Tomer Cohen (CPO at LinkedIn)08 September 2024

How LinkedIn became interesting: The inside story

6Frameworks

14Insights

Episode overview

LinkedIn CPO Tomer Cohen explains how a neglected, promotional activity feed was rebuilt around professional knowledge exchange and matching the right experts with the right audiences. He describes isolating a randomized two-million-member cohort, unifying product and AI teams, and measuring deeper value instead of shallow traffic so the turnaround could proceed without destabilizing the wider business. Cohen also shares his “we might be wrong, but we are not confused” leadership principle, LinkedIn's shift to an AI-first operating model, and his conviction-led approach to product leadership and career growth.

Key ideas

Clarity of thought and clarity of execution matter more than protecting oneself from being wrong.
Strong product principles need meaningful tradeoffs, and declared priorities must be reflected in talent and resourcing.
Turnaround products can be tested in isolated randomized cohorts to demonstrate behavior change without disrupting entrenched organizational metrics.
LinkedIn redefined its feed around professionally useful knowledge exchange rather than promotions, upsells, or traffic distribution.
AI acts as a marketplace matchmaker by connecting creators and experts with the smaller number of professionally relevant people who matter.
AI-first product leadership includes owning algorithm objectives, features, data collection, fine-tuning, infrastructure, and inference—not delegating them as a black box.
During a technological shift, teams can first diverge through broad experimentation and then converge around a few top-down strategic bets.
Cohen advises product builders to pursue the intersection of conviction, personal excitement, company importance, measurable impact, and continuous learning.

Transcript available · source text is retained privately and is not published

Frameworks in this episode

Innovation7 steps

AI-First Product Leadership: Hold the Paddles

If AI steers your product, the product leader must own the objective, features, data and infra.

Read →

Strategy5 steps

Backwards From the Peak: Ambition Set by Potential, Not by Numbers

Derive goals from what the thing could be, then work backwards; today's numbers are not the anchor.

Read →

Mindset6 steps

Conviction-Over-Demand Career Navigation

Stop chasing the most in-demand role; chase the conviction where you can make a dent and learn.

Read →

Innovation6 steps

Diverge-Then-Converge Roadmap Reset

On a platform shift: drop the roadmap, keep the objective, let teams go crazy, then pick 4-5 bets.

Read →

Strategy7 steps

The Carve-Out Cohort Turnaround

To fix a load-bearing product, secede from the metrics system with a small randomized cohort.

Read →

Leadership6 steps

We Might Be Wrong, But We Are Not Confused

Make being wrong cheap and being confused unacceptable, in both thought and execution.

Read →

People & resources mentioned

Attributed to the moment in the episode. Timestamps are approximate.

People · 14

Tomer CohenMentionsChief product officer at LinkedIn overseeing teams responsible for LinkedIn products and experiences.
“today my guest is toer Cohen toer is Chief product officer at LinkedIn”
▶ Watch at ~1:00approx.
Dan RothMentionsmentioned
“a big thank you to Shir gasar Dan Roth Josh Redfern and Spar Aral for question suggestions”
▶ Watch at ~2:00approx.
Josh RedfernMentionsWorked with Tomer Cohen.
“one of the folks that I talked to that worked with you about this phrase Josh Redfern”
▶ Watch at ~10:30approx.
Vlad LoktevMentionsVlad was my manager at Airbnb for many years and he's the person that I credit most for teaching me
“chatting with Vlad who worked at Airbnb for many years he was actually my former manager at Airbnb”
▶ Watch at ~14:00approx.
Brian CheskyMentionsBrian chesky at Airbnb there was a whole article where we unpacked this moment in their product development cycle
“he reported to Brian for a long time”
▶ Watch at ~14:00approx.
Brian CheskyMentionsBrian chesky at Airbnb there was a whole article where we unpacked this moment in their product development cycle
“a recent podcast episode had with Brian chesky where he introduces chaos sometimes”
▶ Watch at ~48:30approx.
Deep NisharMentionsHeld Tomer Cohen's current role at LinkedIn when Cohen joined.
“I met with who was who was in my role today DP nishar that time”
▶ Watch at ~52:30approx.
Steve JobsMentionsSteve Jobs is just uh the bar he held for the company and for technical talent and for excellence was not uh wavering.
“when the Steve job's biography came out everybody read it”
▶ Watch at ~57:00approx.
Dan BarberMentionsChef featured on Tomer Cohen's podcast.
“all the way to like a chef Dan Barber”
▶ Watch at ~57:30approx.
Ed CatmullMentionsthe book Creativity Inc. by I think Ed Catmull where he talks about how new ideas are this ugly baby
“from the co-founder of Pixar at cat mall to the CPO of canva or Spotify Roblox”
▶ Watch at ~57:30approx.
Carol DweckCoinedResearcher cited for a study on ego depletion and author of Mindset.
“Carol wer wrote the book”
▶ Watch at ~59:00approx.
Daniel KahnemanCoinedhis influence was so huge you know and his intellect was so huge and his Insight was was so change making
“thinking fast and slow um by Daniel canaman”
▶ Watch at ~59:30approx.
Andy GroveCoinedBusiness leader cited for distinguishing competency from motivation.
“high output management uh by Andy Grove”
▶ Watch at ~1:00:00approx.
Lenny RachitskyMentionstoday we've got another very special compilation episode something I've been pulling on more and more with the podcast and the newsletter
“Lenny thank you this our pleasure”
▶ Watch at ~1:08:30approx.

