long-form-interview· Rich Barton

How Rich Barton Built Expedia and Zillow from $0 to $35B

Rich Barton's playbook for building $35B+ of consumer brands (Expedia, Zillow, Glassdoor) is a tightly coupled system: pick a big pond with locked-up data; free the data with a power-to-the-people product; build a made-up brand-name with rare letters and dog-name energy; market the product through provocation rather than ad spend by feeding constantly-changing data to the local-news newshole; and protect the in-house intrapreneurs from corporate immune-system rejection.

rich-bartonexpediazillowglassdooravvotim-ferrissprovocation-marketingzestimatebill-gatessteve-ballmerbarry-dillerdara-khosrowshahibrad-chasegreg-maffeibill-gurleybenchmarkmicrosoft-spinoutgive-to-gettim-ferriss-show94% confidence

Why this is in the corpus

Rich Barton (co-founder of Expedia, Zillow, Glassdoor; Microsoft alum, Benchmark venture partner, Netflix board) on the operator playbook he reused across four consumer brands: Provocation Marketing, the give-to-get UGC mechanic for Glassdoor, the Zestimate as a byproduct of a failed auction strategy, the made-up-word naming framework (Z/X/Q + 2 syllables + dog-name + verb-able), the spin-out-as-HR-experiment that birthed Expedia, and the leadership lesson that almost all his mistakes were leaving the pitcher on the mound too long.

Summary for skimmers

Rich Barton on Tim Ferriss: Brad Chase's "what's your next big idea?" response to a $10-20M Microsoft failure as the culture of intrapreneurship; Expedia born inside Microsoft as Bill Gates's personal venture; the Microsoft spin-out framed as an HR retention experiment; Zillow's pivot from home-auction to the Zestimate (the byproduct that became the killer feature); the Provocation Marketing playbook (find an emotional sacred cow, address it without scaring people, wire constantly-changing data into the local-news newshole instead of buying ads); the give-to-get UGC mechanic at Glassdoor; the made-up-word naming framework (rare Scrabble letters, 2 syllables, dog-name energy, verb-able); board-seat selection criteria (Local / Fun / Lucrative); and the leadership rule "almost all my mistakes were leaving the pitcher on the mound too long."

Briefing

What survives the editorial filter

This page should feel like a smart colleague already listened for you and left only the operating logic worth keeping. Not everything said in the episode makes it through.

Trust signal

direct_practitioner_account

Guest type: practitioner.

Best used for

Rich Barton (Expedia, Zillow, Glassdoor) on Provocation Marketing, the made-up-word naming framework, the give-to-get UGC mechanic, the Zestimate as failed-auction byproduct, the Microsoft spin-out as HR retention experiment, and the leadership rule that almost all his mistakes were leaving the pitcher on the mound too long.

Hold lightly

No explicit downgrade reason stored yet for this episode.

Principles

Durable claims that survive beyond the speaker's biography — each with explicit limits, transferability judgment, and evidence.

Principle

Protect the intrapreneurs from the corporate immune system

Corporates lose intrapreneurs by default; senior leadership must explicitly protect them after failures, or the variance is selected out.

When an intrapreneur on your team takes a big swing and misses, the next 1:1 question is "what is your next big idea?" — not the post-mortem.

“Brad sat me down. I thought I was for my review. I thought I was going to who knows — it was a 10, 20 million dollar mistake. And he said: alright, what is your next big idea?”Rich Barton

“Mainline corporate culture often rejects the innovators. The leadership needs to hire, cultivate, protect, and invest in those people — be the foreign bodies so they do not get rejected by the corporate immune system.”Rich Barton

Principle

Power-to-the-people positioning — free the locked-up data

Building a consumer brand on top of a locked-data industry produces a structurally winning position — customers love it, incumbents lobby against it, and elected officials side with constituents.

When evaluating a category, find the data the industry has worked hardest to keep opaque; that is the wedge.

“We were the power to the people guys. We were the guys who freed all the information for the regular traveler. We were like, well, want to be power to the people here too. Right.”Rich Barton

“In a lot of places we had real estate professionals lobbying to have us outlawed. The strategy was just activate the latent love for the product itself — the lobbyists got nowhere.”Rich Barton

Principle

Big pond + good fishermen — the only opportunity-selection rule that matters

Real problems are not enough — opportunity selection requires a big pond AND operators who can fish it. Rule out the small ponds at the top of the funnel.

