Compass: Real Estate Revolution - [Business Breakdowns, EP.226]

Summary

• Geoff Collette’s core thesis: Compass has transitioned from a VC-funded growth story toward a potentially low-cost, scaled operator of US residential brokerage — “Today Compass is not the Compass of a few years ago.” It’s the largest US brokerage by sales volume with ~6% of gross transaction value, ~40,000 agents across 38 states plus DC, expected ~$7B revenue and ~$200M free cash flow in 2025, and the stock is up 4x from its lows.
• The moat is four reinforcing advantages: a proprietary end-to-end software platform (~$2B invested since inception), national scale, a leading top-agent network, and exclusive listing inventory depth. Compass has 18% of the top 10,000 agents — 50% more than the next competitor — in an industry where the top 20 agents do roughly 85% of transactions, and 18% of Compass agents’ business comes from in-network referrals. Collette argues the capital window that funded the platform “likely won’t present itself again for a competitor.”
• The 2022 housing downturn was the proving event: mortgage rates spiked from under 3% to over 7%, home sales fell from ~6M to ~4M (a 30-year low), and Compass cut opex from a $1.5B run rate to ~$850M while retention held ~90% for principal agents even as multi-year incentives rolled off. Collette’s open question flips the bear case: “Was the downturn for housing the best thing that could have happened for them?”
• The operating-leverage math is the tradeable part: ~$1.2B gross profit at a high-teens margin against ~$1B opex committed to only 3–4% annual growth, so incremental free cash flow margins “can be in the double digits” — modest growth compounds into rapid FCF growth. SBC is now effectively a fixed cost, and M&A at 4–6x EBITDA falling to 2–3x post-synergy (the @properties–Christie’s deal was 9x toward 5–6x) is central to hitting the “30 by 30” target of 30% average share in the top 30 markets by end-2026, from just over 20% today.
• The three-phase marketing strategy is delivering measurable seller outcomes — listings using it sold ~3% higher, offers were accepted 20% faster (about 8 days quicker to close), and price drops occurred on only 13% versus ~40% market-wide — and ~50% of Q1 Compass listings used it, with 94% still hitting MLS. That undercuts the “double-ending” critique: privately double-sided deals are “a low single-digit percentage” of transactions.
• The Zillow fight is “a fight over the top of funnel”: Zillow’s rule pressures listings publicly marketed for more than 24 hours without being placed in MLS, pressuring phase two of Compass’s process, and Compass is suing Zillow in an antitrust lawsuit. Collette’s asymmetry framing: if Compass wins or pre-marketing flexibility spreads, that likely benefits Compass; “if the status quo persists nothing really changes.”
• Cyclical setup adds a kicker: at ~4M existing home sales — a 30-year trough, while midcycle is about 33% higher — “there’s a lot more upside than downside,” while the disintermediation risk hasn’t materialized (close to 90% of buyers used an agent in 2023, up from 83% in 2010). An AI assistant in beta this summer, full release by year-end, “could really bend the curve on the value proposition.”

Deep dive

1. Four advantages that reinforce each other

• Collette’s opening frame: Compass is the only major brokerage with a fully integrated, proprietary software stack — “everything from CRM to marketing to transaction management” — built mobile-first for agents on the go, funded by nearly $2B of technology investment since inception at “a point in time… that likely won’t present itself again for a competitor.” Owning the platform and data gives a unique AI edge: not just content creation but automated “workflows, follow-up, and client engagement,” with a voice-activated assistant in beta this summer and full release by year-end.
• Scale and talent density compound it: ~40,000 agents in virtually every major metro amortize R&D across a growing base, and Compass holds about 18% of the top 10,000 agents — 50% more than the next competitor — in an industry where the top 20 agents do ~85% of transactions. Top producers “like to work with one another,” and 18% of Compass agents’ business comes from in-network referrals, boosting revenue per agent, retention, and recruiting.
• The stated ambition is to escape commoditization: “not compete on commission splits, but an offering that really enhances the agent’s ability to grow their business more profitably.”

2. From VC-fueled land grab to religion on costs

• The founding model was NYC rentals, quickly pivoted to home-sale brokerage, then scaled with “an immense amount of venture capital funding” — sign-on bonuses and multi-year contracts incumbents couldn’t match. Collette concedes the sustainability question mark “remained with the company for a long time.”
• The 2022 downturn was the forcing function: rates jumped from under 3% to over 7%, sales fell from ~6M to ~4M. Compass paused expansion, stopped M&A, ran three rounds of layoffs in a year, eliminated all agent incentives, and — with CFO Kalani Reelitz — cut opex from a $1.5B run rate to ~$850M, “nearly cut in half.”
• The proof point: despite the cuts, Compass did not stop investing in its platform; by late 2022, agents could conduct their entire workflow on it. As multi-year incentives rolled off, organic agent growth continued and principal-agent retention held around 90% — “the technology-oriented value prop was able to kind of stand on its own by early 2023.” They “really found religion around cost discipline in 2022.”
• Today’s snapshot: largest US brokerage by volume, ~6% GTV share, ~21,000 principal agents and nearly 40,000 total agents across about 38 states and DC, with ~$7B expected 2025 revenue and ~$200M FCF. It also has the January acquisition of Christie’s International Real Estate, a high-margin global franchise letting independents join without selling outright, title/escrow operations in most markets, and an early-stage mortgage JV. The “30 by 30” goal is 30% average share in the top 30 markets by end-2026, versus just over 20% now — against average number-two competitors in the single digits.

