Compass, Inc.

Q2 revenue was $4.3 billion, up 14.3% year over year on a pro forma basis, versus a guide of $4.0 billion to $4.2 billion. Adjusted EBITDA was $363 million, and GAAP net income was $92 million, up from $39 million a year ago. Brokerage revenue was $3.96 billion, franchise revenue was $135 million, and integrated services revenue was $211 million; brokerage GTV was $155.2 billion, up 16% year over year pro forma. Free cash flow was $180 million, cash on hand was $694 million, and net debt was $2.45 billion; net debt to adjusted EBITDA was 3.3x versus 4.2x in Q1. For Q3 2026, management guided to revenue of $3.85 billion to $4.05 billion and adjusted EBITDA of $275 million to $305 million. Full-year non-GAAP operating expenses are expected to be $2.75 billion to $2.80 billion, and management said Q3 free cash flow is expected to be positive.

Robert Reffkin framed the quarter as proof that Compass is gaining share and improving economics at the same time. He emphasized outperformance versus the market, calling the platform rollout, Coming Soon adoption, and Rocket-Redfin partnership evidence that the company is “infusing competition” into real estate and expanding seller choice. His tone was very upbeat and combative on industry structure, while also highlighting long-term earnings potential and operating leverage.

Scott Wahlers focused on the mechanics behind the quarter: revenue beat, adjusted EBITDA of $363 million, and free cash flow of $180 million. He said non-GAAP operating expenses were $699 million in Q2 and that full-year OpEx guidance was raised by $50 million to $2.75 billion to $2.80 billion, mainly due to $35 million from a recent brokerage acquisition and some additional compensation expense. He also cited $300 million of actioned cost synergies already reached, with as much as $330 million expected in 2026, and noted $1.8 billion of NOLs that should shield future taxable income.

Analysts pressed on how repeatable the Chicago Coming Soon/Redfin traffic gains are, and Reffkin said Chicago is not unique but rather a sign that the model works when sellers and agents have more flexibility; he said more markets are moving this way and that the trend is becoming the norm. Questions also focused on transactions per agent and agent growth; Wahlers said transactions per agent should keep rising as lower-productivity agents are traded out, while recruiting for the newly acquired brands should ramp more fully by early 2027. On synergies, management said the pace will slow as deeper systems integrations begin, but they remain confident they can meet or exceed the $500 million target over time.