Resources · 43

LinkedInUseswebsite · LinkedIn
“I check it at least 10 times a day”
▶ Watch at ~0:00approx.
LinkedIn FeedMentionsproduct · LinkedIn
“led the effort to revamp the LinkedIn feed”
▶ Watch at ~1:00approx.
LinkedInMentionswebsite · LinkedIn
“toer is Chief product officer at LinkedIn”
▶ Watch at ~1:00approx.
Lenny's NewsletterMentionsnewsletter · Lenny Rachitsky
“drop in a link to your favorite Lenny's newsletter post”
▶ Watch at ~4:30approx.
GammaMentionssoftware · Gamma
“gamma an entirely new way to present your ideas powered by AI”
▶ Watch at ~4:30approx.
Google DocsMentionssoftware · Google
“gamma works with all types of formats from Google Docs PDFs to PowerPoint”
▶ Watch at ~4:30approx.
Gamma ProRecommendsproduct · Gamma
“use promo code Lenny to get a free month of gamma Pro”
▶ Watch at ~5:00approx.
PowerPointMentionssoftware · Microsoft
“setting its sights on becoming the modern alternative to PowerPoint”
▶ Watch at ~5:00approx.
WarrantMentionssoftware · Warrant
“work also recently acquired warrant the fine grain authorization service”
▶ Watch at ~5:30approx.
LoomUsessoftware · Loom
“hundreds of companies are already powered by work OS including ones you probably know like versel webflow and Loom”
▶ Watch at ~5:30approx.
VercelUsescompany · Guillermo Rauch
“hundreds of companies are already powered by work OS including ones you probably know like versel webflow and Loom”
▶ Watch at ~5:30approx.
WebflowUsessoftware · Webflow
“hundreds of companies are already powered by work OS including ones you probably know like versel webflow and Loom”
▶ Watch at ~5:30approx.
ZanzibarMentionssoftware · Google
“a groundbreaking authorization system called Zanzibar which was originally designed for Google”
▶ Watch at ~6:00approx.
WorkOSRecommendscompany · WorkOS, Inc.
“you should consider workos it's a dropin replacement for azero”
▶ Watch at ~6:00approx.
AirbnbMentionscompany · Brian Chesky, Joe Gebbia, and Nathan Blecharczyk
“chatting with Vlad who worked at Airbnb for many years”
▶ Watch at ~14:00approx.
TwitterMentionswebsite · Twitter
“most of my traffic to my newsletter comes from LinkedIn not Twitter”
▶ Watch at ~16:30approx.
LinkedIn PagesMentionsproduct · LinkedIn
“we had a few at LinkedIn pages is another one helping businesses build their presence on LinkedIn”
▶ Watch at ~18:00approx.
LinkedIn FeedMentionsproduct · LinkedIn
“the transformation to the LinkedIn feed was exactly like you said”
▶ Watch at ~19:00approx.
LinkedIn LearningMentionscourse · LinkedIn
“we B link that to make it LinkedIn learning a while back”
▶ Watch at ~19:00approx.
MetaMentionscompany
“we don't compete for volume like it's not we're not in the same kind of category of meta”
▶ Watch at ~24:00approx.
PlaidMentionscompany · Zach Perret and William Hockey
“think of merge like plaid but for everything B2B SS”
▶ Watch at ~34:30approx.
MergeMentionscompany · Merge
“the folks at merge are obsessed with Integrations their single API help SAS companies launch over 200 product Integrations”
▶ Watch at ~34:30approx.
DrataUsescompany
“organizations like ramp drada and electric use merge”
▶ Watch at ~35:00approx.
