Zillow Group, Inc. (NASDAQ: Z) — The #1 Housing Portal, Now Priced Like Its Moat Is Google’s to Take

⚡ Claude’s Take

The author’s own independent opinion and general information — not investment advice. The analysis below carries no position and no price target.

Verdict: HOLD / accumulate-on-weakness. Great franchise, unresolved existential question, price finally reasonable but not yet cheap enough to pay for the risk. Not a short. Directionally, I’d anchor fair value at roughly 3.0–3.5x sales (~$40–48/share) if Rentals keeps compounding, operating leverage keeps landing, and the Google/aggregation fear proves overdone — with a genuine ~2.0–2.2x sales (~$24–28) downside if native search-listings and Homes.com structurally bleed Zillow’s traffic-monetization engine. That asymmetry is why I’d accumulate only in the high-$20s to ~$30, not at $34, and size it small.

Zillow is the best asset in residential-real-estate technology: the #1 consumer destination by a wide margin (Comscore #1; a genuine two-sided network of consumer traffic ↔ agent advertisers), an asset-light 74%-gross-margin model, net cash, and — critically — a Rentals franchise growing 39% that is a better business than the transaction-dependent Premier Agent core and less hostage to mortgage rates. The market has repriced the whole thing as if the moat is already gone: the stock is down ~64% from its September-2025 high, trades at ~3.1x sales (the 14th percentile of its own decade, cheapest-ever on sales ex-2022 crash), with textbook falling-knife factor characteristics (negative 12-1 momentum, a −76% five-year max drawdown, rate-sensitivity beta of −0.65, deeply negative Sharpe at every horizon). What broke it in June 2026 was Google inserting native home listings into search — a real threat to the top of Zillow’s funnel — stacked on a frozen housing market (existing-home sales at a ~30-year trough), post-NAR commission compression shrinking the pool that funds Premier Agent, and CoStar’s Homes.com outspending everyone to give agents free leads.

Here is why I stop at HOLD rather than buy the dip. First, the business has never earned its cost of capital — ROIC is ~0% to negative across the whole five-year set, there is a −$1.86B accumulated deficit, and the “profitability” everyone cites ($622M adjusted EBITDA) is 63% stock-based-comp add-back; strip the $390M of SBC and real EBITDA is ~$232M, i.e. the honest EV/EBITDA is ~33x, not 12x. Second, management burned ~$1B+ on the Zillow Offers iBuying misadventure and then repurchased $670M of stock in 2025 at ~$70 that is now worth ~$34 — capital allocation has been expensively pro-cyclical. Third, and most important, the Google/native-listings and commission-pool threats are the rare kind that could permanently impair the aggregation moat, and they are unquantifiable today. When the central risk is a step-change in the moat rather than a cyclical air-pocket, you demand a bigger discount than 3x sales — and you get paid to wait, because there is no dividend and buybacks just fund dilution. Framing: a franchise falling-knife — a real, dominant, cash-generative platform caught in a genuine structural scare, not a broken business. Conviction: medium. Flips bullish on evidence the Google integration is not diverting Zillow’s engaged-shopper traffic (stable app/direct-traffic share + For-Sale revenue holding through 2026) and SBC falling toward ~10% of revenue. Flips bearish on a measurable step-down in Premier Agent revenue per visit or a Compass/CoStar-led breach of Zillow’s listing-access position. Tag: “You don’t pay a network-effect multiple for a network the market thinks Google just entered.”

📈 Stock Price Action — Five-Year Event Map

Factual price history — no recommendation, no price target. Price moves are FACT; attributed drivers are INTERPRETATION.

Over five years Zillow round-tripped a full cycle and then some: from a post-COVID/iBuying-euphoria ~$120 (mid-2021, itself down from a ~$212 Feb-2021 peak), down to a ~$26 rate-shock trough (Oct 2022), back up to a cycle high of $93.88 (17 Sep 2025) — and then a violent ~64% collapse to a $29.23 low (25 Jun 2026). It now trades at $33.91 (close 2026-07-17), −63.9% off its 52-week high ($93.88) and −72.7% off the five-year high, sitting below a falling 200-day EMA (~$49.6) with the 50-EMA (~$35.3) rolling over — an unambiguous, still-intact downtrend.

Cycle narrative. (1) The 2021 top and first leg down was Zillow’s own unforced error — the capital-destructive exit of Zillow Offers. (2) 2022 was macro: a rate shock that hit every long-duration, unprofitable growth name and froze the housing market that Zillow monetizes. (3–5) 2023–2025 was a genuine operational recovery — the company cut costs, refocused on the asset-light “housing super app,” and grew Rentals and Zillow Home Loans, driving the adjusted-EBITDA margin from negative to 24% and the stock back to a fresh cycle high near $94. (6) From late 2025 the housing freeze simply would not thaw — existing-home sales stuck near a 30-year low — and the multiple began compressing on a business whose For-Sale revenue is levered to transaction volume. (7) The June-2026 Google catalyst was the qualitative break: the fear shifted from “cyclical air-pocket” to “the aggregator’s traffic funnel is under structural attack,” and the stock lost another third in weeks. (8) The subsequent bounce off $29 is oversold-reflex plus a Q1-2026 revenue beat, not a resolution of the structural question. Every attribution here is INTERPRETATION cross-referenced to earnings dates, 8-K events, the news feed, and the FactorsToday tape; the price levels themselves are FACT from the five-year CSV.

1. Executive Summary

Zillow Group is the dominant consumer gateway to United States residential real estate — the #1 real-estate destination by audience (Comscore), a genuine two-sided network of consumer traffic and agent advertisers, monetized through an asset-light, ~74%-gross-margin advertising-and-services model spanning four revenue categories: Residential ($1.70B, 66% of revenue, +7%), Rentals ($630M, 24%, +39%), Mortgages ($199M, +37%), and Other. FY2025 revenue grew 16% to $2.583B, gross profit was $1.92B, and management-defined adjusted EBITDA reached $622M (24% margin) — the culmination of a multi-year operational recovery from the 2021–22 near-death experience (the capital-destructive exit of the Zillow Offers iBuying business, followed by a rate-shock that froze the housing market). Operating leverage is real and improving: the GAAP operating margin recovered from −12.7% (2023) to −1.2% (2025), and the company printed its first positive GAAP net income ($23M) in years.

That is the bull half of the ledger. The bear half is equally concrete. Zillow has never earned its cost of capital — ROIC sits around zero-to-negative across the entire five-year window, and the balance sheet carries a −$1.86B accumulated deficit against $2.82B of goodwill (58% of book equity; tangible book equity is only ~$1.78B, ~$7.4/share). The celebrated “profitability” is substantially cosmetic: adjusted EBITDA adds back $390M of stock-based compensation — roughly 63% of the total and ~15% of revenue — so real EBITDA net of SBC is closer to $232M, and the ~$101M of reported free cash flow is entirely underwritten by paying employees in stock. Management then spent $670M repurchasing stock in 2025 at ~$70/share (now ~$34) — largely to offset SBC dilution rather than to return capital — a pro-cyclical, value-destructive use of cash that echoes the iBuying error.

The investment debate now hinges on a structural question, not a cyclical one. In June 2026, Google began inserting native home listings directly into search results — a credible threat to the top of Zillow’s traffic funnel — layered on a frozen housing market (existing-home sales at a ~30-year low), post-NAR-settlement compression of the buyer-agent commission pool that ultimately funds Premier Agent, and an aggressive, well-capitalized assault from CoStar’s Homes.com. The stock has repriced accordingly: down ~64% from its September-2025 high to ~3.1x sales — the 14th percentile of its own decade-long range. The body below assesses, without a recommendation or price target, whether that de-rating reflects a permanently impaired aggregator or an oversold #1 platform with a genuinely compounding Rentals franchise and unrealized operating leverage.

2. Business Overview

Zillow Group operates the largest consumer-facing digital marketplace for U.S. residential real estate. At its core it is not a brokerage, a lender at scale, or a homebuilder — it is a lead-generation and advertising platform that aggregates the nation’s home-shopping audience and rents that audience, plus a growing suite of transaction software, to the professionals who actually get paid on the transaction: real-estate agents, multifamily property managers, mortgage loan officers, and home builders. The company monetizes attention and intent, not the sale itself. (Fact — FY2025 10-K, “Business,” filed 2026-02-11.)

Scale of the audience. The Zillow Group portfolio (Zillow, Trulia, StreetEasy, HotPads, Out East) drew an annual-monthly high of 259 million unique users in July 2025 and roughly 9.6 billion visits in 2025 (+3% YoY), with average monthly unique users of 235 million (+6% from 221 million in 2024). Zillow is the most-visited and, per third-party studies, most-trusted online residential real-estate brand, and — a genuinely telling data point — the word “Zillow” is now searched more often than the term “real estate” in the United States, and Zillow leads the category in active app users. (Fact — 10-K, “Competitive Advantages”; sources cited therein: Comscore Dec-2025, 2025 Google Trends, Sensor Tower Jan–Dec 2025, Life Story Research 2025.) Management estimates roughly 80% of U.S. residential transactions involve an agent who uses at least one Zillow product (Premier Agent, Follow Up Boss, ShowingTime, Showcase, or dotloop). That figure blends the consumer platform with the acquired software footprint and should be read as reach, not monetized share, but the aggregation position is real. (Fact/Interpretation — 10-K.)

The four revenue categories (FY2025, total revenue $2,583M, +16% YoY):

(Fact — 10-K Consolidated Statements of Operations and MD&A, lines reconciled to the filing.)

How Residential is monetized. The engine is Premier Agent, which sells buyer/seller “connections” (validated leads) to agents. It runs on two pricing models. The legacy market-based pricing (“share of voice”) is a subscription-like auction: agents buy a share of advertising in a ZIP code and receive connections in proportion to that share. The newer Zillow Preferred (rebranded in Q4-2025 from the “Flex” program) is a pay-for-performance model — invite-only, for top agent teams — where Zillow delivers leads for free and collects a success fee (generally 15–40% of the agent’s commission) only when a transaction closes, typically within two years. This is a structurally superior monetization: it aligns Zillow’s take with closed volume rather than upfront ad budgets and lets Zillow capture a share of the commission rather than a fixed subscription. Management reports that Enhanced-Market connections exited 2025 at over 40% of the total, and that pushing more volume through the integrated/Preferred funnel lifted For Sale revenue per Total Transaction Value (TTV) to 8.5 basis points in 2025 from 7.9 bps in 2024 (+8%) — the single most important internal KPI, because it shows Zillow extracting more revenue per dollar of industry transaction volume even as that volume stays frozen. (Fact — 10-K MD&A “Key Metrics.”) Residential revenue per visit rose 4% to $0.178. (Fact — 10-K.)

Rentals is a two-sided advertising marketplace: property managers (multifamily and “longtail” landlords under 25 units) pay for exposure and lead volume; renters search, tour, apply, sign, and pay rent on-platform. Zillow had 2.4 million average monthly active rental listings including 72,000 multifamily properties (+44% YoY) and ~33 million average monthly rental unique visitors. Reach is amplified by syndication agreements with Realtor.com and Redfin — Zillow’s listings appear on rivals’ sites — plus partnerships with AppFolio and Esusu. (Fact — 10-K.)

