3 Numbers From Compass Q2 That Should Change Your Brokerage Math

Compass just posted a record quarter. Revenue hit $4.31 billion, up 109% year over year. The press release used "record" eleven times. Wall Street liked it. The stock moved.

But three numbers buried in the filing tell a different story if you are an agent deciding where to hang your license this fall. One of them should make you rethink everything you assumed about mega-brokerage growth.

The Three Numbers and What They Mean for Working Agents

Compass reported $4.31 billion in Q2 2026 revenue, $92 million in net income, and $363 million in adjusted EBITDA. Those are the numbers in the press release. The numbers that matter for your brokerage decision are different: 1,003 agents lost, roughly 100 agents gained organically at eXp, and a 95.5% retention rate that sounds better than it is. Each one tells you something about where the industry is headed and what it means for your split, your support, and your tech stack.

1,003 Brokerage Agents Gone in a Record Quarter

Compass posted record revenue and lost agents in the same quarter. The headline and the trend line are telling two different stories.

Compass ended Q2 2026 with 83,184 brokerage agents, down from 84,187 at the end of Q1. That is a net loss of 1,003 agents in the same quarter the company posted record revenue. Compass added 2,816 agents on a gross basis during the quarter, which means roughly 3,819 agents left.

The 95.5% quarterly retention rate improved from 94.1% in Q1, and Compass highlighted that retention among agents earning $20,000 or more in annual gross commission income held at 98.7%. The message: they are keeping the top producers and losing lower-volume agents. That is strategically intentional. But if you are a newer agent or one producing under $20K GCI annually, the signal is clear. Compass is optimizing for revenue per agent, not for headcount. The $4.31 billion revenue figure divided across fewer agents means higher per-agent production, which looks great on an earnings call and means less support for agents who are still building.

For context, Compass's total revenue includes franchise segment revenue from the Anywhere brands it acquired (Century 21, Coldwell Banker, Sotheby's International Realty, and others), which brings its total network to roughly 340,000 agents worldwide. The brokerage segment, where Compass directly employs agents, is where the 1,003-agent decline occurred. The franchise network operates under different economics entirely.

About 100: eXp's Real Organic Growth Number

eXp added 5,006 agents in Q2. Only about 100 came from agents choosing to join. The rest came from buying NextHome.

eXp Realty, operating under parent company AGNT Inc. since its rebrand, reported 87,338 agents at the end of Q2 2026. That is up 6% year over year and up from 82,332 at the end of Q1. Those numbers look like strong growth until you read the filing detail. Most of the gain came from the NextHome acquisition. eXp's organic agent additions during the quarter were approximately 100.

One hundred agents. For a company that built its brand on explosive agent recruitment through revenue-sharing incentives and stock awards, that number is a sharp departure from the growth-at-all-costs era of 2020 to 2023, when eXp was adding thousands of agents per quarter. Revenue hit $1.4 billion (up 11% year over year) and transactions grew 12% to 132,497, but the company still posted a net loss of $2.7 million. Adjusted EBITDA more than doubled to $25.7 million, which shows operational improvement, but profitability at the net-income level remains out of reach despite nearly 90,000 agents.

The productivity story is more encouraging. Agent productivity rose 6% year over year, and real estate sales volume grew 15% to $60.5 billion. eXp is getting more deals per agent even as agent growth flattens. But the stock-incentive recruitment model that made eXp famous is clearly slowing. If you are considering eXp for the revenue share or stock awards, model your expected income from those programs using 2026 recruitment rates, not 2021 rates.

95.5% Retention Sounds Good Until You Annualize It

A 95.5% quarterly retention rate compounds to roughly 17% annual churn. At 83,184 agents, that's 14,000 replacements needed per year just to hold steady.

Compass's 95.5% quarterly retention rate means 4.5% of brokerage agents leave every quarter. Compounded over four quarters, that works out to roughly 17% annual attrition. At 83,184 agents, that pace means Compass needs to recruit approximately 14,000 agents per year just to hold steady, never mind grow. The company added 2,816 on a gross basis in Q2, which annualizes to roughly 11,264. That is below the replacement rate.

eXp's retention numbers are harder to parse because the NextHome acquisition muddies the organic figures. But the theme is the same: both mega-brokerages are running on an acquisition treadmill where M&A replaces organic recruitment as the primary growth engine. That is a fundamentally different business model than what either company pitched to agents five years ago.

