. Others entered product from founding or building motivations and are the builders being hired today. Everybody wants a builder. Engineers, designers, and marketing folks are often builders. 'Builders wanted' will be the tagline for the next couple of years.
Building is fun. Prediction: choosing between preparing a podcast or coding in Cloud Code, prefer the latter for the progress and satisfaction of seeing things improve.
When I was a PM I used to take a break and go change the light bulbs in the house. The reason was the light bulb was broken, but replacing it made the light come on—satisfying because product jobs rarely offer that immediate fix. It's the challenge of responsibility without authority. Until now, PMs can participate in the joy: create designs without waiting for designers or backlogs.
Builders will have the time of their lives, focused on judgment, invading other functions.
In his group started five years ago, first 2-3 years to 60-70 people had one founder. Last 12 months to 125 people, now 14 founders—many choosing founding CEO over product executive roles.
One senior person interviewed for CHRO (Chief HR Officer) position because they wanted product manager background. Forward-leaning companies need obsolescence skill, judgment skill, empowerment skill, builder skill in functions like HR—the function might be easier to learn.
Product builders will have broader opportunities up and down the stack: engineers and PMs becoming product leaders, blurring lines.
Non-builders who don't love building are in trouble—about half the people. Those not into tech, who saw it as lucrative for communication and information moving, team building (all valuable skills), face challenges. They may leave tech, build new businesses using AI tools, or find jobs outside tech.
Hot take from Arnold, head of growth at Anthropic (PM): as engineers do more, PMs squeezed managing many features, ideas, docs. Push for PMs to build themselves, but leverage higher not coding but staying on top. Need more PMs as engineers speed up—for prototyping, ideating, feedback, alignment. Better to do higher leverage work than shipping.
Depends on 'shipping': if engineering team of 50, being the 51st is cheap engineer knockoff. If building ways to stay on top (past: tickets, backlogs, standups)—that's information overload. CPOs building internal tools for efficiency: automated product reviews, standups.
With 10x stuff happening, need judgment on good/bad changes. If manual, overwhelmed. Combine hiring product builders with judgment and folks building internal tools for decisioning—not through human management but technology.
Wow, if I do this, I don't have to ever do a status report... I hate doing status reports. Now I wrote something and my boss is happier with more detail.
Exciting: building internal software to scale. Make yourself scale through software—big opportunity, fun like building your own thing.
Adults still needed—companies grow quickly, driven by founders. AI supplements, but judgment from expertise and wisdom. Attractive for hands-on wisdom, credibility with founders, speaking their language, prior experience.
Pace increases: feed LLMs running at night, seize opportunities—dizzying, risks burnout.
Geography and diversity step back: AI wave Bay Area-heavy, hiring fewer like themselves (age, gender, ethnic backgrounds). High pace impacts: women having kids in power years can't spend nights/weekends in Cloud Code.
Brands matter less than modernity in delivering product. Past 10 years: experience at scale brands signaled competence. Now, ways of building software alien to past—feedback: 'What tools do you use? Your judgment? How do you think?' Not 'Five years ago you shipped this.'
How modern you are now is career advice, not established brands. Big brands may not be AI-forward; 6 years there leaves you outdated. Logos on resume may hurt if company not current.
Big brands hard to highlight impact: e.g., at Meta, optimizing algorithm piece had huge impact, hard navigation—but falls flat vs. living in future.
Skills valued flipping: start as builders, taught leverage/scale/organizing/enabling others. 'Stop working on the car, work on the factory.' Now scale done differently, care about your opinion and what you build—jarring.
Change hard—humans not designed for it easily. Kids reinvent constantly: crawl to walk, learn languages fast, unafraid of mistakes—teach kids to ski early. Adults trained for stability: happy medium, few changes, stay in job.
Mental block: 'I worked hard to get here, why change? I did the deal: school, job, brand, management, income, family—now start over?'
