LLenny's Podcast
← All episodes
Nan Yu (Head of Product)30 January 2025

Linear’s secret to building beloved B2B products

7Frameworks
15Insights

Frameworks in this episode

Leadership3 steps

Deadlines as P-Zero

Have few deadlines, but make each one a P0 that nothing else can touch — and cut scope to hit it

Read →
Communication3 steps

Emotional Discovery: Feel Bad Like the Customer

On customer calls, dig until you feel the exact emotion the customer wants to avoid, then build to remove it

Read →
Influence4 steps

Interview as Discovery: Be the Solution to Their Burning Problem

Treat a job interview like product discovery — find the hiring manager's burning problem and position yourself as its solution

Read →
Innovation4 steps

Systematized Creativity by Extremes

Build the most extreme version along one attribute, feel it, repeat in the opposite direction, then find where they meet

Read →
Productivity4 steps

The 10% Working-Version Rule

Have a testable working version by the time 10% of your time budget elapses, not at the halfway point

Read →
Marketing3 steps

The Double Triangle: PM as Go-To-Market Discipline

Sit the PM between two triangles — build (eng/product/design) and sell (sales/marketing/PM) — and originate the customer's language

Read →
Strategy3 steps

The Middle-Manager Reporting Veto

Automatically say no to customization features that ease manager reporting at the cost of IC workflow

Read →

Insights & moments

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

Hot Take· 5

Hot Take08:00

Why "Speed vs Quality" Is a False Tradeoff

Nan argues people treat speed and quality as opposed only because they equate speed with rushing or sloppiness, when they should equate it with competence. Experts at the peak of their craft go fast precisely because the work is second nature, and in software, speed enables more iterations, which is what actually drives quality. The trick is holding the first version loosely, so no one wastes time perfecting something that may be wrong.

  • People over-index on 'rushing/sloppy' when they think about speed instead of 'competent/expert'.
  • You can predict a craftsperson's output quality by how fast they move.
  • In software, speed means more iterations, and iteration count drives how good the product gets.
  • The first version is a best guess, not a masterpiece, so no one needs to be a perfectionist.

“you can basically tell how good the output is going to be of their work product by how fast they're going”

Nan Yu · 08:30

“the first version of it is our best guess in the general direction of what we want to actually ship in the end”

Nan Yu · 11:00
#product#speed#quality#craft
Hot Take16:30

The One Feature Category Linear Always Says No To

Linear refuses one specific category of request without debate: customization features asked for by middle managers to make reporting easier, at the cost of making the IC workflow worse. Nan explains this is exactly what leads to the bloat that makes ICs hate enterprise software. When ICs disengage from tedious required fields, the reporting data becomes wrong anyway, so the tradeoff is false for everyone.

  • The non-negotiable 'no' is customization for middle-manager reporting that degrades IC workflow.
  • This category is the exact thing that produces enterprise-software bloat.
  • ICs are paid for their real job, not filling tickets, so they disengage or fill fields at random.
  • Once ICs fill required fields randomly, the manager's reporting data is wrong regardless.
  • Holding this line is a core promise of Linear, not a temporary stance.

“customization features requested by middle managers in order to make reporting a little bit easier at the cost of making IC workflow is worse”

Nan Yu · 16:30

“there's nine choices I don't know what any of them meet so I'm just going to pick one at random”

Nan Yu · 18:30
#product#b2b#bloat#prioritization
Hot Take54:00

B2B Software Doesn't Just Solve Problems, It Teaches You How to Work

Nan's view is that adopting B2B software means adopting a practice, not just a tool. Tools usually start as an internal process someone found valuable, then get spun out, so buying one is buying into a way of working with baked-in best practices. The starkest example is ERP: companies redo their entire internal processes to conform to the tool because they trust it encodes a battle-tested way to operate.

  • Most B2B tools originate as an internal process that worked and got productized.
  • Adopting the tool means adopting the idea that its practice is worth doing at all.
  • Buyers import a baseline level of competency, whatever the tool's defaults are.
  • ERP is the extreme case: firms rebuild their processes to conform to the tool's best practices.

“when you adopt that tool you're not just adopting like the the actual software you're adopting the idea that this is a practice that you…”

Nan Yu · 54:30
#b2b#product#workflow#adoption
Hot Take1:09:30

Deadlines Are Fabrications, So Have Few and Make Them P-Zero

Nan calls deadlines complete fabrications, and the only way to make one real is to treat it as a P-zero problem where nothing else matters. Have few deadlines, and when you commit, the engineers on it can't be pulled onto anything else, and the PM's job is to cut as much scope as possible so you reach a real product you can say yes or no to. He also does almost no estimating; shipping early is what makes deadlines hittable.

  • Deadlines are fabrications, made real only by treating them as P-zero.
  • Have few deadlines; when you set one, engineers on it are off-limits for other work.
  • The PM's job is to cut scope so you have a real product to make the go/no-go call on.
  • Linear does almost no estimating; shipping early is the mechanism for hitting deadlines.
  • Deadlines are usually worth it only around external marketing moments, and missed windows don't come back.