What This Means for Your Brokerage Decision

If you are evaluating brokerages this fall, these earnings tell you three things worth weighting in your decision:

Mega-brokerage growth is now M&A-driven, not agent-attraction-driven. Compass doubled revenue by buying Anywhere. eXp grew agent count by buying NextHome. Neither company is growing meaningfully through agents choosing to join. This matters because the value proposition pitched to you during recruiting (network effects, brand momentum, growing agent base) is being sustained by acquisitions, not by agents voting with their licenses. Ask your recruiter what percentage of their office's agent growth in the last 12 months came from acquisitions versus organic recruitment.

Per-agent economics differ by a factor of 3x. Compass's $4.31 billion in quarterly revenue across 83,184 brokerage agents is roughly $51,800 per agent per quarter. eXp's $1.4 billion across 87,338 agents is roughly $16,000 per agent per quarter. This gap reflects different business models (Compass takes a larger split; eXp provides a lower split plus stock incentives), different agent production levels, and different market segments (Compass skews luxury). But the ratio tells you something about the support infrastructure each dollar of agent production is funding.

Your tech stack is baked into the brokerage choice whether you realize it or not. Compass agents use its proprietary platform, which includes CRM, marketing, and transaction management tools built on the technology it acquired. eXp agents get access to eXp's internal tools plus kvCORE for CRM. In both cases, the brokerage's technology decisions shape your workflow. If you are comparing CRMs independently, factor in whether your brokerage choice will override that decision. Compass agents evaluating BoldTrail/kvCORE pricing should know that cost may already be included or required depending on your franchise affiliation.

The Solo Agent Calculus Is Different From the Team Lead Calculus

Solo agents producing 8 to 15 deals per year care most about split percentage, monthly fees, and the quality of leads or referrals the brokerage provides. For them, the mega-brokerage earnings are background noise. What matters is the net-to-agent number on every closing statement. A 70/30 split at Compass on a $400,000 median sale with a 2.5% commission yields $7,000 per deal to the agent before fees. An 80/20 split at eXp on the same deal yields $8,000, minus monthly fees and before any stock-award value. The $1,000-per-deal gap over 12 deals is $12,000 per year. That is the number worth arguing about, not the headline revenue.

Team leads producing 40 to 80 deals care about something else entirely: whether the brokerage's infrastructure can support their operation without requiring a parallel tech stack. A team lead paying $1,200 per month for a CRM the brokerage also provides is wasting $14,400 per year. But a team lead stuck on a brokerage-mandated CRM that cannot handle their lead routing needs is losing deals. The Q2 earnings do not answer this question directly, but they tell you where each company is investing. Compass invested in integrated services (title and escrow revenue per transaction hit $3,654 in Q2). eXp invested in operational efficiency (operating expenses grew only 2% despite 11% revenue growth). Both choices shape the agent experience differently.

If your brokerage decision hinges on lead generation costs, remember that neither mega-brokerage's earnings report tells you what leads cost at your specific office. That number comes from your local market, your portal spend, and your conversion rate, not from a corporate earnings call. If you're building a lead gen and CRM stack that doesn't depend on brokerage-mandated tools, start with RobinFlow's onboarding to see what an independent setup looks like.

FAQ

Did Compass actually grow, or did they just buy Anywhere?

Both. Compass's pro forma revenue (which adjusts for the Anywhere acquisition as if it had been owned for the full comparison period) grew 14.3% year over year. That is real organic growth. Gross transaction value outperformed the overall market by roughly 1,000 basis points on a pro forma basis. But the headline "109% revenue increase" is almost entirely the Anywhere acquisition. The organic growth is solid; the headline is M&A math.

Is eXp still a good option for agents who want revenue share?

The revenue-share model still works, but the math has changed. When eXp was adding thousands of agents per quarter, building a downline was relatively easy. With organic growth at roughly 100 agents per quarter companywide, the recruitment pool is much smaller. Model your expected revenue-share income using current growth rates. If your projection depends on recruiting 10 agents per year in your market, check whether eXp added that many organically in your state during Q2 2026. The stock-incentive value also depends on AGNT's share price, which reflects the company's path to profitability. The adjusted EBITDA doubling to $25.7 million is positive, but net income is still negative at minus $2.7 million.

Should I care about my brokerage's Q2 earnings?

You should care about what the earnings reveal about where your brokerage is investing and whether its growth model is sustainable. If your brokerage is growing only through acquisitions and losing agents organically, the culture and support infrastructure may shift in ways that affect you directly. If your brokerage is profitable and investing in technology and services you use, the earnings are confirming that your fees are funding something useful. Either way, your per-deal economics matter more than the company's headline revenue. Run your own numbers first, then check whether the earnings story aligns with your experience at the desk level.