Lenny's PodcastCompeting with giants: An inside look at how The Browser Company builds product | Josh Miller (CEO)
CHAPTERS
- 0:00 – 4:41
Cold open: Why The Browser Company optimizes for feelings over graphs
Josh challenges Silicon Valley’s default obsession with metrics, arguing that focusing only on numbers misses what makes products beloved. He introduces the idea of explicitly designing for emotions like joy, speed, and focus as a practical product strategy—not just a romantic ideal.
- •Metrics are useful but can leave “a lot on the table” when over-optimized
- •Arc teams start from desired user feelings (joy, fast, organized, focused)
- •Reframing product goals around human experience, not dashboards
- 4:41 – 7:40
What Arc is—and the single metric they use to track growth (D5D7)
Lenny and Josh discuss the perennial challenge of describing Arc succinctly, while grounding the business in one core metric. Josh explains why D5D7 (using Arc at least five days a week) captures retention, engagement, and growth in one hard-to-game number.
- •Arc is a replacement for your default browser, but hard to summarize in one line
- •D5D7 as the north star: 5+ active days out of 7
- •Why D5D7 is harder to game and blends key signals
- •They focus on week-over-week growth rate more than absolute scale
- 7:40 – 9:47
Retention benchmarks: what Arc’s D5D7 numbers look like and how they think about cohorts
Josh shares Arc’s D5D7 retention range and emphasizes that the real goal is steady improvement cohort over cohort. He notes they’re proud not only of strong early retention, but that it has continued to inch upward even as they broaden beyond early enthusiasts.
- •Arc’s D5D7 retention is roughly in the low/mid-30% to low-40% range by cohort
- •They avoid DAU/WAU-style fixation and look at the single metric consistently
- •The key is retention improving over time across cohorts
- •Acknowledges retention will likely trend down eventually as adoption broadens
- 9:47 – 15:00
The anti-dogma product philosophy: feelings first, metrics as guardrails
Josh traces his perspective back to his time at Facebook and the industry’s metric-centric culture. He argues that beloved brands historically optimize for experience and emotion at the moment of creation, using metrics later to stay honest.
- •Josh’s background includes an acquihire into Facebook and learning product there
- •Critique: metric obsession is powerful but incomplete for creative work
- •Examples of brands (Disney, Nike, Apple) that optimize for human experience
- •Snapchat vs. Facebook example: feelings like closeness vs. mechanical activity metrics
- 15:00 – 19:38
Operationalizing “feelings”: how it shows up in day-to-day feature building (Peek example)
Lenny presses on how “optimize for feelings” works in practice, and Josh explains it’s currently more conversational than formal. Using Peek, he describes designing for “airiness” and effortless context switching, with the expectation they may formalize more as they scale.
- •They don’t yet have a formal PRD template listing target emotions
- •Small, tight teams align through direct conversations and shared context
- •Peek feature designed around lightness, speed, and minimal context switching
- •They expect process to evolve as team size grows
- 19:38 – 23:27
Shipping fast by “shipping values”: heartfelt intensity, beginner’s mind, and bias to action
Josh attributes the team’s rapid weekly shipping cadence to values rather than process. He breaks down specific cultural principles that produce speed: intrinsic motivation, humility, ambition, and a default to action.
- •“Heartfelt intensity” as the primary hiring and execution accelerant
- •“Assume you don’t know” → move fast, learn by doing, avoid over-planning
- •“Start by asking what could be” → ambition as motivation and quality driver
- •Swarming and team ownership help unblock shipping
- 23:27 – 28:21
Building values without corporate cringe: the “Notes on Road Trips” essay and the team interview process
Josh explains his skepticism of traditional corporate values and how the company arrived at a more authentic format. They interviewed every team member, extracted recurring phrases, and turned them into a narrative essay that serves as a cultural manual.
- •Values emerged organically from repeated patterns in team feedback
- •They conducted 1:1 interviews with the entire company (not surveys)
- •Most value phrases came directly from employees’ words
- •Turning values into a story (“Notes on Road Trips”) took ~3 months and heavy collaboration
- 28:21 – 38:07
Hiring a dream team: building the company as the product (and celebrating people publicly)
Josh describes an unusually deliberate approach: the founding goal wasn’t “build a browser,” but “build the dream team” and a way of working that attracts exceptional talent. He shares examples of senior leaders joining as ICs and explains how public celebration of contributors reinforces a talent flywheel.