Mortgages is Zillow Home Loans (available in 49 states + DC), which originates purchase loans and sells them into the secondary market for gain-on-sale, plus a mortgage-marketplace lead product. It is small ($199M) but strategically central: it is the “attach” that lets Zillow monetize the same shopper twice and is the financial backbone of the “housing super app.” (Fact — 10-K.)

Software / “Workspace.” Zillow has acquired a stack of agent SaaS: Follow Up Boss (CRM, acquired Dec-2023 for ~$400M), ShowingTime (tour scheduling), dotloop (transaction management), and Zillow Showcase (premium listing media). These are sold on subscription and, more importantly, deepen switching costs by embedding Zillow into the agent’s daily workflow. (Fact — 10-K.)

Recurring vs. transactional. The revenue base is a hybrid. Market-based Premier Agent and the software suite are recurring/subscription-like; Zillow Preferred and Mortgages are transactional (tied to closings); Rentals multifamily advertising is contract-based and recurring. There is no long-dated contractual backlog — revenue is a function of housing activity, traffic, and agent ad budgets, all of which can move quickly. The total addressable market management sizes at $202B (residential transaction fees $108B, mortgage origination $50B, title/escrow $18B, rentals advertising $17B, property-management software $9B). Against that, Zillow’s $2.6B of revenue represents a low-single-digit penetration — the bull’s headroom and the bear’s reminder that Zillow captures only ~1.8% of the ~$108B commission pool it sits adjacent to. (Fact — 10-K “Total Addressable Market.”)

The strategic thesis binding it together is the “housing super app” / integrated-transaction model, expressed geographically as Enhanced Markets: metros where Zillow stitches touring → financing → closing into one funnel served by Zillow Preferred agent partners and Zillow Home Loans officers. The wager is that owning more steps of one transaction multiplies revenue per transaction (the rising revenue-per-TTV metric) and raises switching costs for both consumers and agents.

Verdict. Zillow is the category-defining aggregator of U.S. home-shopping demand with an audience no rival matches and a genuinely improving monetization model (share-of-transaction, not just share-of-voice). It is a high-gross-margin (74%) advertising-and-software business, not a capital-heavy brokerage or an iBuyer — a materially better business model than either Compass (§ Competitive Position) or the wound-down Zillow Offers experiment. But it is fundamentally a toll-collector on housing activity that owns neither end of its own value chain — not the listing data (accessed via terminable MLS memberships) and not the top-of-funnel discovery (largely intermediated by Google). That structural dependency is the whole game, and it is exactly what is now under attack.

3. Industry Dynamics

The transaction pool is frozen, and that is not a passing cyclical dip. Roughly 4.8 million existing and new homes were sold in the U.S. in 2025 — near a multi-decade low and a level last seen in the mid-1990s despite a far larger population and housing stock. The cause is rate lock-in: a large majority of outstanding mortgages are fixed below 5%, so homeowners will not sell into 6–7% mortgage rates and forfeit their cheap debt. The result is a market where TTV was ~$2.2 trillion in 2025 (+3% YoY, essentially flat) and unit volume is stuck, even as prices remain elevated because the few available homes are competed over. (Fact — 10-K “Competition” and MD&A; NAR Economic Outlook Dec-2025, Census new-home sales.) Rental demand has moved the other way — rising vacancy and turnover, more advertising demand — which is precisely why Rentals is Zillow’s growth engine while For Sale grinds.

The profit pools. The industry’s transaction-fee pool (commissions) is ~$108B on ~5–6% blended commission rates; purchase-mortgage origination added >$1.3 trillion of volume and ~$50B of origination revenue; title/escrow ~$18B; rentals advertising ~$17B; property-management software ~$9B. (Fact — 10-K TAM, sourced to NAR, Fannie Mae, MBA, ALTA, Census, Fortune Business Insights.) The online-portal advertising profit pool — the slice Zillow actually competes for — is a fraction of the $108B commission pool: it is the share of agents’ and property managers’ marketing budgets that flows to digital lead-gen. Historically this was a highly attractive, high-margin pool because a handful of portals aggregated demand and agents had no better way to reach in-market buyers. That attractiveness is precisely what is now drawing in a flood of capital.

Regulation is a live, thesis-relevant force, not background noise. Three vectors:

1. The NAR commission settlement (finalized/implemented 2024). In response to the Burnett/Sitzer class actions, NAR agreed to rule changes decoupling buyer-agent compensation from the MLS and requiring written buyer-broker agreements before tours. The bear fear was that buyer commissions would collapse, shrinking the commission pool Zillow’s economics ride on. Two years in, the evidence is that blended commissions have compressed modestly, not collapsed — but the DOJ continues to investigate NAR, and the direction of travel is toward lower, more transparent, more negotiable buyer commissions. For a company whose best growth lever (Zillow Preferred) takes a percentage of the agent’s commission, structural commission compression is a direct headwind to revenue-per-transaction. (Fact/Interpretation — 10-K risk factors on NAR Settlement and DOJ.)
2. MLS / listing-access rules. Zillow does not own listing data; it accesses it through MLS memberships and listing-provider agreements that are terminable on limited notice and governed by ~500 different, inconsistent MLS rule sets. Rule changes — including changes “targeted…in response to our operations” — can restrict or cut Zillow’s data access. This is the supply-side chokepoint (see Competitive Position). (Fact — 10-K, “If we fail to comply with the rules…of MLSs.”)
3. Mortgage/consumer regulation. Zillow Home Loans operates under RESPA, TILA, ECOA, the Fair Housing Act, Dodd-Frank and CFPB oversight; the integrated “super app” (steering a Zillow shopper to Zillow’s own loan officer) sits squarely in RESPA anti-steering / referral-fee territory — and Zillow was in fact named in 2025 complaints alleging RESPA violations, among others (see below). (Fact — 10-K “Legal Proceedings.”)

Value chain and capital-cycle read (Marathon lens). The chain runs: listing data (MLS/brokers) → aggregation/display (portals: Zillow, Homes.com, Realtor.com, Redfin) → consumer discovery (increasingly Google / AI search) → lead delivery → agent/lender monetization → closing (title/escrow). Zillow occupies the high-margin aggregation/lead-delivery middle — but it is now being squeezed from both ends. On the supply side, brokerages (Compass) are experimenting with withholding listings from portals (“private exclusives”), and CoStar has built a rival aggregator. On the demand side, Google is inserting native home listings above the portals, threatening to disintermediate discovery. This is a textbook late-capital-cycle picture: historically high portal returns attracted an enormous inflow of competing capital — CoStar has poured well over $1 billion into Homes.com, Alphabet is entering with effectively unlimited resources, and Rocket rolled up Redfin (2025) — the classic Marathon warning sign that supernormal returns are about to be competed away. Barriers to entry (brand, data scale) are real against small entrants but do not exist against Google, which owns vastly more consumer intent and brand than Zillow. (Interpretation, capital-cycle framework, the relevant section.)

Verdict — structurally mixed, and deteriorating at the margin. The demand side of the industry (transactions) is in a multi-decade freeze with no clear catalyst to thaw, capping For Sale volume growth. The one genuinely attractive structural feature — a concentrated, high-margin portal-advertising oligopoly with Zillow as clear #1 — is under simultaneous assault from a deep-pocketed challenger (CoStar), a platform-owner turning predator (Google), and consolidation (Rocket/Redfin), while regulation (NAR/DOJ) pressures the commission pool the economics depend on. This is not a structurally good industry in the Greenwald sense of stable, defended profit pools; it is a good position inside an industry whose barriers are being tested by an influx of capital and a platform shift. Zillow’s near-term growth has to come from taking more revenue per transaction and from Rentals — because the transaction pie is not growing.

4. Competitive Position

Name the moat precisely. Zillow’s advantage is a two-sided demand-aggregation network effect fused with brand and data-scale, best classified in the Greenwald taxonomy as customer captivity (habit + brand) reinforced by economies of scale on the data/technology side — not a classic locked-in network effect. The mechanism: consumers go to Zillow first because it has the most listings and the best-known brand (“Zillow” > “real estate” in search); that audience concentration is what agents and property managers pay to reach; the resulting revenue funds product and data investment (the 173-million-home “living database,” the Zestimate, 224 patents), which improves the consumer experience and reinforces the habit. It is a real flywheel and it has produced a genuine financial outcome — Zillow monetizes its audience at a multiple of any rival despite comparable raw traffic (see below). (Fact/Interpretation — 10-K “Competitive Advantages.”)

But pressure-test it hard, because two of its three legs are contestable. A true network effect locks users in because the value depends on other users being present and they cannot easily replicate the network elsewhere. Zillow’s does not clear that bar cleanly: agents multi-home (they advertise on Zillow and Realtor.com and buy Google leads), consumers multi-home (a buyer will use whatever site surfaces the listing they want), and Zillow owns neither the listings nor the discovery layer. What Zillow really has is a habit and brand advantage sitting on top of someone else’s data and someone else’s traffic funnel. That is a strong position, but it is an aggregator’s position, and aggregators are vulnerable to whoever controls the layer above them. The layer above them is Google.

Threat 1 — Google inserting native listings (the June 2026 catalyst that halved the stock). On June 11, 2026, Google took its home-listing search format (Home-Listing Local Services Ads, powered by a HouseCanary partnership) nationwide to all 50 states, surfacing price, photos, and features directly at the top of mobile search results. Because a large share of home-search sessions begin at Google, a native listing unit above the Zillow link is a direct disintermediation threat to Zillow’s top of funnel — the demand-aggregation leg of the moat. The market read it as existential: Zillow fell sharply on the news and the stock has roughly halved from its 2025 highs (~$68 → ~$34). (Fact — The Real Deal, TechRepublic, Inman, Benzinga, 2026-06-11/16; price move per CSV.) The nuance that keeps this a threat rather than a confirmed break: the format is currently fed by only three MLSs, and Zillow states pilot-market data showed no meaningful decline in its traffic or search interest. (Fact — Inman 2026-06-16; note this is management’s claim, a hypothesis to be validated against future traffic prints — Open Question.) The point is not that the moat has already fallen; it is that the market has re-rated the durability of a moat whose top-of-funnel was always rented from Google, and Google has now demonstrated both the intent and the capability to compete for it. Alphabet has infinitely more brand, data, and consumer intent than Zillow; if it presses, Zillow cannot out-aggregate it. This is the single most important fact in the entire thesis.