When evaluating an opportunity, audit two questions explicitly: is the addressable pond truly big, and are we structurally good fishermen for it. Reject if either is no.

“My business criteria for doing stuff is, is it a big pond and are there good fishermen? A lot of entrepreneurs make this mistake of identifying a really big problem, but it is just a small opportunity.”Rich Barton

“With Zillow we knew 100% that there would be a leading digital real estate marketplace in the US at some point. Inevitability and business model — who knows, who cares, it is giant, it is a big pond.”Rich Barton

Principle

Almost all leadership mistakes are leaving the pitcher on the mound too long

Almost all leadership mistakes are firing too late — pull the pitcher quickly and frame the release as a partnership.

When you have known for >30 days that someone needs to go, schedule the conversation this week; the team is paying the cost while you delay.

“All of my mistakes as a leader have been leaving the pitcher on the mound too long, hoping that the arm would get better.”Rich Barton

“If you are not happy with the performance of this person, I guarantee you the person is not happy either. You can increase love in the world by releasing that person to find where that person belongs.”Rich Barton

Frameworks

Reusable systems and operating models — including when they help and when they break.

Framework

The made-up-word naming framework — Z/X/Q letters, 2 syllables, dog-name + verb

For consumer brands, made-up names with rare Scrabble letters, two syllables, dog-name energy, and verb-ability dominate literal-domain names — you own the definition rather than renting it.

Run your candidate names through the Z/X/Q test, the syllable test, the dog-name test, and the verb-able test before committing.

“Pick the super rare letters — Z, X, Q — they are very distinctive, jump off a page, stick in people brains. Fewer syllables is better. Two syllables is the sweet spot. The word should be a good dog name. It can be turned into a verb. Double letters and palindromes — people remember them.”Rich Barton

“The hard way and the best way for consumers is to make up a word. Once you do, you own that word. Apple, Amazon — middle path. You have to build a new definition for that word.”Rich Barton

Framework

The Provocation Marketing curriculum — emotional sacred-cow + good-feeling provocation

Provocation Marketing — emotional sacred-cow + good-feeling provocation + constantly-changing data fed to the newshole — produces owned attention that ad spend cannot match.

For your category, identify the data the industry keeps opaque and the emotional zone customers care about; design a feature that fuses both and wire the data into local news distribution.

“Find a seven deadly sin zone — something emotionally core to us. Then go address some sacred cow or taboo in that space. Most of the ideas would be really negative. Get rid of all those — a cheap way to get attention is to scare people. You want people to be provoked but feel good or tickled or entertained.”Rich Barton

“There is an infinite newshole for housing data at local newspaper level. We set up a mechanism to feed that data — terrific brand-builder, rather than spending ad money.”Rich Barton

Framework

Local / Fun / Lucrative — board-seat selection criteria from Greg Maffei

Board seats should pass at least 2 of 3 — Local, Fun, Lucrative — and you should exit boards that tip into professional-director theater.

Audit your current board commitments against Local/Fun/Lucrative; exit any that fail 2 of 3.

“Greg Maffei told me earlier in my career: is it local, is it fun, is it lucrative? At least 2 of 3 — trifecta scores all 3.”Rich Barton

“If a board tips into professional directors who are really worried about their board director reputation, it becomes more about CYA. Versus let's build a company.”Rich Barton

Signals

What appears to be shifting, for whom it matters, and what happens if you ignore it.

Signal

Cloud HQ as inclusivity flywheel — distributed work brings off-rampers back

Cloud HQ is the durable inclusivity flywheel — distributed work re-includes off-rampers (often experienced moms / caregivers) and widens the talent pool permanently.

If you are a return-to-office-by-default leader, run a 90-day distributed-friendly experiment with explicit inclusivity goals before reverting.

“Cloud HQ enabled really smart very experienced moms who may have decided to take the off-ramp into primarily mom-hood. It enabled them to come back. Likewise for fathers. Companies that get angry when they see you in your car on Zoom because you are at a dentist appointment are leaving talent on the table.”Rich Barton

Opportunities

Only included where there is a buyer, a real wedge, and a plausible revenue path — not vague idea theater.

Opportunity

Opportunity: Data-as-marketing in regulated industries

$10B+ opportunity.

“Publish the data. The hoarders lose.”Rich Barton

Durability: Time-sensitive.

Named.

Lessons still worth keeping

Useful takeaways that did not fully clear the bar for durable principle status.

Lesson

The Microsoft Expedia spin-out as HR retention experiment

Spin-outs are a high-leverage retention tool at scale — they re-couple performance to reward for high-leverage builders who would otherwise leave for the wild outside opportunity.