Power years: equally disappoint everyone (parents, kids, health, family, friends, partner, work, retirement)—prioritization with limited hours vs. demands.
Psychological factors: take week off to learn latest, but 3 months later antiquated—target keeps moving, never in flow.
Best at old system find reinvention hardest—shadow superpower: mastering old blinds to new. Weaker performers more open to change.
#1 advice: courage to cross mental threshold, prioritize staying modern/current above all. Embrace reinvention—that's the world now.
Analogy: AI lab heads (Demis, Sam Altman, Dario) would slow if all agreed, but game theory prevents it—can't stop, huge advantage not to.
Baby version optimistic: past shifts like hardware PM (HP, Cisco, AMD long cycles) to internet (Google APM program birthed new breed—jarring, not business school direct). Microsoft/Meta adapted. Next years chaos: agents, judgment forms, role blurring, backgrounds shifting. Then settles to optimization.
Job not moving information but evaluating: what success looks like, if it makes sense for system.
It's a totally different skill. But after a couple years, there's going to be some routinness to it. There's going to be some training. There's going to be consistency. The job you had next is going to look like the previous job. You need to cross the threshold to stay modern because things are changing. Right now is the moment when if you love building, you have to stay current because you will be happier and you'll be more relevant. And if you don't love building, you have to recognize the industry is moving away from you.
This is not forever that it's a couple years potentially maybe a little bit longer but it's not a forever thing where you have to lock in and sacrifice forever. This is the moment to get on the rocket to get on the ship.
Compensation has never been higher, there's a lot of open roles, people that are embracing this are having a lot of fun. Most PMs are sitting there all day waiting for stuff to happen, waiting for meetings and approvals and all this alignment and all this PR reviews and now you can ship stuff so often.
For someone who's afraid, who's a builder, and nervous about their career: Do you really love your day? If you put green, yellow, and red next to the meetings you had and did that for a week, the vast majority of PMs in a product organization show mostly yellow and red. The ones moving into build mode show mostly green and yellow. Yet most are sitting very afraid feeling stuck.
Transition mentally and physically from fear and sidelines to being in the game. There is a moment where they experience the first joy in using the new tools. Everyone has a story: "Oh, I was doing it a certain way and then all of a sudden I built this thing." Examples include: partner and I use this app that I built; chief of staff app to keep track of my inbox; manage the lights in my house using this thing. Usually accompanies a story like stayed up all night, spent time talking to friends or hacking away or talking to Claude.
Even my wife has a story about how she has this business that she's thinking through and she went from using the AIs to do business plan to actually do test market. Everyone has this moment of joy and then they're hooked. Joy is the biggest antidote to burnout. The moment you have joy, it doesn't feel like work. Product management will be building and joy, not work.
Find a path to get there. The moment you do, your psyche creates time, doesn't feel like you're disappointing others and builds energy because people have more energy than they realize. They're just exhausted by the monotony of what was defined as product management.
Number one piece of advice: Have you found joy? If none of this is joyful, you're probably not in for the next version of our industry. Be respectful and honest with yourself. But a lot of product people that are anxious are going to be happier once they cross the chasm. Find a way to create that moment of joy, whether an app on the side or at work.
If you're a leader, find those moments of joy in your staff because that's contagious. I like building stuff and caught the bug. I try to find TV shows that I can vibe code in parallel to.
Amazon Prime shows are good because they're built on books, structured. Alex Cross, Jack Ryan. Hilarious: binge watched a season of 24, remembered I loved it but not the plot, vibe coded the whole time.
Pretty all-in on Claude these last three months. For a month pretty aggressive on Cursor because their newer stuff with highest level of reasoning is advanced. Hard to switch between tools, try to standardize.
Build a bunch of web properties for my community. Anytime I see something I can obsolete in code, I build code around it. For 100 people, thoughtful matching: who's the best person to meet, haven't met before, their haves or wants. Used to be manual like "Jay-Z would appreciate meeting Annie, but don't know if they've met." Now write software: agent that does matching; agent to figure out jobs head of products are hiring, make available, build mailing list for folks interested in work matched automatically. Next generation of recruiting.