Threat 2 — CoStar / Homes.com, and why traffic ≠ moat. CoStar has spent enormous sums — outspending Zillow ~4x (~$412M) on advertising in 2023 and running tens of thousands of commercials — to build Homes.com into the #2 U.S. portal by traffic (~104M monthly uniques in Q1-2025) on a free “your listing, your lead” model that sends the buyer inquiry to the listing agent rather than a paying advertiser. Here is the decisive competitive fact: despite roughly half of Zillow’s traffic, Homes.com monetizes at a tiny fraction of it. CoStar’s entire North American residential segment revenue was ~$16.5M in Q1-2025 (~$66M annualized), against Zillow’s $1.9B of For Sale revenue. (Fact — CoStar Q1-2025 results via Inman/RISMedia 2025-04; Zillow 10-K.) This is the strongest evidence that Zillow’s real moat is not raw traffic but the monetization apparatus — the agent-advertiser relationships, the Premier Agent auction, the Zillow Preferred success-fee model, the embedded CRM/software. CoStar bought an audience but has not been able to convert it into agent ad dollars, because agents pay for validated buyer leads they can convert, and that is a relationship and product problem, not a traffic problem. Zillow’s monetization edge is durable against CoStar. It is far less clearly durable against Google, which can both aggregate demand and sell the ad unit.

Threat 3 — losing listing access (Compass private-exclusives / the “Zillow ban”). In April 2025 Zillow imposed Listing Access Standards: a home marketed publicly for more than one day without being on the MLS is banned from Zillow. Compass — the largest U.S. brokerage — was building a “three-phase” private-exclusive strategy (private → coming-soon → MLS) and sued Zillow in June 2025 for antitrust, alleging Zillow was protecting a monopoly. The sequence since matters for the thesis: a New York federal judge denied Compass’s preliminary injunction in February 2026, finding “the record fails to show that Zillow possessed the power to exclude competition from the online home search market”; Compass dropped its suit in March 2026 after Zillow introduced a “Preview” product; and Zillow countersued Compass and the Chicago-area MLS (MRED) in May 2026. (Fact — NAR/Courthouse News/Inman/Real Estate News, Feb–Jul 2026.) Net read: Zillow won the near-term skirmish and preserved its all-listings comprehensiveness, but the episode exposed the deeper vulnerability — Zillow’s product depends on universal listing access it does not own, and the largest brokerages have both the incentive and, increasingly, the scale to withhold inventory. Ironically, the same court finding that Zillow lacks monopoly power (good for the antitrust defense) undercuts any claim to a dominant, defensible moat.

Direct comparison.

• vs. Homes.com/CoStar: Zillow wins decisively on monetization (~30x the residential revenue on ~2x the traffic); CoStar’s deep pockets keep it a permanent margin-pressuring irritant but not, so far, a share-of-wallet taker.
• vs. Realtor.com (News Corp/Move): the historical #2 by monetization, now arguably #3 by traffic behind Homes.com; a rational, less-disruptive competitor.
• vs. Redfin (now Rocket, July 2025): Redfin’s ~49M-visitor portal is now inside Rocket’s mortgage-servicing-and-origination flywheel (Rocket also bought Mr. Cooper, Oct-2025) — a vertically integrated “search-to-servicing” rival that competes with Zillow’s own super-app ambition, especially on the Home Loans attach.
• vs. Compass: a no-moat brokerage (per prior coverage) that is nonetheless a supply-side threat via listing-withholding.

Verdict — a durable monetization advantage sitting on an eroding aggregation foundation. Zillow’s moat against every portal competitor is intact and, on the evidence, wide — no rival converts audience into agent revenue anywhere near as efficiently. But the moat is not the classic defended-network-effect the bull narrative implies; it is brand + habit + a superior monetization stack layered on rented listing data and, critically, a rented top-of-funnel. The Google entry attacks the one leg Zillow can least defend. The correct characterization is a strong incumbent aggregator whose barriers are being actively tested by the one competitor (Google) it cannot out-scale — durable today, structurally exposed tomorrow. If a senior investor asks “what deteriorates without the moat,” the answer is concrete: if Google captures even 15–20% of top-of-funnel discovery, Zillow’s connection volume and revenue-per-TTV both compress, and the whole super-app economic case weakens.

5. Growth History and Forward Opportunities

The five-year revenue picture is distorted by the Zillow Offers wind-down and must be read on continuing operations. Zillow’s headline revenue collapsed from the iBuying era (Zillow Offers, the “Homes” segment, generated multi-billion-dollar gross revenue by buying and flipping houses) after management shut Zillow Offers in late 2021, taking write-downs and moving it to discontinued operations (2021 net income included a ~$630M discontinued-ops loss). Read on the continuing marketplace business, the trajectory is: FY2021 $2,132M → FY2022 $1,958M → FY2023 $1,945M → FY2024 $2,236M → FY2025 $2,583M. The 2022–2023 trough reflects the mortgage-rate shock freezing transactions and gutting the refinance-heavy mortgage business; the 2024–2025 re-acceleration (+15%, +16%) reflects Rentals scaling and For Sale share gains, not a housing recovery — TTV was flat. (Fact — DATA_BRIEF/ROIC; 10-K.)

Revenue by category, the durable read (2023→2025):

(Fact — 10-K income statement.)

Rentals is the growth engine, and it is high-quality growth. Rentals grew +39% to $630M in FY2025 (24% of revenue, up from 20%), driven by a 58% jump in multifamily revenue as multifamily property count rose +44% to 72,000 and property managers upgraded to richer advertising packages — revenue per rental visitor rose 31% to $4.77 on only 6% visitor growth. This is the best kind of marketplace growth: monetization deepening faster than traffic, funded by a genuine product/inventory advantage (largest U.S. rental listing set) and amplified — capital-light — by syndication to Realtor.com and Redfin. Rentals has a long runway: there is no MLS for rentals, so aggregating the ~50 million U.S. rental units is a hard, defensible problem Zillow is winning, and the rentals advertising TAM (~$17B) plus property-management software (~$9B) dwarfs current revenue. This is the most credible durable-growth story in the company. (Fact — 10-K; Interpretation.)

For Sale growth is share-of-transaction, not volume. With unit sales frozen, Residential grew only 7% — but the quality signal is that Zillow grew despite flat TTV by lifting For Sale revenue per TTV to 8.5 bps from 7.9 bps (+8%). The mechanism is the Enhanced-Markets / Zillow Preferred flywheel: convert more of the audience into paid connections, attach Zillow Home Loans, and capture a commission-linked success fee. Management says Enhanced-Market connections exited 2025 >40% of total, with a runway to the majority. This is the bull case’s core: even a frozen housing market can yield double-digit For Sale growth if Zillow keeps taking a bigger slice per transaction. The bear caveat is that revenue-per-TTV expansion (a) is partly because commissions are the base — NAR-driven commission compression works against it — and (b) is inherently self-limiting; you cannot raise your take-rate on agents forever before they push back, and CoStar’s free model is explicitly designed to make them push back.

Zillow Home Loans is the highest-leverage forward option — and the least proven. Mortgages grew +37% to $199M on an origination ramp, with “double-digit customer adoption” of Home Loans across Enhanced Markets. If Zillow can convert even a modest share of its 235M-user audience into purchase-mortgage originations, the attach economics (monetizing the same shopper twice, capturing gain-on-sale) are powerful and would materially raise revenue-per-transaction. But it is sub-scale versus Rocket (now armed with Redfin and Mr. Cooper’s $2.1T servicing book), it is capital- and rate-cycle-sensitive, and the RESPA/steering scrutiny around routing Zillow shoppers to Zillow’s own loan officers is a real regulatory ceiling. High-potential, unproven, contested. (Fact/Open Question — 10-K; Rocket disclosures.)

Organic vs. acquired. Growth has been a blend. The software/CRM layer was bought — Follow Up Boss (~$400M, Dec-2023), ShowingTime, dotloop, Aryeo, Virtual Staging AI — a deliberate roll-up to embed Zillow in agent workflows and raise switching costs. But the two live growth vectors — Rentals and For Sale revenue-per-TTV — are predominantly organic, product- and monetization-driven, not M&A-bought. That is a favorable mix: the growth that matters is not being manufactured by acquisition accounting.

Verdict — genuinely high-quality growth in Rentals and share-of-transaction, but riding on a contested foundation. The Rentals engine is durable, capital-light, and economically real (deepening monetization, defensible inventory advantage). The For Sale share-gain engine is real but bounded — it depends on an ever-rising take-rate against agents in a market where a free-lead competitor (CoStar) and commission-compressing regulation (NAR/DOJ) both push the other way, and on a top-of-funnel Google is now attacking. Home Loans is the swing option: large if it works, unproven and regulator-exposed. The growth is not ad-spend-bought or M&A-manufactured — it is product-led — which is a real positive. But it is not the growth of a widening moat; it is a strong operator squeezing more revenue out of a frozen, increasingly contested market. High-quality growth on a thinning foundation is the honest characterization.

6. Financial Quality

Zillow’s income statement tells two stories at once, and the tension between them is the whole analytical problem. On the surface the top line has re-accelerated impressively: revenue grew 16% to $2,583M in FY2025 (from $2,236M in 2024, $1,945M in 2023), materially outpacing a residential real-estate transaction market still frozen near multi-decade lows (Fact, FY2025 10-K MD&A, “Financial Overview”). Beneath that, the business has never earned its cost of capital, carries an accumulated deficit of −$1,859M, and funds a share-based-compensation bill that in most years has exceeded its entire free cash flow. The job of this section is to reconcile the genuine operating-leverage story with the equally genuine quality-of-earnings problem.

Revenue composition and the deliberate margin give-up. Management now reports one segment with four revenue categories. FY2025 splits (10-K MD&A revenue table) are: Residential $1,704M (66% of revenue, +7%), Rentals $630M (24%, +39%), Mortgages $199M (8%, +37%), and Other $50M (2%). “For Sale” (Residential + Mortgages) is $1,903M (74%). Residential — legacy Premier Agent advertising plus ShowingTime, Zillow Showcase, new-construction and Follow Up Boss — remains the profit engine, and Residential revenue-per-visit rose 4% to $0.178 (Fact, MD&A). But the growth is being led by the two lowest-margin categories. Rentals, growing at nearly 40%, is now a quarter of revenue; Mortgages (Zillow Home Loans) is the fastest-growing “For Sale” line off a small base.

That mix shift is the mechanical driver of the gross-margin slide, and it is a choice, not an accident. Gross margin has compressed every year: 84.8% (2021) → 81.3% (2022) → 78.4% (2023) → 76.4% (2024) → 74.1% (2025) (Fact, ROIC/10-K). In FY2025 cost of revenue rose 27% ($141M) to $668M against revenue up only 16% — and the 10-K names the culprits precisely: +$87M of lead-acquisition costs tied to the Redfin rentals syndication agreement, +$18M D&A, +$17M ad-serving to support the rentals marketplace, and higher mortgage loan-processing costs (Fact, MD&A “Cost of Revenue”). Rentals and Mortgages carry direct variable acquisition/origination costs that the pure-advertising Premier Agent model never did. Interpretation: this is Zillow deliberately trading gross-margin percentage for a larger, more diversified revenue base and deeper transaction attach — defensible strategically, but it means the “74% gross margin” headline is structurally drifting toward the low-70s and should be modeled down, not up, as Rentals and Mortgages keep out-growing Residential.