3. NAR, MLS, and an industry built to stay fragmented

• Collette’s structural map: NAR’s 1.5M members must join local, state, and national Realtor associations to access MLS — “a patchwork of over 500 local databases,” ~80% association-controlled — making the trade group “more of a regulator for the industry than many other industries.” That construct supported tens of thousands of independent brokerages: the top 22 control only ~25% of volume, and ~60% is done by firms under $1B in GTV.
• The 6% commission is “a little bit of a misnomer” — averages drifted from ~5.3% toward 5% through 2020 and ticked up in 2023. The listing side has remained around 3%, while the buy side has faced more pressure. The Burnett settlement ended the MLS offer-of-compensation in August 2024 and introduced written agreements between buyers and their agents, but rates and buyer representation stayed stable. The real shift is toward professionalism: buyers’ agents are “having to articulate their value to their client in a way that they really didn’t have to before.”
• The stat that carries the argument: 70% of agents who paid MLS dues and were brokerage-affiliated in 2024 did zero transactions — a market shifting share toward the most professional agents and the brokerages that support them.

4. Inventory is power: three-phase marketing and the Zillow war

• The premise: “ultimately, inventory is power in residential real estate.” There’s real demand for off-MLS marketing — Illinois’s MLS had, on average, ~21% of volume that eventually went to the full market pass through its private listing network, ~40% of listings are dropping price, and even Zillow’s own studies found 31% of sellers want pre-marketing. Compass’s three phases are private exclusive (validate pricing without days-on-market and price-drop metrics), “coming soon” on Compass.com, then full MLS/portal launch — modeled on homebuilders, “very sophisticated repeat sellers who aren’t really putting all their inventory on MLS.”
• The private-exclusive phase is accessible to Compass agents and their buyer-side clients as well as other brokerages; Compass says it is not withholding listings or trying to prevent co-brokering. The results as told: ~50% of Q1 Compass listings used the process, 94% still hit MLS, sold ~3% more on average, offers were accepted 20% faster (about 8 days quicker to close), and price drops occurred on only 13% versus 40% broadly. On the double-ending critique — “it’s not the one-phase marketing strategy”; privately double-sided deals are a low single-digit percentage, and “a lot of this is overblown to be quite honest.”
• The Zillow conflict traces to 2019’s Clear Cooperation rule, which required listings to enter MLS within 24 hours of public marketing — “even coined by some as the Compass rule.” After March 2025 changes created more pre-marketing flexibility, Zillow went further: if a property is publicly marketed and not put into MLS within 24 hours, “your listing will be banned from Zillow.” Collette calls it “a fight over the top of funnel”; Compass is suing in antitrust, enforcement “might be kind of tricky,” high-end privacy could bifurcate the market, and CoStar’s Homes.com is giving free advertising boosts to banned listings. Chicago has operated comparable private listing networks for over a decade, and Austin recently implemented something similar. Collette’s asymmetry is that greater flexibility likely benefits Compass; if the status quo persists, nothing really changes.

5. The model: 82/18 splits, capped opex, and double-digit incremental margins

• The revenue equation: agents × productivity × average price (~$1M — “premium” rather than luxury, per Collette’s correction of Matt’s framing) × commission rate × the split — with organic share gains in all 17 quarters as a public company. Splits average ~82% agent / 18% Compass, with a ~9-percentage-point band by production level; growth mix-shifts from the top 10% of agents toward the top 50%, “improving overall margins without actually changing commission splits.” Rival models with richer splits come with “less support, more out-of-pocket costs.”
• The leverage: 2025 FCF margin ~3% on ~$1.2B gross profit at a high-teens margin against ~$1B opex including ~$190M annual tech spend — with opex committed to 3–4% annual growth. Incremental FCF margins “can be in the double digits,” so “not too high levels of growth can result in very rapid free cash flow growth.” SBC is “basically a fixed cost now,” making FCF per share “very attractive.” Competitors, Collette notes, are “starved” of the scale to invest — stuck with “a hodgepodge of third-party software tools.”
• M&A “could be very material” to reaching 30% share: Tennessee and Louisiana brokerages, the @properties–Christie’s deal at 9x EBITDA heading toward 5–6x with synergies ahead of plan, a 4–6x target falling to 2–3x post-synergy while keeping opex flat, plus “walkover situations” where brokerages shed costs and join Compass.

6. Risks, and the questions that decide the story

• The foundational assumption: “the agent remains central to the home transaction.” Disruption attempts “largely have not panned out” — close to 90% of buyers used an agent in 2023, up from 83% in 2010, because home sales are “an infrequent high stakes situation where consumers overwhelmingly prefer expert guidance.” AI is the residual risk, but Compass’s own AI rollout “should actually be both offensive and defensive.” Other watch-items are retaining productive agents, continuing to provide technology, training, and support, M&A execution and integration, and pre-marketing flexibility reversing — “that would not help them differentiate.”
• Macro is framed as asymmetric: ~4M existing sales is a 30-year low, while midcycle is about 33% higher — “a lot more upside than downside,” and “a recession without inflation could actually help mortgage rates.”
• Collette’s closing lesson: residential brokerage “triggers an immediate hard pass” for many investors — low margins, churn, regulatory noise — “but sometimes downturns reshape companies in ways that are easy to miss.” His open questions: has the tipping point already been reached, what upside sits in the court cases, and long term, “could it evolve into a home search destination, not just an agent platform — and what would the economic situation look like then?” Matt’s admission closes the loop: “It’s a name I will admit I had a very specific view on… it’s opened up my mind to the reality of the situation.”