“deadlines they're just they're they're complete Fabrications”

Nan Yu · 1:09:30

“we do almost no estimating in order to hit deadlines”

Nan Yu · 1:13:00
#product#deadlines#scope#execution
Hot Take1:17:00

"The Correct Amount Is Too Much Minus One"

Nan's life motto ties directly to his 'build the extreme version' philosophy: to find the right amount of something, overshoot to the edge, discover what's too much, then back off by one. His pizza example, five slices felt bad, four was right, illustrates it. Lenny connects it to Elon Musk's cutting rule that if you don't add back 10% of what you cut, you didn't cut enough.

  • Find the right amount by deliberately overshooting to find what's too much.
  • This mirrors the product method of building the extreme version to locate the ideal.
  • Pizza example: five slices too many, four about right.
  • Related to Musk's rule: if you don't restore ~10% of what you cut, you didn't cut enough.

“the correct amount is too much minus one”

Nan Yu · 1:17:00
#mindset#decision-making#creativity

Explainer· 4

Explainer31:00

Real Empathy Is Feeling as Bad as the Customer Feels

Nan's goal on customer calls is to feel bad in the same way the customer feels, not just run a standard five-whys analysis. By having a long, deep conversation, he uncovers the emotional valence behind a request, the actual moment someone felt bad. In one case, a person marked a Dec 30 ship date, marketing panicked, and the pain led to Linear letting users set target dates at any granularity so they never feel forced into false precision.

  • The aim is to feel the customer's bad feeling, not just extract stated goals.
  • Five-whys can miss the emotional reason someone feels bad about a gap.
  • People won't tell you directly how they feel until you've built enough rapport.
  • The Dec 30 deadline pain led to granular target dates (month, quarter, half-year) in Linear.

“my goal is to feel bad in the same way that customers feel bad”

Nan Yu · 31:00

“so I don't ever want to put dates on things ever again”

Nan Yu · 32:30
#product#empathy#customer-research#emotion
Explainer34:30

Emotional Hooks and Paul Graham's "Schlep Blindness"

In crowded categories, the obvious goal-oriented opportunities are already picked, so Nan looks for emotional hooks, the places people feel bad during their workday without noticing. He ties this to Paul Graham's idea of schlep blindness: people are numb to the annoying grind around them. An outsider can spot the rhythm of a user's frustrations, like dreading Mondays because of a tedious report, and design that pain away.

  • In competitive industries the low-hanging goal-oriented ideas are already taken.
  • Look at things from an angle others haven't seen: where users feel bad daily.
  • PMs and engineers avoid the 'touchy-feely' emotional layer, which is the opening.
  • Paul Graham's 'schlep blindness': people are blind to the annoying grind of their own routine.

“you have to kind of look at things from uh an angle that other people might not have seen”

Nan Yu · 34:30

“Paul Graham has a word for this um he calls it he calls it schle blindness”

Nan Yu · 35:30
#product#emotion#differentiation#discovery
Explainer45:00

Systemized Creativity: Build the Most Extreme Version

Nan's method for non-obvious ideas is to ask how extreme you can push a product along some trait, echoing Brian Chesky's 11-star experience. Throwing away cost and practicality expands the search space so you can see beyond your defaults. The biggest risk in product isn't picking wrong among your options, it's never seeing the right option because it sat in a corner you never looked at.

  • Ask 'what's the most outrageous version of this?' along a specific trait.
  • Deliberately discard cost and practicality to explore the possibility space.
  • The goal is to break past unconscious constraints and see beyond your defaults.
  • The real risk is not seeing the right choice at all, not choosing wrong among the visible ones.

“what's the most outrageous version of this along some you know some some trait”

Nan Yu · 45:30

“the biggest risk is you didn't see the right choice to begin with”

Nan Yu · 46:30
#product#creativity#ideation#design
Explainer58:30

The PM "Double Triangle": Product Management as Go-To-Market

Beyond the usual engineering-product-design triad, Nan treats product management as partly a go-to-market discipline alongside sales and marketing. The PM sits between the building side and the selling side, using deep customer knowledge to help marketing pick the exact native language customers use, which sales then validates in the field. His advice to B2B PMs: the untapped impact is usually on the sell side, so lean into product marketing and the actual words customers read.

  • PM is partly a go-to-market discipline, not only the eng/product/design triad.
  • Visualize a double triangle: PM in the middle connecting builders to sellers.
  • PMs know the customer's native language best and should help originate marketing copy.
  • The impact you're leaving on the table is usually the sell side, not eng collaboration.

“we think of product management as as partially like a go to market discipline in the same way that sales and marketing are”

Nan Yu · 58:30
#product-management#go-to-market#marketing#collaboration

Story· 3

Story28:00

How Digging Behind "Custom Fields" Became a Real Feature

Users kept asking for fully customizable fields, which Linear resists because it burdens ICs. Digging in, Nan found ~40% wanted the fields to track which customer (e.g. Walmart) asked for what, so CSMs could show evidence of work. Instead of custom fields, Linear built Customer Requests that auto-pulls feedback from support tools and CRMs and tags escalations automatically, giving managers the reporting without making ICs do anything extra.