The GAAP-loss vs. Adjusted-EBITDA gap — and what closes it. Zillow remained GAAP-unprofitable at the operating line through 2024 and only just crossed into black net income in 2025. Operating loss narrowed from −$247M (2023, −12.7% margin) → −$190M (2024, −8.5%) → −$34M (2025, −1.3%), and net income swung to +$23M (first GAAP profit since 2021), aided by $77M of “other income, net” (interest on the cash pile). Against that, the company’s headline Adjusted EBITDA was $622M (24% margin) in 2025, up 25% from $498M (22%) in 2024 (Fact, MD&A Adjusted-EBITDA reconciliation). The ~$656M gap between the −$34M operating loss and $622M Adjusted EBITDA is bridged overwhelmingly by two add-backs: D&A of $264M and share-based compensation of $390M. SBC alone is 63% of the bridge. This is the crux: Zillow’s flagship profitability metric is “profitable” primarily because it excludes the single largest real cost of running the business. The company’s own disclosure concedes “Adjusted EBITDA does not consider the potentially dilutive impact of share-based compensation” (10-K) — a limitation that, for this issuer, is not a footnote but the headline.

The SBC problem, quantified. Share-based compensation expensed through the P&L has run $312M (2021), $451M (2022), $451M (2023), $448M (2024), $390M (2025) — i.e. 23.2% of revenue in 2023, 20.0% in 2024, and 15.1% in 2025 (Fact, cash-flow statement / computed). And that understates the true grant cost: Zillow capitalizes $62–73M of SBC per year into software ($62M in 2025, $72M in 2024, $73M in 2023; Fact, cash-flow “Noncash transactions”), so total SBC granted was $452M in 2025 (per the shareholders’-equity statement) — the reported $390M expense is net of the capitalized slug. Two honest points cut in opposite directions. First, the fair one: SBC is falling both absolutely (from $451M to $390M) and as a share of revenue (from 23% to 15%) — real progress, driven partly by the August-2022 equity-award actions rolling off. Second, the damning one: even at 15% of revenue, SBC ($390M) is ~3.9x reported free cash flow and is a genuine transfer of ownership from existing holders to employees that must be either bought back (with cash) or absorbed as dilution.

Buybacks that offset dilution rather than return capital — the net share count. Here the two problems collide. Zillow spent $1,395M on repurchases across 2023–2025 ($424M + $301M + $670M). Yet the share count rose: from 234.3M (Jan 1 2023) to 240.4M (Dec 31 2025), +6.1M shares (+2.6%) despite that $1.4B outlay (Fact, Consolidated Statements of Shareholders’ Equity). RSU vesting added ~6.1–6.4M shares each year, option exercises another 1.8–5.5M, and the 2024/2026 note settlements issued 6.5M in 2024 — collectively swamping repurchases. Only in 2025 did the diluted count finally decline (242.5M → 240.4M, −2.0M) as the larger $670M buyback plus a −3.1M capped-call share settlement outran issuance. Interpretation: over the full three-year window, $1.4B of buybacks did not even hold the line flat — they were an incomplete SBC-dilution offset, not a return of capital. Calling this program “capital return” flatters it; it is the cash cost of the stock-comp model, paid on the open market.

And the buybacks were done at the wrong price. Weighted-average repurchase prices were ~$44.5 (2023), ~$42.4 (2024), and ~$70.5 (2025) — the 2025 buyback alone bought 8.1M Class A at $70.11 and 1.4M Class C at $73.19 (Fact, Note 10 repurchase table). Against today’s $33.90, the three-year blended repurchase price of roughly $53 implies the company retired stock ~55% above where it now trades; the 2025 tranche is down ~52%. Buying back the most stock ($670M) at the highest price ($70+), just before a ~50% decline, is the textbook value-destructive buyback pattern. (The one redeeming note, covered in the relevant section, is the March 2026 $1.25B re-authorization at a far lower price.)

FCF quality. Reported operating cash flow was $368M in 2025 (down from $428M in 2024, as a $227M build in mortgage loans held for sale consumed working capital). Total capital spend was $267M — $133M PP&E plus $134M “purchases of intangible assets” (up sharply from $28M, and embedding the $100M initial Redfin rentals-partnership payment plus capitalized software). Simple FCF ≈ CFO $368M − capex $267M ≈ $101M. The problem is scale: FCF of ~$101M against SBC of $390M means economic free cash flow net of the real stock-comp cost is roughly −$289M. The business does not yet self-fund its own compensation. Management instead promotes “Adjusted free cash flow” of $420M, reconciled by adding back $219M of net warehouse borrowings (defensible — a pass-through that funds loans-held-for-sale) and the $100M Redfin payment as “one-time and nonrecurring” (aggressive — it is a real cash outlay for a growth partnership) (Fact, MD&A AFCF reconciliation). Ex the Redfin add-back, AFCF is ~$320M; and no version of it charges the $390M SBC. Additionally, capitalizing $62M of SBC and $134M of development spend shifts cost off the current P&L and defers it into amortization — a legitimate but flattering accounting choice that inflates near-term reported margin and Adjusted EBITDA.

ROIC/ROE and the balance sheet. Returns remain the weakest part of the story: FY2025 ROE is ~0.5% ($23M net income on ~$4,866M average equity), ROIC is essentially zero-to-negative, and on a cost-of-capital basis Zillow has never cleared the bar — the −$1,859M accumulated deficit is the cumulative scoreboard. GAAP P/E (~135x) is meaningless and should be ignored (use EV/Sales and EV/Adjusted-EBITDA instead). The offsetting positive is a genuinely clean, near-fortress balance sheet — and here I correct the working brief: as of 12/31/2025 the convertible notes are fully settled ($0 outstanding); the only debt is $364M of mortgage-warehouse borrowings, fully matched by $386M of loans held for sale (a self-liquidating pass-through). Against $768M cash + $527M short-term investments = $1,295M, Zillow is effectively debt-free with ~$1.3B of net cash — management itself stated it became “debt-free, other than credit-facility borrowings for Zillow Home Loans” in May 2025 (Fact, 8-K 2025-05-16). Book equity is $4,884M, but 58% of it is goodwill ($2,823M, flat YoY — no impairment); net intangibles are $279M, leaving tangible book equity ≈ $1,782M (~$7.4/share). (The brief’s $1,463M used gross intangibles; the balance sheet’s net figure gives $1,782M.) The goodwill is mostly the legacy Trulia deal plus ShowingTime and Follow Up Boss — accounting equity that reflects past M&A, not tangible earning assets.

Quality-of-earnings verdict. Earnings quality is low but improving. The positives are real: cash generation is positive, the balance sheet is pristine, SBC is falling as a share of revenue, and there are no accrual red flags (no impairments, contingent consideration marked conservatively, net income and CFO directionally consistent). The negatives are structural: the flagship profitability metric (Adjusted EBITDA) is manufactured by excluding the largest real cost; “adjusted FCF” is further embellished by adding back a genuine growth outlay; GAAP profitability is razor-thin and flattered by interest income and cost capitalization; and buybacks are a dilution treadmill dressed as capital return.

Verdict — do economics improve with scale? Yes, mechanically, but not yet enough to earn the cost of capital. The operating-leverage evidence must be weighed fairly, and it is strong: operating margin improved from −12.7% to −1.3% and Adjusted-EBITDA margin from 22% to 24% while revenue grew 33% over two years, with sales-and-marketing up only 7% and G&A actually declining — the incremental-margin structure of an advertising-and-marketplace platform is clearly present. Fixed technology and brand costs are being spread over a bigger revenue base. But the same period shows the business still cannot cover fully-loaded (SBC-inclusive) costs, ROIC/ROE remain negative-to-negligible, gross margin is structurally eroding as the growth mix skews to lower-margin Rentals/Mortgages, and free cash flow net of stock comp is negative. The correct read is not “bad business” but “unproven unit economics on a full-cost basis”: operating leverage is appearing and the trajectory is genuinely positive, yet Zillow has not demonstrated it can convert its traffic dominance into returns above its cost of capital. Until GAAP operating income and SBC-inclusive FCF turn durably and materially positive, the “economics improve with scale” thesis remains a trajectory, not an achievement.

7. Capital Allocation

Zillow’s capital-allocation record is a study in a management team that is intelligent, founder-aligned, and financially conservative — yet has presided over one catastrophic strategic bet and a buyback program that, however well-intentioned, has destroyed value at the margin. The verdict has to weigh a genuinely disciplined balance sheet against a history of expensive M&A and mis-timed repurchases.

The defining error: Zillow Offers / iBuying. Between 2018 and 2021 Zillow built an algorithmic home-flipping business (Zillow Offers, the “Homes” segment) that bought, renovated, and resold houses off its own balance sheet. In November 2021 management abruptly wound it down, citing an inability to forecast home prices accurately enough to run the model at scale. The financial damage was enormous: Zillow disclosed a Homes-segment loss of roughly $880M pre-tax in FY2021, including ~$300M+ of inventory write-downs, and the 2021 net loss carried ~$630M of discontinued-operations charges (Fact/Interpretation — FY2021 disclosures, outside the 5-year primary window; magnitude widely reported). The wind-down triggered a securities class action (amended May 2022) alleging materially false statements about Zillow Offers, and shareholder derivative suits (2022) alleging the board breached its fiduciary duties over the losses (Fact, FY2025 10-K legal-proceedings note). Interpretation: this was a large, unforced, capital-intensive error — a light-capital advertising platform levering its balance sheet into a spread business it did not understand. That management recognized the mistake and exited decisively is to its credit; that it made the bet at all, with shareholder money, is the single biggest black mark on the record and the reason to treat any future balance-sheet-intensive pivot (e.g., deeper mortgage-balance-sheet risk) with skepticism.

M&A: mostly bolt-on “software-around-the-transaction,” at full but not reckless prices. Post-iBuying, capital deployment shifted to tuck-in acquisitions that deepen the agent/consumer workflow: ShowingTime (~$500M, 2021) for tour scheduling; dotloop earlier for transaction management; and most significantly Follow Up Boss (Enchant, LLC), December 2023, for $399M cash net of cash acquired plus up to $100M of contingent consideration over three years (Fact, 10-K Note; total potential ~$499M). Smaller deals (Aryeo listing media, Virtual Staging AI, Out East, Spruce/closing) round out the “housing super-app” build. Cash paid for acquisitions was $433M in 2023, ~$7M in 2024, nil in 2025. Interpretation: strategically coherent — each acquisition attaches Zillow more tightly to the transaction and cross-sells into its traffic — but the prices are full (Follow Up Boss at a rich revenue multiple for a CRM), and the returns are unproven; the $2,823M goodwill balance is the cumulative price paid for growth that has yet to show up in ROIC. Notably, integration has not produced impairments, and Zillow trimmed the Follow Up Boss earnout accrual as performance came in below the aggressive case (a $3M G&A benefit in 2025) — a sign of conservative deal accounting rather than empire-building.