RampUsescompany · Eric Glyman and Karim Atiyeh
“organizations like ramp drada and electric use merge”
▶ Watch at ~35:00approx.
LinkedIn AI AcademyMentionscourse · LinkedIn
“we basically established an AI Academy every PM had to go through training”
▶ Watch at ~41:00approx.
Building OneCoinedpodcast · Tomer Cohen
“I buil that into a podcast”
▶ Watch at ~57:00approx.
PixarMentionscompany
“from the co-founder of Pixar at cat mall”
▶ Watch at ~57:30approx.
CanvaMentionscompany · Melanie Perkins, Cliff Obrecht, and Cameron Adams
“the CPO of canva or Spotify Roblox”
▶ Watch at ~57:30approx.link ↗
SpotifyMentionssoftware · Spotify
“the CPO of canva or Spotify Roblox”
▶ Watch at ~57:30approx.
RobloxMentionscompany
“the CPO of canva or Spotify Roblox”
▶ Watch at ~57:30approx.
Building OneRecommendspodcast · Tomer Cohen
“building one on Apple or Spotify it's it's short and it's really about showing you different disciplines”
▶ Watch at ~58:00approx.
MindsetRecommendsbook · Carol Dweck
“my fundamental books is one mindset it's about growth mindset”
▶ Watch at ~59:00approx.
Thinking, Fast and SlowRecommendsbook · Daniel Kahneman
“second book is thinking fast and slow um by Daniel canaman”
▶ Watch at ~59:30approx.
High Output ManagementRecommendsbook · Andy Grove
“lastly and on the fundamental side is high output management uh by Andy Grove”
▶ Watch at ~1:00:00approx.
BlueyRecommendstv show · Joe Brumm
“yesterday I saw Bluey”
▶ Watch at ~1:00:30approx.
BlueyUsestv show · Joe Brumm
“I can watch it with my six-year-old 9-year-old and 12year old and we're all going to enjoy it”
▶ Watch at ~1:01:00approx.
Spark AmpUsesproduct · Positive Grid
“there's a spark amp I have which allows me to kind of play with effects easily”
▶ Watch at ~1:02:00approx.
Ultimate GuitarUsessoftware
“I coupled that with the it's called the ultimate guitar app”
▶ Watch at ~1:02:30approx.
LinkedIn CoachMentionsproduct · LinkedIn
“we build this coach aspect where you can go to any job and you can Riff on the job”
▶ Watch at ~1:05:30approx.
Never Search AloneRecommendsbook · Phyl Terry
“there's this book called Never search alone that we're going to have the author on the podcast soon”
▶ Watch at ~1:06:30approx.
Friend.comRecommendsproduct
“check out friend.com it's like a digital friend that just is with you all the time”
▶ Watch at ~1:07:30approx.
LennyPodcast.comRecommendswebsite · Lenny Rachitsky
“you can find all past episodes or learn more about the show at Lenny podcast.com”
▶ Watch at ~1:08:30approx.
Apple PodcastsMentionssoftware · Apple Inc.
“you can subscribe to the show on Apple podcast Spotify or your favorite podcast app”
▶ Watch at ~1:08:30approx.

Spot an error or want something removed? Request a correction or removal.

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Hot Take· 4

Hot Take13:00

Don't Set Goals From Today's Numbers — Start From the Potential

Cohen sets ambitious goals by working backwards from what a product could be, not from where its metrics sit today. He uses a mountain metaphor: you can see the peak and base camp clearly even if the middle is blurry. Without that ambition, he argues, you'll never make the climb.