Use AI for content: AI takes questions from people, trained on my content like yours, gives answers. I read questions, evaluate answers, identify themes where LLM and I disagree, then write.
Anything I do that I can replicate, I try to obsolete myself. First job, asked best engineer: definition of great engineer? "The best engineer I know is my dad" (before tech). "An engineer is someone who obsoletes themselves from everything they do."
Taken that to every job. People say "No, I don't want to obsolete myself. That's my job." If I obsolete, there'll be a better job for me. AI puts an A-pop on that: obsolete by asking AIs to do it.
Stack is: what can I do to obsolete anything and everything I do on a daily basis.
Best tactic: solve your own problem. Think about something not great in your day-to-day or want to improve. Go to Lovable or install Claude Code, download Cursor, tell it "I want to build the dashboard to control my Sonos" and it walks you through everything in English.
You don't need to be an engineer. Just be opinionated on what you want to see and know what you want, know what good looks like. Moment you get it, light bulb turns on. Used to have to work to do this, now agent does it. What's the next agent?
Your job can be much more joyful as PM. If you look at your pie chart today of how happy you are during the day, you can actually be much happier.
You're staring at change, which is scary. Change is such an alarm bell. The things changing are the things you don't love. There is a world which is better, but you have to go through the tunnel and the tunnel sucks. May mean changing jobs.
Prediction: vast majority of people listening will be in a different job in next 5 years because they choose to move (company struggles to stay modern) or companies shedding and returning staff.
Find moment of joy. Engineering mindset: obsolete yourself on something today, reduce less joyful parts.
Increase pace. Next two years requires fire in the belly, agency. If you leave a job and start new, year one not at year five pace of last job. New relationship, first year put best foot forward. Find that reserve, make time, maybe disappoint others to create time to stay current, find joy, obsolete.
Swallow your ego. Don't say "I was XYZ leader, only consider same level." Everything changing, past brand doesn't matter. Egoless perspective: willing to take something smaller to go through the tunnel correctly.
Stay long-term focused. Spend next couple years on the boat leaving station to new world. Once there, cream rises to top.
My property called The Skip. Best career advice: not next move, but move after. Skip job, skip opportunity. Make sure skip opportunity saved, get high salary, premier builder role.
PMs go to every industry as agents of change: can talk, broadest view through technical lens. PMs first to tools, change agents in company.
Next year, product organizations change how they build. 12 months later, marketing, sales, HVAC company bought by PE firm, school: all not current, need change agent who works in future.
Bullish PM leaders like dandelion seeds blowing everywhere.
Influx to product from design, data science, engineering: those with judgment, can talk, want to stay current, have point of view.
Designers may move into product: platform not pixels but product itself.
Crazy influx, exodus, and some can't get in: three changes happening instantaneously explain anxiety.
Tweet: only four jobs remain: 1) product engineer, vibe coder, PM slop cannon; 2) security/infra person; 3) hot people (present easy UX to world, customer service); 4) grown-ups (adults).
Agree. Adults exist: top of field that can talk, opinionated, go broad (like hot people).
Anthropic head of growth: so much time now alignment as PM. Harder: aligning people in company or AI/AGI?
Alignment: getting people right information, ground truth. That sucked (status reports, docs passed up chain with spin). Now dramatically better.
Now: decide, have opinion, fight for it. Conversation clearer with ground truth, less spin. CEO can ask agent: situation? Performance? Customer wants? System effect? Product change?
Fight with real credence. Forum for opinionated PMs.
Alignment won't go away, just less theatrics. Larger companies have theatrics AI removes. Most PMs: waste of time.
80% of PMs don't want boss's boss job day: back-to-back meetings. Answer changes in next two years.
Engineering changing even more. Engineers thriving become more PM-like: coding solved, now what to build, great? Direction? Success?