Buybacks: the right instinct, the wrong prices (mostly). Zillow has authorized a cumulative $3.5B of repurchases ($2.5B originally, raised $1B in May 2025), of which $711M remained at year-end 2025 — implying ~$2.8B deployed to date (including some convertible-note repurchases). As detailed in the relevant section, the $1,395M spent in 2023–2025 came at a blended ~$53/share versus $33.90 today, and the largest tranche ($670M in 2025) was executed at ~$70 — value-destructive in hindsight and, worse, largely an SBC-dilution offset rather than a genuine per-share-value return (the share count still rose 2.6% over the period). This is the central capital-allocation critique: the company bought high, and bought to mop up its own stock issuance. The redeeming counter-evidence is timing-improved: on March 4, 2026 the board authorized an additional $1.25B (Fact, 8-K 2026-03-05) with the stock in the mid-$30s–low-$40s — buying near the lows is far more defensible and, if executed there, will finally be accretive. Interpretation: the buyback philosophy (return excess cash, exploit weakness) is sound; the execution track record through 2025 was poor, and the 2026 authorization is the test of whether they have learned to buy low.

Liability management: genuinely well done. Against the iBuying black mark, Zillow’s handling of its convertible-note stack deserves real credit. It carried well over $1.6B of low-coupon converts (0.75% due 2024, 1.375% due 2026, 2.75% due 2025). It settled the 2024 Notes (Sept 2024), 2026 Notes (Dec 2024), and the remaining $419M of 2025 Notes (May 2025) — mostly in cash, from its own liquidity — unwound the associated capped calls in August 2025 for +$38M of proceeds, and put in place a fresh $500M revolving credit facility (January 2026, maturing 2031) for flexibility (Fact, 10-K; 8-Ks 2024-09-04, 2024-12-19, 2025-05-16, 2025-08-25, 2026-01-30). The result is an issuer that entered a housing downturn effectively debt-free with $1.3B of net cash — exactly the balance-sheet posture you want in a cyclical trough, and a marked contrast to iBuying-era balance-sheet risk.

R&D and S&M intensity. Zillow spends heavily on product and go-to-market, as a platform must: Technology & development $607M (23% of revenue) and Sales & marketing $843M (33%) in 2025 (Fact, MD&A). The encouraging signal is the trend — both grew far slower than revenue (T&D +4%, S&M +7% vs. revenue +16%), and G&A fell 5% — the source of the operating leverage. This is disciplined; management is not buying growth by re-accelerating opex.

SBC as capital allocation, incentives, and founder control. The compensation architecture is the most consequential capital-allocation choice, because it is the largest cash/dilution cost and it is unusually structured. Zillow runs no formal cash-bonus program; executive pay is base salary plus time-vested RSUs and market-priced stock options — with no performance/TSR-based PSUs (Fact, DEF 14A CD&A; “we have not established a formal cash incentive program”). CEO Wacksman earned $7.1M in 2025 (down from $15.7M in 2024); co-founders Barton and Frink took ~$6.85M each, entirely in options; Barton’s 2023 package was $26.4M (a large option grant). Interpretation: pure time-and-price-based equity aligns leadership with the share price over the long run — and notably, the founders’ large option grants at strikes of $49–$135 are now deeply underwater ($34 stock), so insiders only monetize them if the stock recovers, which is real alignment. The flip side is twofold: (1) the absence of operating-metric-linked incentives means no explicit accountability for margins, returns, or capital efficiency — only the stock price; and (2) time-based equity is precisely what drives the $390–450M annual SBC bill and the buyback treadmill. Governance is founder-controlled: Class B shares carry 10 votes each, and Barton (36.2% of votes) plus Frink (23.5%) — with all officers/directors at 60.6% of total voting power on ~6.2M Class B shares — hold entrenched control despite a modest economic stake (Fact, DEF 14A “Security Ownership”). Outside holders (Caledonia 27% of Class A, Vanguard, FMR) have little say. Dual-class control concentrates capital-allocation authority in the two founders who both built the traffic moat and authored the iBuying debacle.

Verdict — has management allocated capital intelligently? Mixed, trending better, but the record is blemished by a major unforced error. The honest scorecard: one large value-destroying bet (iBuying, ~$500M–$880M+ of losses and ongoing litigation) that permanently colors the risk assessment; value-destructive buybacks through 2025 ($1.4B at ~$53 vs. $34, and largely a dilution offset); and full-priced, unproven M&A ($2.8B of goodwill not yet earning its keep). Against that: exemplary liability management (fully deleveraged into a downturn, $1.3B net cash); disciplined opex (growth without opex re-acceleration); conservative deal accounting (no impairments, trimmed earnouts); genuine founder alignment via underwater equity; and a better-timed 2026 buyback authorization. This is not a management team that squanders cash carelessly — it is financially prudent day-to-day — but it is a team whose two biggest capital decisions (iBuying and the 2023–25 buyback) both destroyed value, and whose dual-class control removes the external check that might have prevented them. Net: capital allocation is adequate and improving, not exemplary — credit the balance-sheet discipline, but do not forget the iBuying scar or the buyback-at-the-top pattern.

7A. SEC Filings Sweep & Insider Read

Insider transactions — the Form 4 read. The corpus holds 375 Form 4s plus 13 Form 4/A over the trailing five years, with heavy clustering on specific dates (2025-03-03/04, 2025-08-11–15, 2026-03-03/04, 2026-05-15) — the fingerprint of routine RSU-vesting and same-day option-exercise-and-sell activity, not discretionary trading. I sampled ~17 filings across the last 18 months covering every named officer and director: co-founders Barton and Frink, CEO Wacksman, CFO-org (CAO Rock, GC Owens), COO Choo, and outside directors Blachford, Gurley, Cormier Thielke, Stephenson.

The signal is unambiguous and one-directional: zero code-P open-market purchases across the entire sample. Every transaction was either an option exercise (code M) at legacy strikes of ~$21–$32 immediately paired with a market sale (code S), a routine RSU-vest disposition, or a director equity grant (code A). The large majority of sales were executed under Rule 10b5-1 plans (diversification, not conviction). Representative activity: Frink (Co-Executive Chairman & President) systematically exercised and sold via 10b5-1 throughout the window — selling into the $80s in mid-2025 and continuing down into the $44–$55 range in early 2026, while still holding ~2.23M shares; Wacksman sold only 9,072 shares at $36.54 (May 2026, 10b5-1) and owns very little stock outright (468 Class A plus mostly-unvested Class C); director Gurley received a 18,996-share award (comp, code A). Interpretation: the complete absence of any discretionary open-market buying — even with the stock down ~50%+ and trading at $34 — is a notable non-endorsement; no insider has stepped in to buy the decline. The mitigant is that insiders’ unvested options/RSUs carry strikes of $49–$135, now deeply underwater, so leadership retains substantial leveraged upside exposure and is not “cashing out” its core stake, merely monetizing vested legacy grants on schedule. Net insider direction: 100% net selling, entirely routine/10b5-1, with a conspicuous absence of conviction buys.

8-K material-event timeline (trailing ~5 years).

• Nov 2021: Zillow Offers wind-down announced (iBuying exit) — the pivot that reset the entire capital structure and strategy.
• Jan 6, 2024: J. William Gurley (Benchmark) appointed to the board (8-K 2024-01-09, Item 5.02).
• Aug 7, 2024: Q2 earnings and the CEO transition — Jeremy Wacksman named CEO; Rich Barton to Co-Executive Chairman alongside co-founder Frink (8-K 2024-08-07, Items 2.02 + 5.02; corroborated by DEF 14A “CEO since August 2024”). CFO Jeremy Hofmann had been appointed May 2023.
• Sep 2024 & Dec 2024: Settlement of the 2024 and 2026 convertible notes (8-Ks, Item 3.02 unregistered equity issuances on conversion).
• May 15, 2025: Final settlement of the 2025 Notes — company declares itself “debt-free” ex-warehouse (8-K 2025-05-16, Item 8.01).
• May 2, 2025: Buyback authorization raised by $1B to $3.5B cumulative.
• Aug 25, 2025: Capped-call unwind with Morgan Stanley/Citi/Barclays/JPMorgan (+$38M proceeds; 8-K 2025-08-25).
• Jan 30, 2026: New $500M revolving credit facility (maturity 2031) entered (8-K 2026-01-30, Items 1.01/2.03).
• Mar 4, 2026: Board authorizes an additional $1.25B of repurchases (8-K 2026-03-05, Item 8.01) — at a far lower share price than the 2023–25 buybacks.
• Routine Item 5.07 8-Ks (annual-meeting votes) in Jun-2024, Jun-2025, Jun-2026.

Litigation / one-time items distorting run-rate. The active legal backdrop: (1) Zillow Offers securities class action (amended May 2022) and shareholder derivative suits (2022) over the iBuying losses — legacy, but unresolved reputational/contingent-liability tail; (2) the industry-wide NAR commission settlement (finalized 2024), which changed how buyer-broker commissions are offered/negotiated — a structural risk to agent marketing budgets (Zillow is not a named party but is exposed); and (3) per external reporting, a 2025 Compass antitrust suit challenging Zillow’s private/exclusive-listings (“Zillow ban”) policy — Open Question, to confirm against the legal-proceedings note; flagged from news, not yet reconciled to a filing here. On one-time items: FY2021 is distorted by the ~$630M iBuying discontinued-ops charge; FY2024–25 results are relatively clean, though note that $62M of capitalized SBC and $134M of capitalized development/partnership spend shift cost off the current P&L, and the $100M Redfin payment is treated as one-time in Adjusted FCF — analysts should normalize these when assessing run-rate cash economics. Goodwill of $2,823M has not been impaired, and contingent-consideration marks have been conservative (earnout trimmed).

Verdict (insider/filings input to the relevant section): the filing record corroborates the fundamental read — a fully-deleveraged, founder-controlled platform emerging from its iBuying mistake, with clean recent filings but a total absence of insider buying and 60.6% founder voting control that limits external accountability. Nothing in the Form 4 or 8-K record contradicts the thesis; the insider tape is neutral-to-slightly-negative (steady 10b5-1 selling, no conviction buys), and the 8-K timeline confirms both the balance-sheet cleanup (positive) and the concentrated-control governance structure (a risk).

8. Changes and Headwinds — Last Two Years

The last twenty-four months reshaped Zillow along three axes at once: a leadership handoff, a decisive strategic pivot from a lead-generation advertising portal toward an integrated “housing super app,” and an unusually hostile external environment — a frozen transaction market, a nest of antitrust and IP litigation on both sides of the “v.”, and, in June 2026, the first genuinely structural threat to the aggregation model in Google. The stock has responded violently: the Class C shares (Z) fell from roughly ~$68 in mid-2025 to $33.905 at the 2026-07-17 close, a decline of approximately 50% over twelve months (FACT, price CSV / ROIC stale pr_last). What follows separates the corporate changes management controls from the exogenous headwinds it does not.

