LLenny's Podcast
← All episodes
Tamar Yehoshua (Product at Glean, ex-Google and Slack)26 September 2024

Lessons in product leadership and AI strategy from Glean, Google, Amazon, and Slack

7Frameworks
16Insights

Frameworks in this episode

Insights & moments

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

Myth Buster· 1

Myth Buster45:30

Stop Over-Indexing on the Users Who'll Be Unhappy

PMs often over-worry about the vocal minority upset by a launch or a removed feature, but today's users are usually far fewer than tomorrow's, so design for the bigger future audience. The key is doing it respectfully: be transparent, authentic, and actually listen. Tamar illustrates with Google PR chief Rachel Whetstone, who, instead of sending an FAQ, phoned an upset individual engineer directly, listened, and completely changed his mind.

  • PMs get trapped by the vocal minority when removing or redesigning features
  • Design for the larger number of users who'll use the product tomorrow, not just today's
  • Be respectful, transparent, and authentic, with real reasons, not marketing speak
  • Rachel Whetstone phoned an upset engineer rather than sending an FAQ, and won him over

the number of people using your product today is usually unless you're at Google far smaller than the number of people are going to be…

Tamar Yehoshua · 46:30

she did not respond to my email she picked up the phone and she called him

Tamar Yehoshua · 48:00
#product-management#user-research#leadership

Hot Take· 4

Hot Take11:30

You Don't Have to Be a Well-Run Company to Win

Tamar prefers well-run companies but insists being well-run isn't correlated with success. She's seen companies with high executive turnover, people getting yelled at, constant firings, and miserable staff whose numbers are still growing like crazy, and the reverse: an amazing, well-oiled company that flatlined. What matters most is a product people value plus distribution, a working sales team, and enough money in the bank.

  • Broken marketing, HR gaps, and turnover don't reliably predict failure
  • She's seen miserable, chaotic companies grow like crazy and well-run ones flatline
  • What actually matters: product-market fit, distribution, sales, and runway

there's like high executive turnover where people get yelled at there are lots of people fired there's re all the time people I talk to…

Tamar Yehoshua · 12:00

amazing CEO well-run well oiled machine everything hired good Executives and they flatlined

Tamar Yehoshua · 12:30
#startups#product-market-fit#leadership
Hot Take18:30

You Don't Need a Five-Year Plan

Tamar has never had a five-year plan and still doesn't know what she wants to do in five years. She tells anxious younger people that planning is fine if that's who you are, but it isn't required, and watching a stressed 20-something relax at hearing this convinced her it's worth saying out loud. Early-career forks feel scarier because MBA-vs-job or PM-vs-engineer choices lead down very different paths.

  • She has never had, and still doesn't have, a five-year plan
  • Planning suits some personalities but isn't a requirement for success
  • Early-career forks feel higher-stakes because they diverge sharply

I said I Never Had A Five-Year Plan

Tamar Yehoshua · 19:00

I still have no idea what I want to do in five years I've never had an idea what I want to do in five…

Tamar Yehoshua · 19:30
#career#advice#mindset
Hot Take50:00

AI Will Blur the Lines Between PMs, Engineers, and Designers

Tamar, who holds a master's in AI, believes we're underestimating how much AI will change work, not overnight, but the people who figure out how to leverage it will pull far ahead. In five to ten years she expects the lines between PMs, engineers, and designers to blur as AI lets each build prototypes and designs. Tools like Figma AI and Copilot aren't fully there yet, but the dangerous move is trying once, failing, and reverting to old ways.

  • We're underestimating how much AI will change how we work, though not suddenly
  • Those who learn to leverage AI will work faster and pull far ahead
  • In 5-10 years the PM/engineer/designer roles will blur
  • Trying AI once and reverting to the old way is a dangerous spot to be in

we are underestimating how much it's going to change how we work

Tamar Yehoshua · 51:00

in 5 to 10 years I think the lines between product managers and engineers and designers are going to blur

Tamar Yehoshua · 51:30
#ai#product-management#future-of-work
Hot Take59:00

The One Thing AI Can't Do, and Why It Sorts Good PMs From Bad

Tamar argues AI's weakness is creativity: it can tell you what customers are asking for and what today's problems are, but not the leap of how to solve them differentiatedly. So AI threatens PMs doing rote execution and project management while rewarding strategic, creative ones. It won't necessarily mean fewer people, since companies always want to build more, but the not-so-good PMs' jobs go away while the great PMs keep great jobs.

