The PodKnowledge|← All Episodes
Based on Lenny's Podcast data
Lenny's Knowledge Sketch

Might Be Wrong,
Not Confused

Tomer Cohen
Chief Product Officer, LinkedIn · product, design, business dev, content, customer ops
DEC 4 2025
The Turnaround

The LinkedIn Feed
Nobody Believed In

FEED, PROMOTIONAL → DAILY HABIT
"I don't start building from the numbers today. I start from what this could be, based on inspiration and excitement."
  • The feed used to be a promotional graveyard; even Lenny doubted it could become a place people wanted to visit
  • Tomer ran the effort to revamp it, and moved LinkedIn to an AI-first mindset before it was cool
  • His starting point: how many professionals exist, how many would want to share and engage — not today's number
  • LinkedIn is a 20+ year-old company; every phase has been underestimated, then quietly gotten better
Framework

Two Halves of Clarity

CLARITY OF THOUGHT"Are we disagreeingor misunderstanding?"CLARITY OF EXECUTION"If the migration is whatengineers do, it IS #1"
"We might be wrong, but we're not confused. That's actually really liberating."
  • Alpha teams get attached to being right; Tomer gets attached to clarity and to moving in one direction
  • The mask: people in the US say "I don't understand" when they mean "I disagree." Push through to which one it is
  • The execution test: your #1 priority is whatever most of your engineers actually work on; anything else is theater
  • Rooted in a founder story: a company hedged its way to death because the team kept quietly betting on other paths
The everyday version"I'm willing to sacrifice or trade off X" is a real opinion. "We should build a simple product" is not. Force the trade-off out of hiding.
The Approach

Start from Potential,
Not From Today

  • Set the goal from what the product could be, based on inspiration — not the metric baseline you can defend today
  • Then walk it backward into something the team can actually pursue this quarter
  • The feed was "cringey and promotional"; Tomer set the goal that millions would want to visit daily
  • LinkedIn has 20+ years of skeptics; the founding move was ignoring them each phase
The Chesky reference

Tomer's former manager at Airbnb watched Brian Chesky set 10x versions of a goal to figure out what it would actually take. Tomer runs the same play at LinkedIn.

One billion members is small

"If you play from that, one billion members is actually pretty small compared to where we can go." The framing is what unlocks the next act.

Playbook

How to Run a Team the Tomer Way

  • Spend real time on the problem statement, not the solution; make the trade-off explicit
  • Push people to name whether they are disagreeing or misunderstanding, then handle each properly
  • Watch what engineers actually do, not what the doc says; the calendar is the priority
  • Set the goal from potential, not the number you have; then start the walk-back to today
Where "not confused" came fromA founder told Tomer, years ago, that their company was on the brink until everyone stopped hedging in different directions. Pulling in the same direction is not a guarantee of success; hedging is a guarantee of failure.
Contrarian

Tomer's Uncomfortable Takes

✗Being right is the pointINSTEAD →✓ Being aligned is the point. Attachment to being right creates confusion; attachment to clarity creates movement.
✗"We might just be misunderstanding each other"INSTEAD →✓ Push through to which one it is. If disagreeing, stop and decide. Misunderstanding hidden as disagreement wastes everyone's time.
✗Set goals from your current baselineINSTEAD →✓ Set them from what could exist. The feed becoming a daily place for millions did not make sense from the numbers of the time.
✗"Build a simple product" is a strategyINSTEAD →✓ Everyone wants a simple product. Only "I'm willing to trade off X" is a real opinion. Force the trade-off.
Based on Tomer Cohen's episode on Lenny's Podcast. All ideas on this page are from the episode.Watch on YouTube
GO DEEPER IN THE EPISODE
04:42 The need for change in product development11:52 The full-stack builder model explained16:03 Implementing AI and automation in product development19:17 Building and customizing AI tools27:51 The timeline to launch31:46 Pilot program and early results37:04 Feedback from top talent39:48 Change management and adoption46:53 Encouraging people to play with AI tools41:21 Performance reviews and full-stack builders48:00 Challenges and specialization50:05 Finding talent52:46 Tips for implementing in your own company56:43 Lightning round and final thoughts
Every sketch here began as a transcript.

Send us yours and we’ll draw it. $1 a transcript.

Sketch mine →
𝕏︎ X / Twitterin LinkedIn📸 Instagram🔗 Copy link
0:00