Leadership: Barton steps back, Wacksman steps up. Co-founder Rich Barton, who had returned as CEO in 2019 to run the Zillow Offers iBuying experiment (and its 2021 wind-down), handed the chief-executive role to long-time insider Jeremy Wacksman in August 2024 and moved to Executive Chair (FACT). This is a continuity handoff, not a break — Wacksman was Zillow’s president and had run product and marketing for years — but it matters for governance because Barton and co-founder Lloyd Frink retain super-voting Class B stock, so control does not travel with the CEO title (FACT; the relevant section). The bull reading is that Wacksman is the operator executing Barton’s platform vision without the founder’s appetite for moon-shot capital destruction; the bear reading is that the same super-vote that let Barton bet ~$1B+ of shareholder value on iBuying remains intact. Interpretation: the transition is thesis-neutral to mildly positive — steadier hands, unchanged control.

The strategic pivot: from Premier Agent portal to integrated transaction platform. The defining change of the period is the deliberate migration of the core Residential business from a market-based-pricing (MVP) lead auction toward “enhanced markets” — an integrated funnel that couples Zillow Preferred agent connections with Zillow Home Loans, ShowingTime touring, dotloop closing, and the Follow Up Boss CRM (acquired end-2023 for ~$400M). Enhanced markets reached 49% of connections in Q1 2026, up from 44% in Q4 2025, on the way to a stated ≥75% target (FACT, Q1’26 call, 2026-05-06). The financial signature is visible: For-Sale revenue grew 12% YoY in Q1’26 against a housing market NAR reported up only ~2%, and mortgages revenue grew 56% YoY on 96% growth in purchase-loan origination volume to a record $1.5B, making Zillow Home Loans a “top-25 purchase lender” (FACT, management; hypothesis on durability, the relevant section rule 8). Interpretation: this is real share-of-wallet capture and the clearest evidence that Zillow can grow through a flat market — the single most important positive change of the period.

Rentals: the growth engine — and the legal lightning rod. Rentals revenue grew 39% in FY2025 and 42% YoY in Q1’26 (multifamily +57%), reaching 76,000 multifamily properties and 2.7M average monthly listings, with a stated path to $1B+ annual revenue (FACT). Two syndication deals powered this: the February 6, 2025 Redfin partnership — under which Zillow paid Redfin ~$100M to exit the rental-listings business and syndicate Zillow’s inventory — and a May 2026 “Preview” collaboration with Realtor.com (FACT, web + call). The Redfin deal is now the company’s most serious live legal exposure (below). Interpretation: Rentals is the highest-quality growth in the business, but management chose to buy competitive exit rather than out-compete, and that choice is what drew the FTC.

The listing-access war: Zillow ban, Compass, and the reversal. In April 2025 Zillow imposed “Listing Access Standards” barring homes publicly marketed off-Zillow for more than a day (the “Zillow ban”), a direct shot at Compass’s three-phase “private exclusive” strategy. Compass sued for antitrust in June 2025; in 2025 Zillow banned 48 listings, ~90% of them Compass’s (FACT, web). The fight then reversed: Compass dropped its suit in March 2026 after Zillow relaxed the standards, and in May 2026 Zillow went on offense, suing Compass and a Chicago-area MLS over private-listing practices (FACT). Zillow’s “Preview” product — putting pre-market listings publicly on Zillow (and now Realtor.com) from day one — is the commercial expression of this stance. Interpretation: Zillow is defending the public-portal model that is its entire moat; it has largely won the Compass skirmish, but the episode exposes how much of the moat depends on continued access to broker/MLS inventory it does not own.

The Google catalyst — the reason the multiple broke. On June 11, 2026, Google expanded its Home-Listing Local Services Ads nationwide across all 50 states, surfacing price, photos, and features directly inside mobile Search and connecting buyers to listing agents for a paid lead, powered by HouseCanary data (FACT, HousingWire / Inman, 2026-06-11). Zillow shares fell 4.9% and CoStar 3.6% on the day (FACT). The threat is existential in theory: Google intercepts the highest-intent buyer at the search box, disintermediating the aggregation function that generates ~80% of Zillow’s direct traffic. The mitigants are real but unproven: only three MLSs currently feed the enhanced format, the program was paused in January 2026 over MLS-consent concerns and relaunched consent-first in mid-May, and Zillow stated on June 16 that pilot-market data showed no meaningful traffic decline and that Google poses “no immediate threat — yet” (FACT for the statement; the claim itself is management hypothesis, the relevant section rule 8). Interpretation: this is the first credible attack on the demand-aggregation moat by a party with more search distribution than Zillow will ever have; even if the near-term traffic impact is nil, it caps the terminal multiple, and the market has repriced accordingly.

The frozen market and the NAR overhang. The macro backdrop is the worst housing-transaction environment in three decades: 2025 existing-home sales totaled 4.06M, the lowest since 1995, versus a ~5.2M historical norm, as rate lock-in kept both inventory and turnover depressed (FACT, NAR/PBS/Wolf Street). A late-2025 dip in rates lifted December to a 4.35M SAAR, but management explicitly plans for a flat market all year (FACT, Q1’26 call). Separately, NAR’s $418M commission settlement (rules effective August 17, 2024) moved buyer-broker compensation off the MLS and mandated buyer-representation agreements — the feared structural compression of the commission pool that funds agent ad budgets. Notably, the compression has not yet shown up: buyer-agent commissions averaged ~2.43% in 2025, flat-to-slightly-up (FACT, Redfin data). Interpretation: the freeze is a demonstrable, ongoing revenue headwind that Zillow has out-run via share gains; the NAR risk is real but so far theoretical.

Litigation stack and the sell-side de-rate. Beyond Compass, three matters accreted in 2025: the FTC plus five state AGs (VA, AZ, CT, NY, WA) sued over the Zillow-Redfin rental deal (Sept/Oct 2025; bench trial set for August 24, 2026); CoStar sued for copyright infringement over ~46,000 watermarked images (July 2025, damages theories reaching ~$1B); and legacy Zillow-Offers securities/derivative suits remain live (FACT, 10-K Note 13). Legal costs are running ~200bps of margin headwind in H1’26 (FACT, Q1’26 call). Sell-side price targets have followed the tape down — Jefferies cut its Buy target to $60 on 2026-07-14, KBW stayed Market Perform (FACT, data brief). Interpretation: the de-rate is partly litigation-overhang, partly Google, partly macro — a classic multiple compression on unchanged (indeed improving) fundamentals.

Verdict — net weakening of the thesis on external factors, partially offset by genuinely improving execution. The controllable changes are positive: the platform pivot is working, For-Sale and Rentals are taking share through a dead market, SBC is falling, and buybacks are aggressive. But the exogenous stack has decisively worsened — Google introduces a structural ceiling on the aggregation moat that did not exist eighteen months ago, the FTC trial and CoStar suit are live tail risks, and the transaction freeze shows no durable break. On balance the last two years weaken the risk-adjusted thesis: the business is executing better than ever into a market that has become materially more dangerous, and the price move reflects the latter more than the former.

9. Risk Analysis

Zillow is a high-operating-leverage, ad-funded intermediary whose value rests on one fragile asset — being the default starting point for U.S. home search — layered over a violently cyclical transaction market. The risks below are ranked by the product of likelihood and impact, with the aggregation-disintermediation cluster (Google, listing access) as the dominant existential threat and the housing cycle as the dominant near-term earnings driver.

Most material risks. The two that dominate are non-symmetric. Risk #1 (search disintermediation) is the only one that can permanently impair the terminal value rather than a single year’s earnings: Zillow’s entire economic model is the conversion of free, habitual, top-of-funnel search traffic into paid agent, mortgage, and rental leads. Google — and, secondarily, LLM answer-engines and CoStar’s subsidized Homes.com — attack precisely that free-traffic aggregation. The near-term data (no pilot traffic loss, only three MLSs live, consent-first friction) argues the threat is slow-moving, and Zillow’s own AI-mode, rich-media (3D tours, Showcase), and transaction-integration are credible defenses because Google does not own the closing workflow. But the risk is directional and one-way: Google’s distribution advantage in Search is structurally larger than anything Zillow can build, and even a gradual traffic leak compresses the multiple the market will pay. Risk #2 (the frozen cycle) is near-certain to persist through 2026 on management’s own planning assumption, but it is a cyclical impairment — a coiled spring, not a broken one — and Zillow has demonstrably grown revenue double-digits through it by taking share. The interaction of #8 (SBC-funded earnings) with #11 (valuation) is the quieter danger: a market that stops crediting adjusted EBITDA and starts demanding SBC-inclusive free cash flow could re-rate the stock lower even with revenue compounding.

Catastrophic / total-loss scenarios. A permanent, total loss is remote. Zillow is effectively debt-free with ~$0.8–1.3B of net cash ($1.3B cash + short-term investments, plus a $500M revolver), generates positive GAAP net income and adjusted FCF, and owns irreplaceable brand and habitual-usage assets — 70% of U.S. movers touch Zillow, and “Zillow” is searched more than “real estate” (management claim). The realistic tail is severe impairment, not insolvency: a scenario in which Google/AI genuinely disintermediates search over 3–5 years while the housing freeze persists and the FTC forces an unwind of the Redfin rental deal would collapse both the growth and moat narratives simultaneously, and a 3x EV/S name with negative ROIC has ample room to compress toward tangible book (~$7/share) before finding valuation support. That is the −76% max-drawdown pathway the tape has already partially traveled (FACT, FactorsToday). The catastrophic case is therefore a multiple-and-narrative implosion, not a balance-sheet failure.

10. Valuation Discussion (Embedded Expectations)

Framework. GAAP earnings are not usable here — trailing net income of $23M puts the P/E at ~135x (the 10th percentile of its own history precisely because the “E” is a rounding error), so the multiples that matter are EV/Sales, EV/Gross Profit, and EV/adjusted-EBITDA, cross-checked against the stock’s own multi-year range and against the (poorer-quality) brokerage and (higher-quality) marketplace comp sets. All figures below are struck at the 2026-07-17 close of $33.905.

Capitalization (recomputed at the current price; note ROIC’s cached pr_last of $68.22 is stale — the stock has roughly halved). On ~248M total shares (Class A + B + C), market capitalization is ~$8.4B. The balance sheet is effectively debt-free: the convertible notes were fully settled by mid-2025, leaving only ~$364M of mortgage-warehouse borrowings (a self-liquidating pass-through matched by ~$386M of loans held for sale) plus ~$93M of leases, against $1.30B of cash and short-term investments — i.e. ~$0.84B of net cash (nearer ~$1.2B if the warehouse line is excluded as an operating pass-through). Taking the conservative ~$0.84B, enterprise value is ~$7.6B. Against FY2025 results that is:

Own-history context (valuation_index, ~10-year percentiles). Composite valuation sits at the 10th percentile; P/S of 3.11x is at the 14.6th percentile and P/B of 1.84x at the 5.5th. Ignore the GAAP P/E percentile (distorted by near-zero earnings) and treat P/B cautiously (book is inflated by goodwill). The clean read — P/S near the cheapest it has been in a decade — is the strongest single point for the bulls: on sales, you are buying the #1 US housing portal at a valuation it has only seen during the 2022 rate crash.