  • AI's one weakness is creativity; it won't give you the differentiating leap
  • Rote-execution PMs are threatened; strategic, creative PMs are differentiated
  • AI should automate drudgery like updating Jira and building launch calendars
  • Not-so-good PMs' jobs go away; great PMs keep great jobs; PMs/designers will blend

I think that AI the one thing it's not good at is being creative

Tamar Yehoshua · 1:00:00

the not so good PMS jobs will go away the great PMS will still have great jobs

Tamar Yehoshua · 1:01:30
#ai#product-management#careers

Explainer· 2

Explainer15:30

Why Hypergrowth Companies Feel Like Chaos (And When It Isn't OK)

When a company hits product-market fit it has to grow fast or growth stalls, and that speed inevitably breaks infrastructure and communication. At any moment half the company may have been there under six months. Tamar notes MySpace died because its product got too slow. Some chaos is normal until scale, but constantly changing strategy, direction, and people's projects so nothing gets finished is the bad kind.

  • Post-PMF, you must grow fast or growth stalls, which breaks things
  • At any point 50% of a hypergrowth company may have under six months tenure
  • Bad chaos = constantly changing strategy and projects so no one can achieve anything
  • Once past ~5,000-10,000 people you need to become well-run

you've got at any given point 50% of the company has been there less than six months

Tamar Yehoshua · 16:30

look at companies like Myspace well they died because their product got too slow

Tamar Yehoshua · 17:00
#hypergrowth#startups#scaling
Explainer38:00

Building Strong PM-Engineering Relationships: Don't Ask Mom and Dad

Tamar calls the PM-engineering partnership the most important thing a product leader manages, and says vetting your engineering partner should be part of evaluating a job. Alignment means clear roles so people can't play the PM against the CTO by asking mom and dad for different answers. With Slack's Cal Henderson she never went around him on his domain. The foundation is respect and trust that people follow up on what they say.

  • Evaluate your engineering partner before joining; it's core to the role
  • Clear divide-and-conquer prevents people playing PM against CTO
  • She routed questions in Cal's domain back to Cal rather than going around him
  • The bottom line is respect and trust that people deliver what they commit to

you don't want any of this like people in the organization they asked Mom they asked Dad and they got different opinions and playing one…

Tamar Yehoshua · 39:00

you have to respect and trust that they actually will follow up on what they say they will

Tamar Yehoshua · 40:00
#product-management#cross-functional#leadership

Story· 5

Story06:00

Reading the Room: What a Psychiatrist Father Taught Her About Users

Tamar argues that building products and leading teams both come down to understanding people's motivations: why they'd click a button, join a team, or work hard. She credits her psychiatrist father, who after every family event would analyze in the car why people said and did what they did, and quiz her on it. It trained her to read the whole room and notice how people react.

  • Product and org building both hinge on understanding people's motivations
  • Ask lots of questions to understand what people are trying to accomplish
  • Her father's post-event analysis quizzes taught her to observe reactions across a room

my father was a psychiatrist and when I was growing up we would you have family occasions go to events whatever and afterwards in the…

Tamar Yehoshua · 07:00

you can understand so much if you're paying attention

Tamar Yehoshua · 07:30
#product-management#leadership#empathy
Story09:30

"This Team Shouldn't Exist": Why Impact Beats Hitting Your Goals

Doing a great job isn't hitting the goals you were assigned, it's driving real impact for the business. Early in her career, offered her first management role, Tamar told her stunned manager the team shouldn't exist because it wasn't the right thing for the company, and explained how it should be reorganized instead. Her manager agreed and found her another role. She distills it to one word: impact.

  • A great job means building something people actually use, not just launching
  • Ask whether you enabled the whole organization to be more productive
  • She turned down her first management job by arguing the team shouldn't exist

here's why I don't think that this team should exist it's not the right thing for the company it's not going to be productive

Tamar Yehoshua · 10:30
#career#leadership#impact
Story26:30

Lessons From Jeff Bezos: Speak Last, Stay Consistent, Fewer People Is an Advantage

From early quarterly meetings with Bezos before AWS launched, Tamar took away several things: in six-pager meetings he speaks last, polling everyone around the table first so decisions are a team effort. His consistency, always customer-driven, hating unlabeled icons, made it easy to operate because you knew what he'd ask. And when she noted competitors had 10x the headcount, he told her that was her advantage, framing product-building as a seven-year hill.

  • In six-pager meetings Bezos speaks last after hearing every leader
  • Consistent principles (customer-driven, no unlabeled icons) made the org easier to run
  • He reframed having fewer people than competitors as an advantage
  • He treated products as a seven-year hill requiring long-term thinking

first he does not speak until everybody around the table speaks so he goes around to all his leads

Tamar Yehoshua · 28:00

he looks at me and he said that is your advantage

Tamar Yehoshua · 29:30
#leadership#amazon#bezos
Story30:00

Stuart Butterfield and the Power of Prototyping ("Put It Behind One Button")

Tamar joined Slack to work with Stuart Butterfield, whom she calls the best product thinker in the valley, praising his mix of long-term vision (a four-box 2014 master plan that never changed) and obsession with details. His biggest lesson was prototyping: a mockup can't tell you how something feels. In one redesign he told the team to put everything in the interface behind a single button, not to ship it, but to reveal what the interface truly needed.