  • About 40% of custom-field requests were really about tracking which customer wanted a feature.
  • The manual workaround was tagging things by hand, copied from how they did it in spreadsheets.
  • Linear instead hooked into support tools and CRMs to auto-tag escalations with the requester.
  • Managers still get reporting, and ICs actually gain context on who asked and why.

“Walmart ask for this feature it's really important I need everyone to know that Walmart needs this”

Nan Yu · 29:00
#product#customer-research#feature-design#b2b
Story48:00

Designing Draft-Saving by Going to Both Extremes

To design draft-saving for issues, Linear first built the fastest possible version (close and it's gone, no popup), which felt too unsafe as predicted. Then they built the safest (autosave every keystroke), which left a paper trail of untitled junk. The shipped solution sits between: a brand-new issue interrupts you to save, while an existing draft autosaves in place. The path to the balanced answer was building both extremes and feeling them.

  • Fastest version: no save prompt, close and lose it, felt too unsafe (internal only).
  • Safest version: autosave every character, created clutter of untitled drafts.
  • Final design: interrupt on a new issue, autosave when editing an existing draft.
  • The route to a balanced choice is to go extreme in one direction, then the other, then meet in the middle.

“the path to get there is to do something totally extreme in One Direction and then totally extreme in another Direction and then find where…”

Nan Yu · 51:00
#product#design#iteration#ux
Story1:18:00

How Defective Men's T-Shirts Became Everlane's Bestselling Women's Tee

At Everlane, a batch of men's t-shirts came in about an inch and a half too short and couldn't be sold. To salvage the cash-strapped inventory, the design and marketing teams cropped them another two inches and marketed them to women as a boxy cropped silhouette. It sold out in a week and became a bestselling women's box-cut tee, now offered in around 20 colors, a reminder that product-market fit sometimes shows up in the weirdest way.

  • A batch of men's tees arrived ~1.5 inches too short and was unsellable.
  • The team cropped them further and repositioned them as a women's cropped boxy tee.
  • It sold out in about a week and became a lasting bestseller.
  • Now offered in ~20 colors; PMF can emerge from salvage and constraint, not a grand plan.

“there was a batch of um defective men's t-shirts they all came in like an inch and a half too short”

Nan Yu · 1:18:30

“it sold out in like a week and I like oh okay I guess we just made a hit product”

Nan Yu · 1:19:00
#story#product-market-fit#everlane#serendipity

Takeaway· 3

Takeaway09:30

Have Something Workable by 10% of the Time Budget

Rather than waiting until the halfway or 80% mark, Linear aims to have a working thing that tests a key hypothesis by the time just 10% of the time budget has elapsed. Getting a directional read early prevents building on a wrong assumption and blowing deadlines. Features then roll out through a widening circle of users, from internal to beta groups to GA.

  • By the time 10% of the estimated time has passed, you should have a workable solution that tests a key hypothesis.
  • Waiting until 80% done to judge whether it's panning out is too late.
  • Most wasted time is spent building things nobody ends up wanting.
  • Features graduate through a gradually increasing circle of users: internal, then beta groups, then GA.

“by the time 10% of it has passed you have a workable solution right”

Nan Yu · 09:30

“you don't want to wait till you're 80% done to be able to make that kind of judgment”

Nan Yu · 10:00
#product#iteration#shipping#validation
Takeaway24:30

Tie Every Feature Request to a Real Named Person

When debating what to build, Linear grounds decisions in specific real users, first name, last name, and email you could actually contact, not a hypothetical Alice or Bob. A beautiful, elegant solution is worthless if it doesn't match the messy reality of how a real person works. Tying decisions back to reality is central to how they discuss and prioritize.

  • Use real, contactable people to justify a feature, not made-up personas.
  • Reality is often ugly, so a beautiful solution that doesn't fit it doesn't matter.
  • A specific person's request proves at least one real need exists.
  • PMs fall into a trap of building for elegance instead of for how people actually work.

“here's the first name last name here's their email you can ask them”

Nan Yu · 25:00

“if you have a beautiful solution that doesn't match with reality it doesn't really matter”

Nan Yu · 25:30
#product#customer-research#prioritization
Takeaway1:04:30

Land the Job by Being the Solution to Their Burning Problem

Nan's job-hunting advice: hiring managers bring on a PM to solve some burning problem, so use the interview to do discovery and find that problem, then make the case that you are its solution. That turns hiring into a binary choice between hiring the fix and rolling the dice on someone else. Concretely, act like you already work there, ask about their OKRs this quarter and how you'd help hit them.

  • Figure out the hiring manager's actual job-to-be-done for this specific role.
  • Position yourself as the solution to that problem so the choice becomes binary.
  • Act as if you already work there and ask what you'd do about their real goals.
  • Invest deeply in the few roles where you can genuinely over-deliver.

“you are the solution to a problem and then everyone else is a you know sort of like a a rule of the dice”

Nan Yu · 1:06:00

“just act like you already worked there what would you do”

Nan Yu · 1:06:30
#career#job-search#interviewing#product-management