Engineers advantage: think in systems, obsolescence. Decide if change sustainable, automated, blur.
Product person edge: judgment, communication. Engineer edge: system scaling, change effects. Designer edge: taste.
People with all do better. Industry safe: more hiring, going to grow. Different class of individual.
Design roles are plateauing, with teams not valuing design as much as expected. As products grow exponentially, design should stand out by building beautiful, thought-through experiences, but that's not happening. AI enables non-designers to become great designers, unlike engineering or PM roles where AI helps but doesn't fully replace skill. Despite AI, design isn't more successful.
We just don't know what to make of design in this era.
There are pixel generators (production-focused) and taste makers. The industry conflates design with production rather than taste. Companies hiring head of design often seek production capacity, not taste makers, biasing hiring plans.
Chaos is a ladder.
Being operational now is a ride with good and bad. COVID had challenges with change; now everything is questioned, including how to define joy.
There's room to be optimistic, but it requires recognizing a small period of change and exhaustion to find joy. Find reserves to cross that chasm—it's infectious, but waiting makes it harder. Urge rethinking your craft. Many in the community are thriving with smiling exhaustion (better than plain exhaustion), though the pace is relentless.
Not a reader; vibes code instead of consuming content. Reading is mostly staying up to date via agents, Reddit, X. Podcasts and reading have suffered.
Read James (Huckleberry Finn from Jim's perspective). Fascinating retelling of a classic through another lens—powerful but haunting. As a kid reading Huckleberry Finn, it's innocent; as adult, from Twain's forward-thinking perspective, it's different. Looks for unnatural stories in plain sight.
Paradise (Season 1 and 2)—apocalyptic character drama on dealing with challenging times and motivations. Season 2 almost wrapping up.
Lioness (Paramount Plus)—covert CIA group protecting U.S. security, committing to a bigger cause. Well-written. Watches TV without vibe coding as needs attention.
Not a car guy; wants reliability. Owned Tesla Model S since launch. Daughter passed driving exam, prompting upgrade. Tried latest Teslas with self-driving—works well (95% of time). Discovered mild driving anxiety that disappears with self-driving. Drove 40 years since age 14 in Kansas; wife noticed anxiety in snow to mountains.
Eliminates anxiety sources. Self-driving reduces it. Early Tesla self-driving versions scared people; newer hardware/software is better. Challenge: Convince past users to retry (early/often releases create bad memories). Like Uber, needs friend referrals.
Suggestion: Gift 30 days free self-driving to turned-off users via referral/endorsement for monthly subscription—viral + enablement growth technique.
Loves Mad Max mode personality. Flips between hurry/standard modes; wishes no lane changes at normal speed.
High school yearbook quote: "Genius is 1% inspiration, 99% perspiration" (Thomas Edison, misattributed to Einstein). Hard work matters.
Through AI lens: AI does perspiration. Moving to world of inspired individuals—focus on inspiration as AI handles perspiration.
Started Skip due to lack of operator best practices for emerging operators. As founder/executive, shares direct experiences. Opposite of content-focused non-operators.
Built community of 125+ head of products reinventing product management (was 20-30). Shares wisdom to advance careers—more important now with changes.
Launching skip.help next month: Agents trained on 50 community leaders. Ask questions (interviews, navigation, building chief of staff app)—wisdom of crowds, powered by Superme startup.
Curated community with infinite waitlist, grows slowly for trust. skip.coach mailing list shares wisdom broadly.
Domains: skip.coach, skip.community, skip.show (podcast/newsletter), skip.help, theskip.substack.com.
Collaborating with Lenny's podcast/newsletter. Approaching 50th episode; inspired by Lenny (past 300). Two dozen Skip guests on Lenny's show.
Fills gap in tactical, non-sugar-coated product career advice—not flashy AI-focused.
Thanks for support; helping product builders together.
Keep Lenny's Podcast in your library
Save the videos and channels worth coming back to, and find them again in one place.