He starts from the peak and works backwards, not from today's existence
For LinkedIn's feed, he imagined millions of daily users when the numbers made no sense
The vision isn't detached from reality, but it isn't hooked to it either
Set the goal to something massive, then aim to overdeliver

“so I don't start from the existence to set my ambition I start from like what this could be like based on really inspiration and…”
Tomer Cohen · 15:00

#ambition#goal-setting#product-strategy#vision

Hot Take37:00

If AI Steers Your Product, Why Isn't the Product Leader Holding the Pedals?

Cohen uses a river-rafting analogy: the crew paddles for accuracy and speed, but the guide in the back holds the two pedals that actually steer the boat — and those pedals are AI. In most companies the guide was someone other than the product leader. He argues PMs delegate AI as black-box magic when they should own the algorithm's objective, features, data collection, and even infrastructure.

AI is the steering, not just the speed, of the boat
In most companies the AI 'guide' was not the product leader
PMs should be able to write the algorithm's objective on a whiteboard
Infrastructure and fine-tuning can be a bigger lever than any new button

“those two pedals navigate pretty much the both and those pedals are Ai and the guy better be you”
Tomer Cohen · 37:00

#ai#product-management#leadership#algorithms

Hot Take43:30

With AI You Control the Ingredients, Not the Experience

One of the biggest shifts in becoming an AI-first leader, Cohen says, is realizing you no longer dictate every step of the experience. It's like being a chef who used to control everything down to the temperature of the broccoli, and now hands over the ingredients and guidelines and lets the technology cook. Because AI isn't deterministic, giving it rope to learn — with safety guards in place — ultimately produces a much better experience.

Product leaders are used to dictating every step of the experience
AI-first means you control the ingredients and guidelines, not the outcome
It's a scary loss of control for many product people
AI is not deterministic, so you must give it rope to learn

“this new technology comes in and say just give me the ingredients give me the you know the the guidelines of how you cook and…”
Tomer Cohen · 44:00

#ai#product-management#leadership#design

Hot Take1:03:30

'Becoming Is Better Than Being'

Cohen's favorite life motto, rooted in growth mindset, is that becoming is better than being. The moment you think you've achieved something is the moment you start to deteriorate. He applies it to product too: the moment you think you've mastered it is the moment you become obsolete.

Achievement is the point where deterioration begins
Growth, learning and evolving are the real goals
In product, thinking you've mastered it means you're becoming obsolete

“becoming is better than being it's like the moment you think you kind of achieved something is the moment you start to deteriorate down”
Tomer Cohen · 1:03:30

#growth-mindset#philosophy#motivation

Explainer· 2

Explainer17:00

The Hardest Products Aren't Zero-to-One — They're Minus-One-to-One

Everyone talks about zero-to-one and scaling, but Cohen says turnaround products (minus-one-to-one) get little attention and are often harder because of entrenched internal flows, processes, and metrics. LinkedIn's feed started as an activity tracker that drifted promotional; fixing it meant changing the inner workings of the system and resetting the feed's purpose to genuine professional knowledge exchange.

Minus-one-to-one turnarounds are harder internally than external perception suggests
Entrenched flows, processes and metrics are the real obstacle
LinkedIn was first with a social feed but started wrong, as an activity/tracking feed
The new purpose: not a springboard or upsell feed, but knowledge exchange among people who matter

“there's so much conversation about zero to one products or scaling products but you don't have much conversations about minus one to run products like…”
Tomer Cohen · 17:30

“it's about knowledge exchange it's about how can I get the right views to the right uh experts in a way that actually helps them…”
Tomer Cohen · 20:30

#product-strategy#turnaround#linkedin#feed

Explainer31:00

AI as the Ultimate Matchmaker in a Two-Sided Marketplace

Cohen's biggest lesson for the feed was that in a marketplace, satisfying the need on the other side is where the magic happens. AI, he says, is the ultimate matchmaker — underutilized and misunderstood — and doing value exchange well is what brings people back and makes them spend more time. That conviction is why AI became core to his role.