If your scaled-org has hidden talent that is at flight-risk to outside startups, model the spin-out as an HR move (not just a financial one) before defaulting to retention bonuses.

“Microsoft was getting so big that people could hide out in random corners and as long as Windows NT or Office succeeded they could make money on stock options. That is a compensation accountability disconnect. The best people had this field-of-dreams wild opportunity outside, and they would have gone.”Rich Barton

“All but two people of the 150 decided to come take the adventure.”Rich Barton

Lesson

The Zestimate as failed-auction byproduct — pursue A, discover B

When your A-plan fails for structural reasons but the underlying goal is still valid, the byproduct of pursuing the goal is often the actual product.

When killing a failed product, ask explicitly: what is the byproduct of our pursuit that customers actually responded to. Build that.

“In pursuit of price discovery we found the Zestimate, which was our killer feature. The visual that popped in our heads on the home auction web experiment was a real-time estimated value over neighborhoods — we wanted prices on every roof.”Rich Barton

The Plays

Try these this week

Verb-first executable actions — each one tied to a stated outcome in the episode.

The give-to-get UGC mechanic — Glassdoor's salary disclosure flywheel

Outcome: When user-generated taboo data is the asset, give-to-get reciprocity converts free-riders into contributors and produces a self-reinforcing supply flywheel.

“Their innovation after kind of hand-cranking it with surveys was give-to-get. You show me yours, I will show you mine. We will give you a little taste, but if you want any more data, you have to share your salary and your title and your company — promise you will be anonymous — and do a company review.”
— Rich Barton

6-12 months to reach self-sustaining flywheel per (proposed)

1
Identify the taboo data asset
Pick a category of data that is socially sensitive (salary, attorney quality, medical pricing, landlord behavior). The taboo factor is what creates the Provocation Marketing surface.
2
Build the disclosure gate
Allow users to see a small taste of aggregated data without contributing. Lock the deeper data behind a contribution gate that requires their own data point plus structured fields (title, company, role, etc) plus an anonymous review.
3
Guarantee anonymity with explicit protocols
Promise anonymity in copy. Implement protocols for unique-position contributors (e.g., the only CFO at a small company) — aggregate within larger buckets or suppress display until N>=3.
4
Amplify with public-figure data
Layer review of public-facing roles (CEO performance, partner reviews, professor reviews) on top of the anonymous individual data. Public-figure provocation generates news cycles; the anonymous data generates the trust.
5
Wire the data into Provocation Marketing distribution
Feed constantly-changing aggregated data (median salary by city/role, CEO approval ratings) to local news + LinkedIn / X. The data is now the marketing.
6
Monitor contribution rate vs view rate
If view-to-contribute ratio drops below threshold, tighten the gate (less free taste). If contribution rate is low at signup, loosen the friction (fewer required fields).

Stop or pivot when

→Free-view-to-contribution ratio >1:20 means the gate is too loose; <1:5 means too tight
→Re-disclosure rate (users updating data each year) <20% means the trust loop is broken — audit anonymity

Scripts

Before you start

· A category where the data is socially sensitive enough to be a Provocation Marketing surface
· Anti-deanonymization protocols engineered before launch
· A patience window of 6-12 months for the flywheel to compound

ugc-mechanicsdata-flywheelsmarketplace-designseedseries-aseries-bseries-c

Provocation Marketing PR data distribution — wire the local-newspaper newshole instead of buying ads

Outcome: When you have constantly-changing branded data, wire it directly to the local-newspaper newshole as PR distribution — it dominates ad spend in cost-per-impression and credibility.

“There is an infinite newshole for housing data at local newspaper level once-upon-a-time. If you could wire that up to constantly feed the endless appetite — a terrific brand-builder rather than spending ad money.”
— Rich Barton

3-6 months to ship infrastructure; 12-18 months to reach mature volume per (proposed)

1
Identify constantly-changing data customers care about
Real-estate prices by zip code, salary by role and city, attorney ratings by metro, etc. The data must be (a) updated continuously, (b) of clear consumer interest, (c) tied to your product's data asset.
2
Slice the data at local granularity
National numbers do not feed local newshole. Slice by city, county, congressional district, school district, neighborhood. Each slice is a separate news story.
3
Ship a data API for journalists
Self-serve, free, with attribution. Journalists pull custom queries; the API returns pre-formatted facts with context. Reduces the local journalist's research time to zero.
4
Pre-format the news copy
For each major data update, ship a press packet with headline, lead paragraph, supporting facts, branded chart image. Make the journalist's job to lightly edit and publish.
5
Schedule spokesperson availability by region
Have a regional spokesperson available for quote on demand. Local journalists need a quote — being responsive within hours is the binding constraint.
6
Track coverage by metro / segment
Build a coverage dashboard. Identify under-covered metros and brief journalists there directly. Treat local-news distribution as a managed channel, not a side effect.

