Follow-up · Oct 28 callIdeas to react to

Less time validating titles. More time on the insight.

Zillow's exec talent and talent intelligence teams already build the maps, the benchmarks and the dashboards. The slow part is checking every title against multiple sources before it reaches a hiring manager. So instead of another demo, we ran four questions your teams ask on live data after the call, plus a mini talent flow for Zillow itself. Tell us which one is closest to a real question on your desk.

See the four ideas ↓ Your own talent flow
Mock Claude session · numbers from live Crustdata pulls · Oct 28, 2026
Found since the call

Four moves in your market that a dashboard won't show you.

7 of 10
AI leadership seats at Realtor.com were filled in the last 12 months.
Idea 02 →
30+
Opendoor leaders out since its new CEO, including its CTPO and SVP of Engineering, both now co-founding startups.
Idea 04 →
7
Former Zillow leaders in director-and-up seats at Realtor.com, including its SVP of Engineering.
Your talent flow →
12 to 1
Zillow leaders lost to real estate peers and AI companies in the last 12 months, against 1 hired from them.
Your talent flow →
What I heard

Rich insights for the hiring manager, without re‑checking every title.

We have to validate data, titling... through multiple resources.
We provide hiring managers with rich insights about the market... how the role is positioned.
How does Airbnb do this? Or how are they structured?
Talent mapping, where do I potentially expand my team and what roles... the way organizations are structured and scoped... leadership archetypes, characteristics and backgrounds.
What it has to do

Four things this has to get right for Zillow.

01

Titles you don't have to check twice

Every role comes with its start and end dates and the date we last checked it, in one record per person. An agent's "Vice President" and an engineering VP get sorted before anything is counted.

02

Benchmarks at the level the question is asked

"How is Airbnb structured?" needs seats by function and level, company by company, with the careers behind them. A headcount total doesn't answer it.

03

Inside the Claude you already use

The team already builds artifacts and dashboards in Claude. Crustdata connects to it as an MCP server, so the same chat can pull the data. Nothing new to log into.

04

Ready for a function that's scaling

Talent intelligence is a year old and growing into more of the business. Exec talent partners should be able to ask a benchmark question themselves and get a dated answer.

Why not build it ourselves?

Keep building in Claude. Change what it can reach.

Everyone has the same model now. What it finds depends on what it can see. So keep building the artifacts and dashboards yourselves, and give Claude more to see while it builds them. Here's the same question from the top of this page, asked two ways.

What this page took
1,400+

career histories read after the call, every role with its dates

Caught before counting
250

agent and broker "Vice President" titles at Compass that would have skewed the leader count

Web searches needed
2

to confirm two deal dates. Everything else came from the people data.

By hand, that's opening 1,400 profiles and reading every job on them. Web search can't read career histories at that scale, so Claude on its own never gets to the talent flow, the archetype or the leaver list.

Claude with web search

The leaders who made the news

  • Leadership pages, press releases and the hires that got announced
  • Titles as each source wrote them, with a start date only when a source happens to mention one
  • No way to count a whole function, or to tell an agent's "Vice President" from an engineering VP
  • Every name still has to be checked somewhere else
Claude with Crustdata

Everyone in the seat, dated

  • Every person at the three companies in our index, out of 1B+ people and 60M+ companies
  • Each role with start and end dates, seniority, and the date we last checked it
  • Past employers you can join on, which is what a talent flow or an archetype is built from
  • Today's answer: 19 leaders, 6 with 3+ years, every record checked between Oct 20 and 27

Web search still has a job. We used it twice for this page, to confirm when Rocket closed its Redfin deal and when Opendoor named its new CEO. It's the people data underneath that Claude can't search its way to.

Ideas to react to

Four questions your teams ask. Which one is closest?

Each one ran on live data after the call, on your peers and your market. They're starting points, not a menu: change the peers, the function or the seat, and we'll run it again. We show companies and titles here; names come back in the pull.

01

One exec search, all the way down

Exec searchFor Jesse

The search that started this conversation: every VP-and-up product, engineering, design and AI leader at Redfin, Compass and Opendoor, with time in seat.