In a marketplace it's all about value exchange between both sides
Do the match well and people come back, engage more, and spend more time
AI was underutilized, misunderstood, and run separately from the product team
This is why AI-first became the engine that moves the whole organization

“AI is the ultimate Matchmaker it's underutilized it's un misunderstood it's run separate from the team”
Tomer Cohen · 31:30

#ai#marketplace#matchmaking#product-strategy

Story· 3

Story22:00

How Cohen Carved Out 2 Million Users to Rebuild the Feed

Mass feed experiments kept scaring the whole company because every team relied on the feed for traffic and their numbers swung wildly. So Cohen carved out two million members as an isolated cohort where he had full liberty to build the new experience without hurting company-wide metrics. Over months that cohort showed dramatic behavior change, giving him the evidence to roll it out to everyone.

Mass experiments spooked teams whose goals depended on feed traffic
He isolated 2M members as 'his' cohort, removed from the metrics that mattered
Behavior change took months, not days or weeks
The randomized cohort's success proved the new feed could grow the whole pie

“so what I did was I carved out two million members and I said those are my members I'm going to focus on building”
Tomer Cohen · 22:30

“unless the CEO says I don't care about you know how the company performs for the next two years we're going to go for it”
Tomer Cohen · 27:30

#experimentation#product-strategy#linkedin#metrics

Story33:00

The Professor Who Advised Prime Ministers After Posting on LinkedIn

To illustrate value exchange, Cohen shares a professor who resisted social media but started posting daily on LinkedIn. The reach transformed his career: speaking engagements worth roughly half his annual university salary, and an invitation to advise prime ministers on national investment strategy. After 20 years of teaching, the platform elevated his ability to reach and influence people.

The professor posted daily years of accumulated content on LinkedIn
He earned speaking gigs worth about half his yearly university salary
He was invited to advise prime ministers on investment strategy
Getting the right content to the right people is the matchmaking at scale

“I was invited to advise prime ministers on their investment uh strategy for the country and it's like and they and I've been teaching for…”
Tomer Cohen · 34:00

#linkedin#content#value-exchange#creators

Story1:04:30

The Job Seeker Who Cried — and Why LinkedIn Built an AI Coach

In a session with job seekers, one person cried midway through, saying they felt alone and couldn't afford a coach. That loneliness drove LinkedIn to build an AI 'coach' experience where seekers can riff on any job with a private, personalized LLM — covering their fit, how to apply, and comparisons — without judgment. Cohen measures the impact by emotion: seekers came away feeling supported.

Job seeking is a lonely journey; one seeker cried describing the isolation
The coach is a private, personalized LLM you can riff with on any job
It covers fit, how to best apply, and comparing opportunities
Cohen measures the work's impact by emotion — the sense of feeling supported

“one started crying Midway into the session because they said I cannot share my journey with anybody because I feel like I'm alone in this”
Tomer Cohen · 1:05:00

#ai#linkedin#job-search#product

Tool· 1

Tool58:30

Tomer Cohen's Three Fundamental Books for Product People

Cohen recommends three fundamentals: Mindset by Carol Dweck (who happened to be his wife's manager) for the growth-mindset belief that skills and abilities can be developed; Thinking, Fast and Slow by Daniel Kahneman as the bible of behavioral economics for anyone building front-end products; and High Output Management by Andy Grove for the basics of good management.

Mindset by Carol Dweck — skills and abilities can be developed over time
Thinking, Fast and Slow by Daniel Kahneman — every page makes you stop and think
High Output Management by Andy Grove — the basics that take a manager to a B
He favors fundamentals, especially for people starting their careers

“after you read this book your managerial skills should start from a B and then you can you know over time become an A”
Tomer Cohen · 1:00:00

#books#recommendations#management#learning

Takeaway· 4

Takeaway02:30

Why 'We Might Be Wrong But We're Not Confused' Beats Being Right

Tomer Cohen's most-quoted mantra is about prizing clarity and alignment over the ego-driven need to be right. He traces it to a founder whose failing company only had a chance if everyone pulled in the same direction, and argues that hedging leaves you dependent on luck. A related habit: force people to name whether they're disagreeing or just misunderstanding, so meetings don't waste time.

Alpha personalities get attached to being right, which creates lingering confusion in a team
Getting attached to clarity and focus matters more than being right
Everyone hedging in different directions means only luck can save the system
A mentor taught him to ask people: are you disagreeing or misunderstanding?