  • Butterfield's 2014 four-box master plan for Slack never changed, though yearly work did
  • A mockup can't tell you how a product will feel; you have to prototype and try it
  • He had the team hide the entire interface behind one button to find what was essential
  • Prototyping requires engineers willing to write throwaway code to move faster

I can't tell you if this is going to work I have to feel it I have to try it and a mockup doesn't tell…

Tamar Yehoshua · 31:00

I want you to take everything in the interface and put it behind one button and everyone's like that's never going to work and he's…

Tamar Yehoshua · 32:00
#product-management#prototyping#slack
Story52:30

The Discord-Into-Gemini Trick to Avoid Falling Behind on AI

Tamar's advice for not falling behind is simply to use the products, as good PMs always have with each new wave. She shares a forward-thinking PM who fed an entire, huge Discord channel transcript into Gemini after its context window expanded, then asked for product sentiment, most-requested features, and complaints, calling it a gold mine of analysis no one would ever have done by hand. She adds Glean examples like summarizing Gong sales-call transcripts.

  • Best defense against falling behind: constantly use new AI products, as with mobile before
  • A PM fed a whole Discord channel into Gemini to extract sentiment and feature requests
  • The analysis was a gold mine that would never have been done manually
  • At Glean, an app reads Gong call transcripts to summarize top requested features

he took the transcript from the Discord Channel which was huge right and he fed it into Gemini the entire Channel and then used it…

Tamar Yehoshua · 53:30

it was like a gold mine

Tamar Yehoshua · 54:00
#ai#product-management#productivity

Takeaway· 4

Takeaway02:30

The Most Overlooked Career Move: Nail the Job You Already Have

Tamar's top career lesson is that ambitious people, especially product managers, over-focus on the next role instead of excelling at their current one. You don't earn the next job until you knock the current one out of the park, however simple it seems. Table stakes in tech roles include being technical, knowing the latest technology, understanding your product deeply, and understanding metrics.

  • Ambitious PMs tend to eye the next level instead of mastering their current role
  • You won't get promoted unless you do the job you're in exceptionally well
  • Table stakes: be technical, know the latest tech, deeply understand your product and metrics

you're not going to get the next job unless you do really well at the job that you're in like knock it out of the…

Tamar Yehoshua · 02:30
#career#product-management#promotion
Takeaway20:00

Follow the Best People, Not Domains or Brand-Name Companies

Tamar's career heuristic is to follow the best people rather than domains: work for the best product thinker, engineer, or salesperson so you learn to be the best at it, and go where great people cluster (like the PayPal Mafia or ex-Googlers). As a leader, take jobs where hiring someone will make their career. You don't need a top brand like OpenAI or Glean, because skills, unlike companies, can never be taken away.

  • Follow the best product thinker, engineer, or salesperson, not a domain
  • Go where great people cluster; the relationships outlast any single company
  • As a leader, take roles where hiring people will make their careers
  • You don't need a top-brand company; skills can't be taken away when one fails

you want to follow somebody who's either the best product thinker or the best engineer or the best salesperson

Tamar Yehoshua · 20:30

skills can't be taken away a company can fail but if you learn a skill you will always have that skill

Tamar Yehoshua · 26:00
#career#hiring#learning
Takeaway1:02:00

Your Product Should Get Better as the LLMs Get Better

Building on non-deterministic LLMs, Tamar warns that too many teams build scaffolding to compensate for current model limitations, and all that work will evaporate as models improve. It's fine to do temporarily, but it can't be your differentiator. Your real differentiator must be something outside the LLM that keeps you ahead as models get smarter, so the whole product gets better, not worse, as the underlying models improve.

  • Much effort spent compensating for LLM weaknesses will be thrown away as models improve
  • Temporary workarounds are fine but can't be your differentiator
  • Your differentiator must live outside the LLM and persist as models get smarter
  • The goal: your whole product improves as the underlying LLMs improve

too many people are building things to make up and compensate for the llms that all that work is going to go away

Tamar Yehoshua · 1:05:30

your differentiator is something that will continue to be there as the llms get better and smarter

Tamar Yehoshua · 1:06:00
#ai#product-strategy#llms
Takeaway1:13:30

"There Are No Right Decisions, You Make a Decision Right"

Tamar's favorite life motto came from her father when she was agonizing over which college to attend: there are no right decisions, you make a decision right. Because you can never know how things will turn out, you commit fully and don't look back with regret. Success comes from how you approach the decisions you've made, so pick a path, make it the best you can, and move forward.

  • The motto came from her father when she was indecisive about college
  • You can't know the outcome, so commit and don't second-guess years later
  • Success is made by how you approach a decision, not by picking the 'right' one

there are no right decisions you make a decision right

Tamar Yehoshua · 1:13:30
#mindset#decision-making#advice