Before counting, 250 of Compass's 446 VP-level titles had to go. They belong to agents and brokers ("Vice President, REALTOR"). What's left is 19 leaders. 6 have held the seat 3+ years, and 10 started in the last 12 months.

person_search → title check → tenure → watch on the 19
The question for you

Which live search should we run this on?

Send the seat and the peers. We'll send back the map, with tenure and the date behind every title.
In seat 3+ years · 6 of the 19Oct 28, 2026
CompanyTitleIn seat sinceLast checked
RedfinVice President, User Growth EngineeringJun 2019Oct 27
RedfinHead of ProductAug 2021Oct 27
OpendoorHead of Valuations and ML FoundationJun 2022Oct 21
RedfinVP of EngineeringJul 2022Oct 27
CompassSVP, Head of ProductJan 2023Oct 27
CompassVice President of EngineeringJan 2023Oct 20
The other 13 are newer, including Redfin's Chief Product and Design Officer (since March) and Opendoor's Chief AI Officer (since June).
02

Where AI sits at your peers

Org designFor Ashley and Shawn

"How is Airbnb structured?" is really a question about seats. We counted director-and-up seats with AI, ML or agentic in the title.

Zillow has the deepest bench of this group at 24. Realtor.com is building fastest: 7 of its 10 are less than a year old. And the top seat differs everywhere: a Chief AI Officer at Opendoor since June, one VP of AI at Airbnb, two SVPs splitting product and engineering at Realtor.com.

person_search by company, level and title → start dates → the most senior seat
The question for you

Which function, and which peer set?

AI is one cut. The same count works for design, data, trust and safety, or any org a hiring manager asks about.
AI leadership seats · director and upOct 28, 2026
CompanySeatsNew in 12 moMost senior AI seat
Zillow246SVP, AI
Airbnb134VP of AI, since Feb 2026
Realtor.com107Two SVPs: engineering (consumer and AI), product and AI innovation
Opendoor52Chief AI Officer, since Jun 2026
CoStar Group30SVP, Matterport engineering (AI/ML)
Compass20Senior Director, data and AI engineering
Counted by title, so a leader who runs AI under a plain "VP, Engineering" title isn't in here. That's the kind of line we'd set with you.
03

What a marketplace product leader looks like

Leadership archetypesFor Shawn

An archetype is a pattern across real careers. We read the histories of 97 VP, head-of-product and CPO leaders at 12 consumer marketplaces, from DoorDash, Uber and Airbnb to Redfin and Compass.

More than half came through big tech. Almost half hold an MBA. And 4 in 10 were promoted into the seat rather than hired into it, which changes where you look for the next one.

person_search on title and company → full work and education history → the pattern
The question for you

Which leader profile is on your desk right now?

Name the seat and the companies to read. Or name two or three leaders you'd hire again: we'll work out what they share, then find others who share it.
VP of Product, consumer marketplace · 97 leadersOct 28, 2026
20 yrs

into their career, at the median

1.8 yrs

in the current seat (4.7 at the company)

54%

came through big tech

47%

hold an MBA

40%

were promoted into the seat from inside

34%

had worked at another marketplace first

28%

founded something along the way

26%

started their careers as engineers

Most common past employersAmazon19Microsoft16Google13Meta13Uber7McKinsey6
DoorDash, Expedia Group, Uber, Airbnb, Etsy, Booking.com, Lyft, Instacart, Angi, Redfin, Compass and Turo. Assistants, product marketing and design titles left out.
04

When a peer reshuffles, leaders come loose

Talent signalsFor Jesse and Corey

Two of Zillow's closest peers changed at the top in 2025. Rocket closed its Redfin deal on July 1, and Opendoor named a new CEO on September 10.

Since then, 30+ Opendoor leaders at director level and up have moved on, and 15+ from Redfin. Their product and engineering leaders went to stealth startups, CTO seats and a direct competitor.

person_search on past company, level and start of the new role → where they went → watch for the next move
The question for you

Which peers should we watch?