Comp context. Pure brokerages (Compass ~0.85x EV/sales) and iBuyers (Opendoor, valued on EV/gross-profit because “revenue” is a home-price pass-through) are structurally lower-multiple, lower-quality businesses; Zillow’s asset-light, high-gross-margin marketplace legitimately earns a premium to them. Versus higher-quality vertical marketplaces and classifieds (which trade high-single to low-double-digit EV/sales at maturity), Zillow’s 3.1x is undemanding if its network effect is durable — and expensive if Google and Homes.com are structurally diverting the funnel. The multiple, in other words, is the entire argument.

Embedded expectations (reverse read). At ~$8.0B EV / 3.1x sales / ~13x adjusted EBITDA, the market is not pricing the super-app dream, but it is pricing continuity: mid-single-to-low-double-digit revenue growth, a stable-to-rising adjusted-EBITDA margin, and — implicitly — that today’s SBC-funded “profit” converts over time into real cash earnings. What it is emphatically not underwriting is either (a) a permanent step-down in traffic monetization from Google/native-listings/commission compression, which would collapse the growth algorithm and the margin path, or (b) a re-acceleration to super-app economics (double-digit For-Sale share-of-wallet gains plus a housing-volume recovery), which would support a 4x+ sales multiple. The stock is priced for “the franchise holds and slowly earns its keep” — the bear disputes the first clause, the bull the pace of the second.

Scenario analysis (illustrative 2028 revenue and exit multiple; assumptions explicit, not a forecast):

The spread (~$24 to ~$70) is unusually wide because the outcome is binary on a single unquantifiable variable — whether the aggregation moat survives Google and commission compression — and because there is no dividend and no reliable post-SBC cash return to anchor a floor. No price target and no recommendation is offered; the scenarios frame the embedded expectations only.

11. Variant Perception

Consensus. The sell-side has converged on a reluctant, de-rating “quality-platform-in-a-storm” view: Zillow is the #1 residential portal executing well on an integrated super-app strategy, but the stock is un-ownable near-term because (a) Google just declared war on the aggregation moat, (b) housing is frozen with no visible catalyst, and © the litigation stack (FTC, CoStar) clouds the print. Price targets have marched down with the tape — Jefferies to $60 (Buy) on 2026-07-14, KBW Market Perform — and the shares have halved in a year (FACT). Consensus, in short, still likes the business and distrusts the moment.

The strongest bull case. Zillow is an oversold #1 platform whose fundamentals are improving while the multiple collapses. Revenue grew 18% YoY in Q1’26 against a flat market; For-Sale is taking share (+12% vs. NAR +2%); Rentals is compounding ~30–40% toward a $1B+ target with a long multifamily runway (76k of ~140–150k buildings penetrated); Zillow Home Loans is inflecting (+96% origination volume); and operating leverage is finally showing (net-income margin +500bps YoY, SBC guided down 15%+). Management is buying stock hand-over-fist — $626M in Q1’26 alone, shrinking the count from 256M to 240M shares, with ~$1.3B authorization left — signaling conviction at these prices. On valuation the stock is at ~3.1x EV/S, near the cheapest level in its own multi-year history (P/S 14.6th percentile), for an asset with 70% mover reach, ~80% direct traffic, and an emerging super-app (search → finance → close → CRM) that carries real optionality no rival can replicate end-to-end. If housing so much as normalizes toward 5M+ sales, the operating leverage on a fixed ~$1.1B cost base is dramatic.

The strongest bear case. Zillow is a structurally challenged toll-booth whose toll road is being paved around it. Google’s nationwide listings launch — with the distribution to intercept buyers before they ever reach Zillow — plus CoStar’s deep-pocketed Homes.com and LLM answer-engines, attack the one thing Zillow sells: aggregated demand. The commission pool that funds Premier Agent faces structural NAR-driven pressure, the content moat is rented from MLSs and brokers (as the Compass and CoStar fights expose), and the vaunted “profitability” is an SBC mirage — $390M of stock comp against $23M of GAAP net income and ~$101M of FCF, with buybacks quietly recycling dilution rather than returning surplus. Housing may stay frozen for years, and a 3x EV/S multiple on a business that has never earned its cost of capital (ROIC ~−0.5%) has plenty of downside toward tangible book. In this reading the share gains are Zillow winning a bigger slice of a shrinking, disintermediating pie.

The 3–5 assumptions that matter most.

1. Does free, habitual search traffic survive Google/AI? The entire moat. Bull assumes brand + integration + owned workflow defend ~80% direct traffic; bear assumes gradual leakage caps terminal value.
2. Does the commission/ad-budget pool hold? Bull assumes agent marketing spend is sticky (commissions flat at ~2.43%); bear assumes NAR + fewer transactions structurally shrink it.
3. Is Rentals durable at 30%+? Bull assumes multifamily wallet-share compounding to $1B+; bear assumes the FTC forces a Redfin-deal unwind and growth decelerates.
4. Does operating leverage convert to real (SBC-inclusive) free cash flow? Bull assumes the fixed-cost base + falling SBC inflect FCF; bear assumes SBC stays structurally high and GAAP profit stays thin.
5. Does housing thaw within the investment horizon? Bull needs only normalization; bear assumes multi-year freeze.

Falsification tests. Bull case breaks if Comscore/quarterly traffic shows a durable YoY decline in Zillow app/direct visits as Google MLS coverage expands, or if the FTC prevails at the August 2026 trial and Rentals growth snaps below ~20%. Bear case breaks if Zillow posts several quarters of double-digit For-Sale growth with SBC-inclusive FCF turning solidly positive and traffic share rising even as Google scales — i.e., the platform proves it monetizes intent that Google cannot capture.

Factor-positioning read — abandoned name, not a crowded short. The tape is unambiguously a falling knife: Z trades below its 200-EMA ($33.9 vs. ~$49.6), with an annualized 1-year return of −55.7% and a −76.6% five-year max drawdown; every horizon’s Sharpe is negative (FACT, FactorsToday/AZI, 2026-07-16/17). The factor signature is a rate-sensitive, high-beta (~1.35), negative-momentum (−0.40) small-ish growth name with an InterestRate loading of −0.65 — i.e., the stock is mechanically a leveraged bet on falling rates and a housing thaw, and it is being sold precisely because that regime has not arrived. Crucially, the negative-momentum and negative-alpha loadings, combined with a still-modest short interest and a business that is compounding revenue, argue this is an abandoned/de-rated growth name rather than a crowded, high-conviction short. That distinction is the variant-perception crux: consensus is offsides on timing/positioning, not on the business quality — the market has extrapolated the Google threat and the freeze into the terminal value while the fundamentals keep beating. Whether that gap is opportunity or value-trap turns entirely on assumption #1 (traffic durability), which no factor model can resolve — it is a fundamental judgment, and the tape simply confirms that the market has already voted “impaired” and dares the fundamentals to prove otherwise. Interpretation, not a call: the positioning says pessimism is priced; it does not say the pessimism is wrong.

12. Fact vs. Interpretation

13. Open Questions

1. How much engaged-shopper traffic does Google’s native-listings integration actually divert? The distinction that matters is casual browsers (low monetization) versus high-intent, bottom-of-funnel shoppers (the ones Premier Agent monetizes). No public data yet isolates this; it is the single most important unknown.
2. What is Zillow’s real, post-SBC, through-cycle earnings power? Management guides to adjusted EBITDA; the market needs a credible path from $232M of SBC-inclusive EBITDA to durable cash earnings. When does SBC fall toward ~10% of revenue?
3. How large and how durable is the buyer-agent commission pool post-NAR-settlement? Premier Agent economics ultimately rest on agents’ willingness and ability to pay for leads out of a commission that may be structurally compressing.
4. Can Rentals sustain ~30–40% growth, and what is its standalone margin? Rentals is the growth engine and (interpretation) the best business in the portfolio, but its segment profitability is not separately disclosed.
5. Does Zillow Home Loans reach a purchase-mortgage attach rate that makes the “super app” economically real, or does it remain a sub-scale, capital-intensive drag competing with Rocket/Redfin’s vertically integrated stack?
6. Where does the listing-access fight (Compass litigation, private exclusives, CoStar) settle — does Zillow retain comprehensive listing coverage, the foundation of its consumer value proposition?
7. What total share count and control structure persists given founder super-voting Class B and continued SBC issuance?

14. What Must Be True

Bull case — what must be true (and its falsification test):

• The network effect survives Google and Homes.com: engaged-shopper traffic and Premier Agent revenue-per-visit hold. Falsifier: two-plus consecutive quarters of declining For-Sale revenue or disclosed traffic/engagement erosion attributable to search disintermediation.
• Rentals keeps compounding at 20%+ and carries attractive marginal economics, becoming a large, less-cyclical profit pool. Falsifier: Rentals growth decelerating below ~15% with no margin evidence.
• Operating leverage converts into real (post-SBC) cash earnings — SBC declines toward ~10% of revenue while adjusted-EBITDA margin holds ≥24%. Falsifier: SBC stays ≥14% of revenue and buybacks continue merely to offset dilution.
• The housing market eventually normalizes off its ~30-year-low transaction trough, re-levering For-Sale. Falsifier: existing-home sales remain frozen through 2027 with For-Sale revenue flat-to-down.

Bear case — what must be true (and its falsification test):

• Google/native-listings and CoStar structurally divert top-of-funnel traffic, breaking the growth algorithm. Falsifier: Zillow’s app/direct-traffic share and For-Sale revenue prove resilient a year into Google’s rollout.
• The commission pool that funds Premier Agent compresses durably post-NAR. Falsifier: Premier Agent revenue per transaction stable or rising through 2026–27.
• “Profitability” never becomes real: SBC-funded adjusted EBITDA never converts to sustained post-SBC free cash flow. Falsifier: two-plus years of positive GAAP net income and post-SBC FCF with a flat-to-falling share count.
• Capital allocation stays value-destructive (buybacks at cyclical highs, dilutive M&A). Falsifier: disciplined, counter-cyclical repurchases and no large goodwill-heavy deals.

The two cases are falsifiable on the same near-term evidence — the trajectory of For-Sale revenue and traffic engagement through Google’s rollout, and whether SBC-inclusive earnings turn durably cash-positive. That is the evidence to watch.

APPENDIX A — Standard Diligence Questionnaire

Zillow Group, Inc. (NASDAQ: Z) — as of 2026-07-18

Supplemental to the memo. Answers grounded in primary sources; Fact/Interpretation/Assumption labeled where it matters.

General — what thoughtful questions have other investors asked? The perennial debate is whether Zillow’s audience lead is a durable network-effect moat or merely a traffic-aggregation position rentable from Google — a question the June-2026 Google native-listings launch made acute. Investors also press on: the quality of “adjusted EBITDA” given ~$390M of annual SBC; whether the buyer-agent commission pool that funds Premier Agent shrinks post-NAR-settlement; the economic reality of the “housing super app” (does Zillow Home Loans ever attach at scale?); Rentals’ true margin and durability; and whether management’s capital allocation has learned anything from the ~$1B+ iBuying loss.