Stop or pivot when

→If <10 local stories/month after 6 months, the data is not local enough — re-slice
→If >50% of stories are negative, audit the data presentation — Provocation Marketing requires good-feeling provocation
→Quote-response SLA must be <4 hours for journalists; missed SLAs collapse repeat coverage

Scripts

Before you start

· A data asset that updates continuously and slices to local granularity
· Engineering capacity to ship a journalist-facing API
· A communications team that operates on a 4-hour quote-response SLA

pr-distributioncontent-marketingdata-productsseedseries-aseries-b

Decision Moments

Actual decisions, real outcomes

Specific decisions narrated in the episode with their outcomes and transferable lessons.

Barton's MS-DOS5 + Dos for Dummies bookstore-bundle product had failed; $10-20M write-off; Barton expected to be fired during his review at Microsoft

Did: Did not punish the failure; opened the next 1:1 by asking "alright, what is your next big idea?"Outcome: Barton retained, eventually founded Expedia inside Microsoft and four other consumer brands; the cultural move propagated through Barton's career as the intrapreneur-protection principle

Corporate culture that protects intrapreneurs from the immune system after failures retains the swing-takers; the "what is your next big idea?" question is the cultural artifact that signals protection

Part of an emerging decision pattern across multiple episodes

Bill Gates green-lit Expedia inside Microsoft; later, the dot-com bubble was peaking and Microsoft had 65% ownership but no board seat; the Expedia team would have left for the wild outside opportunity

Did: Spun out Expedia in November 1999 explicitly framed as an HR retention experiment; pitched the spin-out to Ballmer not on financial value-unlock but on retaining the high-leverage buildersOutcome: 150 of 150 (minus 2) employees chose to spin; Expedia recovered from the 2000 crash within ~2 years and became foundational to online travel; Barton's subsequent Provocation Marketing playbook was incubated outside the parent

At scale, equity-based compensation decouples performance from reward — high-leverage builders subsidize hiders. Spin-outs re-couple performance to reward and retain the builders who would otherwise leave

Part of an emerging decision pattern across multiple episodes

Zillow's original product was an online home auction; the auction failed because housing has no real-time liquid market and required customer behavior change beyond viable adoption

Did: Killed the auction; in the pursuit of price discovery, discovered the Zestimate (algorithmic estimated home values plotted on aerial Google-Maps view, like a stock chart for any home)Outcome: Zestimate became the killer feature; Walt Mossberg coverage drove millions to launch; Zillow became the dominant US digital real-estate marketplace and a $35B+ company

When the A-plan fails on structural grounds but the underlying goal is still valid, the byproduct of pursuing the goal is often the actual product; keep the goal anchored and let the path mutate

Part of an emerging decision pattern across multiple episodes

Bill Gurley (Benchmark, Zillow board) challenged Barton at launch — what if Zillow had no marketing budget? Barton's Expedia instinct was to spend on advertising for brand growth

Did: Accepted the no-ad-budget constraint; built Provocation Marketing instead — Zestimate as provocation feature, server-tipping-over launch-day press, local-news data distribution Bloomberg-styleOutcome: Zillow launched to millions of users without ad spend; Provocation Marketing became Barton's reusable playbook across Glassdoor (give-to-get salary disclosure) and Avvo (rate the attorneys)

A no-ad-budget constraint forces the team to invent provocation features that produce owned attention; constraint-driven marketing dominates per-impression-bought ad spend in cost and credibility

Part of an emerging decision pattern across multiple episodes

Tensions surfaced

Contradictions and trade-offs the episode raises — judgment calls a thoughtful operator has to navigate.

Tension

Tension: Provocation marketing vs trust-building marketing

Provocation short-cycle; trust long-cycle.

“Provocation works when relationships are short.”Rich Barton

Durability: Durable.

Productive tension.

Corpus connection

Where this episode fits for retrieval

What kinds of decisions this briefing is best pulled into.

Primary decisions

• product-strategy
• strategy
• marketing-budget-allocation