A weekly watch flags the week a leader at any of them changes roles, in Slack or email, so the market map stays current on its own.
30+

Opendoor leaders out since Sep 2025

15+

Redfin leaders out since Jul 2025

Where they went · a sampleOct 28, 2026
WasNowLeft
Opendoor · Chief Technology and Product OfficerCo-founder, stealth startupJun 2026
Opendoor · SVP, Head of EngineeringCo-founder, stealth startupApr 2026
Opendoor · Sr. Director of EngineeringCTO, BizeeJul 2026
Redfin · VP of EngineeringCTO, RentSpreeNov 2025
Redfin · VP of ProductHead of Consumer Product, CompassFeb 2026
Redfin · CPO, rentals divisionChief Product Officer, CerboDec 2025
Redfin · Chief Financial OfficerCFO, AncestryAug 2025
One move shows up twice on this page: Redfin's VP of Product left in February and is one of the 19 in idea 01, as Compass's Head of Consumer Product since March.
Your own talent flow, in miniature · for Ashley

The one you can grade us on.

Zillow's director-and-up moves over the last 12 months, read from public profiles. You know the real numbers, which makes this the fastest way to check our data.

Where from, where toJoined Zillow fromLeft Zillow for
Big techin from Amazon 3, Google, Microsoft, Airbnb · out to Netflix, NVIDIA, Salesforce, Expedia 9 6
Real estate and proptechin from Better · out to Realtor.com 2, lpt realty, RETR, Hover 1 7
AI companiesout to Fireworks AI 2, Anthropic, Legora, a stealth AI startup 0 5
Own venture, independent or between rolesfounders, consultancies, independents 3 8
Everything elsein from Modern Health, Docusign, Yext, Lowe's · out to Toast, Chime, Block, Canva 24 20
Total, outside hires and exitsplus 56 director-and-up seats filled by internal moves 37 46
What stands outBig tech is where Zillow hires leaders from. Real estate peers and AI companies are where they go: 12 out, 1 in.
Your alumni, in their leadership18 former Zillow people hold director-and-up seats at your peers today: 7 at Realtor.com, 3 each at Redfin, CoStar and Homes.com, 2 at Airbnb. They include Realtor.com's SVP of Engineering, a Zillow VP of Engineering until 2023, and Redfin's new VP of Product and Design.
The question for youWant this by function, or for a peer? The same table runs for Engineering at Zillow, for Redfin since the Rocket deal, or as a boomerang list of your alumni.

How to read it: Nov 2025 to Oct 2026, director and up, from people in our index. Someone who hasn't updated a public profile yet won't be counted, so we expect to be a little under. You'll know.

Straight answers

The questions from the call.

Q1What does this give us that we can't build ourselves in Claude?

Build it in Claude. That's the right place for it. What Claude can't do on its own is reach a dated, structured record for every person at a company.

Crustdata is that layer: titles with start and end dates, past employers, seniority, and when each record was last checked, across 1B+ people and 60M+ companies. Every table on this page came from it.

Q2Is this a request to you, or a platform we pull from ourselves?

Mostly the second. You get keys to our API and MCP server, so the LLM Zillow already uses can call Crustdata directly. There's also a web platform for the team.

And we'll still run a one-off request when that's faster for you.

Q3Who do you compete with, and what's different?

People Data Labs and Coresignal come up most. We're a primary indexer: we build the data from the public web ourselves, keep it fresh, and match each person across sources into one record. We were also built API and MCP first, so it works inside an LLM without extra plumbing.

The fair test is the data, not the pitch. Pick a list you've already validated and compare.

Q4How would we know the titles are right?

Check us against something you already know. Every role carries its start and end dates and a last-checked date, so a wrong title is easy to spot and trace back.

The quickest test is your own team: the talent flow above covers Zillow's director-and-up moves for the last 12 months, and you have the real numbers.

What happens next

Send the hard one.

Everything below came out of the call. No dates we didn't agree to.

The short version

Send the question that's hardest to validate today. We'll send back the answer.

Pick one of the four ideas, change one, or bring your own.

Grab time with Doug Reply to Doug