“I would rather go forward with everybody in the same direction then necessarily try to hedge all the time which will give me no chance…”
Tomer Cohen · 07:00

“are you disagreeing or misunderstanding if you're misunderstanding let's spend the night let's get to a point where you can articulate uh my point of…”
Tomer Cohen · 09:00

#leadership#decision-making#clarity#product

Takeaway09:30

Your Real Priority Is Whatever Your Best People Are Working On

Cohen says organizations routinely decide on a top priority and then never act on it. The tell is resourcing: if most of your engineers are on a migration, the migration is your real priority, not the initiative you named. He pushes leaders to check that their stated priority is actually manifested in where their top talent sits.

Many organizations reach a decision but never act on it
A stated priority must be manifested in your resourcing to be real
If top talent is on moonshots, those moonshots are your true priority
You solve a lot just by making sure focus is actually there

“you have to like make sure that what you're sharing as a priority is actually manifested in your resourcing then I'm like hey this is…”
Tomer Cohen · 10:00

#execution#prioritization#management

Takeaway45:30

Diverge, Then Converge: How LinkedIn Avoided Shipping Bad AI

When ChatGPT-era AI arrived in Fall 2022, Cohen told leaders to drop their roadmaps and go back to the problems they were solving. He deliberately let teams diverge — building duplicates, going crazy, writing the prompt-engineering playbook — to maximize learning. Then, several weeks later, he converged resourcing top-down onto the four or five biggest bets, reviewing only those weekly.

Start from the objective you were solving, not 'we have cool tech, what can we do'
Allow teams to diverge and experiment to inspire creativity and learning
Capacity and cost are constraints, so converge after the exploration phase
He reviewed only the five key bets weekly to signal what mattered

“we really much allowed people to uh I would say many way diverge but then I would say like several weeks after converge”
Tomer Cohen · 46:30

#ai#innovation#product-strategy#experimentation

Takeaway49:30

Chase Conviction and Impact, Not Prestige and Logos

Cohen's career advice is deeply personal: early on he chose roles by what was most in demand and challenging, dictated by society rather than himself. After moving to the US he flipped 180 degrees to follow where he had strong conviction and could make a dent and learn. He argues product people are fortunate to be measured by the products they built and the impact they had — not their titles or the logos on their resume.

He shifted from 'what's most in demand' to 'where do I have conviction'
He didn't apply to LinkedIn — he pitched how mobile should be built and was told to build it
It's hard to be a strong product leader without genuine conviction
Product people are judged on impact, not company names or titles

“it was really pursuing the conviction I had uh and my excitement”
Tomer Cohen · 53:00

“I don't care about the companies you worked at I don't care about the logos I don't care about the titles”
Tomer Cohen · 54:30

#career#conviction#product-management#advice

Key ideas

Frameworks in this episode

AI-First Product Leadership: Hold the Paddles

Backwards From the Peak: Ambition Set by Potential, Not by Numbers

Conviction-Over-Demand Career Navigation

Diverge-Then-Converge Roadmap Reset

The Carve-Out Cohort Turnaround

We Might Be Wrong, But We Are Not Confused

People & resources mentioned

People · 14

Resources · 43

Insights & moments

◆Hot Take· 4

Don't Set Goals From Today's Numbers — Start From the Potential

If AI Steers Your Product, Why Isn't the Product Leader Holding the Pedals?

With AI You Control the Ingredients, Not the Experience

'Becoming Is Better Than Being'

✶Explainer· 2

The Hardest Products Aren't Zero-to-One — They're Minus-One-to-One

AI as the Ultimate Matchmaker in a Two-Sided Marketplace

❝Story· 3

How Cohen Carved Out 2 Million Users to Rebuild the Feed

The Professor Who Advised Prime Ministers After Posting on LinkedIn

The Job Seeker Who Cried — and Why LinkedIn Built an AI Coach

✦Tool· 1

Tomer Cohen's Three Fundamental Books for Product People

▲Takeaway· 4

Why 'We Might Be Wrong But We're Not Confused' Beats Being Right

Your Real Priority Is Whatever Your Best People Are Working On

Diverge, Then Converge: How LinkedIn Avoided Shipping Bad AI

Chase Conviction and Impact, Not Prestige and Logos

Hot Take· 4

Explainer· 2

Story· 3

Tool· 1

Takeaway· 4