Cyclicality & Earnings Nature

• Cyclical high or low? A cyclical low on the demand side — US existing-home sales sit near a ~30-year trough (~4M annualized) on mortgage-rate lock-in (Fact, industry data). For-Sale revenue (Residential + Mortgages, ~74% of total) is directly transaction-levered and therefore depressed. Rentals (24%, +39%) is far less cyclical and is growing through the freeze.
• External or internal drivers? Both: the revenue level is suppressed by an external housing freeze, but the margin recovery (op margin −12.7% → −1.2%, 2023→2025) is internal (cost discipline, mix shift). Interpretation.
• Revenue stability / outlook? Advertising-and-services revenue is more stable than a brokerage’s transaction commission but still moves with housing activity. Outlook hinges on Rentals compounding and any housing-volume normalization.
• Market size — growing/shrinking? The online real-estate advertising and rentals-marketplace pools are secularly growing; the for-sale-commission pool that ultimately funds Premier Agent may be structurally compressing post-NAR. Domestic (US-only).

Business Quality & Competitive Moat

• Industry more/less competitive? More — Google inserting native listings, CoStar’s Homes.com free-lead assault, Rocket/Redfin vertical integration, and listing-access disputes have all intensified competition for the consumer funnel (Fact/Interpretation).
• Profitability (ROIC/ROE)? Poor — ROIC ~0-to-negative across five years; the business has never earned its cost of capital and carries a −$1.86B accumulated deficit (Fact, ROIC/10-K).
• Industry profitability / barriers? The #1 portal position is a real barrier (audience scale, brand, two-sided network), but barriers to entry by an adjacent giant (Google) are lower than they look. Brokerage/iBuying adjacencies are low-barrier, low-margin.
• Easily understood? Yes — an advertising marketplace plus adjacent services.
• Undermined by foreign low-cost labor? Not materially; some offshore engineering, but the asset is US consumer traffic and data.
• Do brands matter? Yes — “Zillow” is the category-defining consumer brand; Trulia, StreetEasy, HotPads add coverage. Brand is a genuine asset.
• Nature of competition / switching costs? Consumers multi-home across portals at zero cost (low consumer switching cost — the vulnerability Google exploits). Agent switching costs are higher where CRM/ShowingTime/Follow Up Boss are embedded in workflow (moderate).

Financial Condition & Balance Sheet

• Unrecognized assets? The consumer audience, data, and brand are internally generated and largely off-balance-sheet (Interpretation) — the real value versus $2.82B of acquisition goodwill.
• Off-balance-sheet liabilities? Convertible-note structures and capped calls; operating leases (modest). Warehouse lines fund mortgage originations.
• Accounting conservatism? GAAP is conservative relative to the adjusted-EBITDA narrative; the aggressive framing is the non-GAAP add-back of $390M SBC.
• CapEx-hungry? Asset-light on hard capex (~$133M PP&E) but capitalizes ~$134M of software; the real “investment” is expensed R&D/S&M and SBC.

Capital Allocation & Management

• FCF generation & use? Reported FCF ~$101M (FY2025) is entirely SBC-funded; cash has gone to $670M of buybacks in 2025 at ~$70 (now ~$34) — pro-cyclical and value-destructive (Fact/Interpretation) — and to M&A (Follow Up Boss ~$400M, ShowingTime, others).
• Significant acquisitions? Yes — ShowingTime, Follow Up Boss, Aryeo/Virtual Staging, Spruce; earlier, the ~$1B+ Zillow Offers iBuying build-and-unwind, a large unforced capital error.
• Buying back shares? Yes, but chiefly to offset SBC dilution rather than shrink the count materially.
• Issuing stock to insiders? Heavily — SBC ~$390–451M/yr; founder super-voting Class B concentrates control (Barton/Frink).
• Comp & motivations? Founder-led (Rich Barton exec chair; Jeremy Wacksman CEO from 2024). Incentive detail per the DEF 14A; alignment is high on ownership, mixed on the SBC-funded model.

Valuation & Market Data

• ADR/MLP/K-1? No — US C-corp common/capital stock (Class A = ZG, Class C = Z). No K-1.
• Dividend? None; returns via buybacks only.
• Profitability? Marginal GAAP; ~24% adjusted-EBITDA margin, but ~$232M net of SBC.
• Net income vs CFO divergence? CFO ($368M) >> net income ($23M) — normal for an SBC-heavy, D&A-heavy model; the divergence is the SBC/D&A add-back, not accrual quality.

Risks & Downside

• What would cause the stock to decline further? Evidence of Google/native-listings diverting engaged-shopper traffic; For-Sale revenue rolling over; commission-pool compression; a listing-access loss; SBC staying elevated; a housing freeze extending into 2027.
• Catastrophic-loss risk? Low near-term — net cash, no going-concern issue. The catastrophe is slow: permanent moat impairment de-rating the multiple over years.
• Total-loss risk? Very low — solvent, cash-generative, dominant. This is a valuation/quality risk, not a solvency one.

Recent News & Events

• Environment changed? Yes, materially: the June-2026 Google native-listings launch, the ongoing listing-access/private-exclusives fight (Compass litigation), post-NAR commission dynamics, and a still-frozen housing market. Q1-2026 revenue beat on For-Sale strength provided a partial offset. See the Changes and Headwinds section.
• Accounting-policy change? None material identified.
• Recent expansion? Rentals build-out (multifamily + syndication with Redfin and Realtor.com); Enhanced Markets integrated-experience expansion; Zillow Home Loans purchase-mortgage push.

APPENDIX B — Source Appendix

Zillow Group, Inc. (NASDAQ: Z) — as of 2026-07-18

Primary sources prioritized. Every material claim in the memo traces to an entry here or to the underlying primary source. Third-party aggregator data (ROIC.ai, AZI, FactorsToday) is labeled and reconciled to filings where it drives a verdict.

Primary — SEC filings (EDGAR, CIK 0001617640)

• Zillow Group FY2025 Form 10-K, filed 2026-02-11 (z-20251231.htm) — business description, four revenue categories, TAM, competitive advantages, MD&A (revenue/cost detail, Adjusted EBITDA & Adjusted FCF reconciliations, Key Metrics: TTV, revenue-per-TTV, unique users), Legal Proceedings (Zillow Offers securities/derivative suits; NAR; CoStar; FTC), risk factors, balance sheet, cash flow, shareholders’ equity, share-repurchase table. Primary source of record.
• FY2021–FY2024 Form 10-Ks (filed 2022-02-10 through 2025-02-11) — five-year revenue/margin history; Zillow Offers wind-down disclosures (FY2021).
• Form 10-Qs (trailing 15) — quarterly category revenue, Q1 2026 print.
• 8-Ks — CEO transition (Wacksman CEO, Barton Exec Chair; 2024-08-07); convertible-note settlements (2024-09-04, 2024-12-19, 2025-05-16); capped-call unwind (2025-08-25); $500M revolver (2026-01-30); $1.25B buyback authorization (2026-03-05); Gurley board appointment (2024-01-09); annual-meeting results.
• DEF 14A proxy statements — executive compensation (salary + time-vested RSUs/options, no cash bonus, no PSUs), CEO/founder pay, dual-class Class B super-voting structure, security ownership (Barton 36.2%, Frink 23.5%, insiders 60.6% of votes).
• Form 4 / 4-A corpus (375 + 13 over five years) — insider transaction read: zero code-P open-market purchases; 100% net selling via option-exercise-and-sell and routine 10b5-1 plans (Barton, Frink, Wacksman, Choo, Rock, Owens, directors).

Primary — company disclosures

• Zillow Q1 2026 earnings release & conference call (2026-05-06/07) — For-Sale +12%, Rentals +42%, Mortgages/ZHL +56% (origination +96% to $1.5B), enhanced-markets 49% of connections, $626M Q1 buyback, SBC guided down, “flat market” 2026 planning assumption, management’s Google pilot-traffic commentary.
• Zillow Investor materials / “Housing super app,” Enhanced Markets, Zillow Preferred (Flex rebrand) disclosures.
• February 6, 2025 Zillow–Redfin rentals partnership (~$100M); May 2026 Realtor.com “Preview” collaboration.

Industry / regulatory

• National Association of Realtors — existing-home sales (2025: 4.06M, lowest since 1995), commission data; NAR $418M settlement (rules effective 2024-08-17).
• Census Bureau (new-home sales), MBA / Fannie Mae (mortgage origination volume), ALTA (title), Fortune Business Insights (property-management software) — TAM components.
• Redfin commission-tracking data (buyer-agent commission ~2.43% in 2025, flat-to-up).
• DOJ ongoing investigation of NAR (risk-factor disclosure).

Competitive / event (trade press — validate against primary)

• Google nationwide Home-Listing Local Services Ads launch (HouseCanary-powered), 2026-06-11 — HousingWire, Inman, The Real Deal, TechRepublic, Benzinga (2026-06-11/16); Zillow −4.9%, CoStar −3.6% on the day.
• Compass v. Zillow antitrust suit (June 2025); PI denied (Feb 2026); Compass dropped suit (Mar 2026); Zillow countersuit vs Compass + MRED (May 2026) — Courthouse News, Inman, Real Estate News.
• FTC + 5 state AGs (VA, AZ, CT, NY, WA) suit over Zillow–Redfin rental deal; bench trial set 2026-08-24.
• CoStar copyright infringement suit (~46,000 images), filed 2025-07-30.
• CoStar/Homes.com Q1-2025 residential revenue (~$16.5M) and traffic (~104M uniques) — Inman/RISMedia.
• Rocket Companies acquisitions of Redfin and Mr. Cooper (2025).
• Sell-side: Jefferies (Buy, PT cut to $60, 2026-07-14), KBW (Market Perform) — via news feed.

Quantitative aggregators (third-party; reconciled to filings)

• ROIC.ai — five-year income statement, balance sheet, cash flow, profitability ratios (ROIC/ROE negative), enterprise value, valuation multiples, latest stock price ($33.21, 2026-07-17). Cross-check; EDGAR/10-K primary.
• Market-data feed — 5-year daily price/OHLCV CSV (EMAs, beta); valuation_index own-history percentiles (composite 10th, P/S 3.11x @ 14.6th, P/B 1.84x @ 5.5th); news feed. Note: ROIC’s cached pr_last $68.22 is stale; current price $33.905 per CSV / ROIC latest.
• FactorsToday — factor loadings (Market +1.31, Momentum −0.40, InterestRate −0.65, SmallSize +0.88, Quality −0.11); leaderboard (y1 −55.7%, m6 −74.5% annualized, 5yr max drawdown −76.6%, negative Sharpe all horizons); stock-info (beta ~1.35, negative alpha). Statistical estimates; positioning overlay, not a call.

Related prior analysis (same sector)

• Compass (COMP) — brokerage economics, NAR commission dynamics, SBC-funded profitability framing.
• Opendoor (OPEN) — iBuying economics, SBC/quality-of-earnings lens, housing-cycle framing.