- •Josh didn’t set out to build a browser; the ambition is an “internet computer”
- •They design the organization to attract top talent across disciplines
- •Examples: Chrome’s early leader, Vimeo product exec, Slack design leader joining as ICs
- •Interviews are framed to help candidates self-select out, not be pitched
- •They publicly celebrate hires and feature contributors to reduce CEO-centric hero worship
- 38:07 – 45:56
Building in public: radical trust-building, plus the risks of founder-centered storytelling
Josh frames transparency as a prototype with a specific goal: rebuilding trust in tech—especially for a browser handling sensitive data. He also shares his discomfort about the CEO becoming too central and the danger of sliding into “internet celebrity” dynamics.
- •Motivation: trust is broken in tech; transparency helps people know the humans behind the product
- •They share meetings and internal struggles to signal honest intent
- •Acknowledges potential downside: making the founder too central to the story
- •The goal is trust-building, not total transparency or attention-seeking
- 45:56 – 50:16
A prototype-driven org: “odd teams” like Membership and Storytelling (and what they replace)
Josh explains how first-principles thinking leads them to structure teams around end-to-end human relationships rather than standard startup functions. “Membership” unifies support, success, and research; “Storytelling” unifies marketing, PR, and investor narrative into a single craft.
- •Prototype-driven culture extends to org design, not just product features
- •Membership team owns the full lifecycle relationship with users
- •Storytelling team reaches people who don’t use Arc yet (press, investors, future users)
- •Team design focuses on incentives, craft, and holistic ownership
- 50:16 – 1:07:52
Why they don’t have traditional PMs (yet): PM as a project role, not an org
Josh argues that PM work is a set of verbs that different disciplines can lead depending on the project. He acknowledges scaling pressures and predicts they may eventually add more conventional PM structure, but wants “multidisciplinary makers” rather than process-heavy roles.
- •PM leadership should match the project (infra engineer for performance, membership for user-facing needs)
- •They’re hiring people who could be PMs elsewhere, but avoid a formal PM org
- •They prefer “mutts”: multidisciplinary builders who do what the work requires
- •Acknowledges likely future need for more dedicated PM capacity as time gets squeezed
- 1:07:52 – 1:13:54
Product-UX-focused companies and competitive strategy: why craft wins in commodity consumer markets
Josh explains why the “best experience” strategy is not anti-business for Arc—it’s the path to winning. In a universal, commoditized, lucrative category like browsers, emotional connection and delight are the rational differentiators that drive growth and revenue.
- •Browsers are universal consumer software used by almost everyone
- •The market is commoditized (often the same Chromium base) and highly lucrative
- •Differentiation requires emotional brand/product connection, not minor feature parity
- •Choosing a market where craft is how you win is a strategic decision
- 1:13:54 – 1:15:45
Nailing fundamentals: latency as table stakes (Tuple example) and the limits of “feelings”
Josh emphasizes that great UX can’t compensate for failing core functional requirements in certain categories. He uses remote collaboration tools and Tuple as examples where low latency is essential, and craft plus performance together create love.
- •In some markets, fundamentals like audio/video latency decide winners
- •Over-indexing on delightful features without performance can doom products
- •Tuple as an example: beautiful design plus exceptional audio quality
- •Feelings matter, but only after core functional bar is cleared
- 1:15:45 – 1:23:50
The long-term vision: Arc as an “internet computer” and the web as the next platform
Josh lays out the macro thesis: computing is shifting fully to the cloud, where apps and files become URLs and devices become interchangeable shells. Arc aims to be to the browser what the iPhone was to the cell phone, ultimately enabling a richer platform for developers on top of the web.
- •Cloud shift: apps and files increasingly live on the internet, not local devices
- •Future devices become interfaces to a persistent “computer” in the cloud
- •Arc’s ambition: iPhone-like leap for the browser—an “internet computer”
- •Web is a powerful open platform (cross-device, low friction, no 30% tax)
- •Incumbents’ incentives (Apple, Google) can limit web-native richness; Arc aims to expand capabilities
- 1:23:50 – 1:28:37
Lightning round: books, media, interview style, favorite tools, and an Arc usage mindset
Josh shares personal recommendations and a few tactical practices from how he hires and works. He also reiterates a key Arc adoption tip: approach it with a beginner’s mind rather than as “just another browser.”
- •Book recommendations: Harold and the Purple Crayon; Seeing Is Forgetting the Name of the Thing One Sees; God Saved Texas
- •Recent favorite: Episode 3 of The Last of Us; Adam Curtis documentaries
- •Favorite interview question: “What would you like to ask me?”; looks for authenticity
- •Tool shout-out: Tuple, plus Notion and Linear
- •Arc tip: don’t think of it as a browser—relearn how you use the internet
