Will Compass maintain combined agent count above 323,000 (95% of ~340,000) through Q3 2026?

The Q4 2025 earnings data provides the first post-announcement retention data, and it is uniformly positive. Compass Q4 retention at 96.8% is consistent with the 97%+ historical baseline — no deterioration from merger uncertainty. Anywhere's top-quartile retention at 'all-time highs' in January suggests the franchise-side agents are not fleeing despite the acquisition. The 830 principal agent adds in Q4 is a record, indicating Compass's recruiting engine is operating at peak capacity. Combined, this data shifts the probability meaningfully upward from the prior 67%. The Rocket/Redfin partnership creates an additional retention tailwind by offering agents access to a major alternative distribution channel.

The positive retention data from Q4 and January is encouraging but does not fully de-risk the July 2026 platform rollout, which is the most dangerous single event for agent retention. Migrating owned-operations agents (Compass side) to a unified platform could create friction, confusion, and feature loss that drives departures. Historical precedent from tech platform migrations shows that forced transitions often trigger user churn. The July timing means the disruption would occur during the highest-activity period (Q3), when agents have the most to lose from platform instability. I weight the platform migration risk as the remaining key uncertainty, moderating my upward revision from the prior 67%.

The reporting change (total agent count only, no principal count) introduces a second-order consideration. Total agent count includes part-time, referral, and less active agents — a broader denominator that is likely more stable than principal-only count. If the market resolves against total agent count, the threshold may actually be easier to meet than anticipated, since total agent counts are less susceptible to competitive recruiting (part-time agents are less likely to switch). The Rocket/Redfin partnership is genuinely significant for retention: agents at Compass/Anywhere brands now have access to an alternative public listing channel, which addresses the Zillow ban concern that could otherwise drive departures. This is a unique competitive advantage no other brokerage can offer.

The earnings update strengthens the YES case on multiple dimensions. First, 96.8% Q4 retention at Compass demonstrates that the merger announcement (pre-close) did not trigger agent flight. Second, Anywhere's all-time high retention in January (post-close) shows the acquired franchise agents are stable during the most uncertain period. Third, 830 principal agent adds is a record, providing additional cushion above the 323,000 threshold. The prior assessment of 67% was appropriately cautious given no post-announcement data; now with favorable early data, an upward revision is warranted. The July platform rollout remains the primary risk, but is one event among many favorable signals.

Moderating optimism with structural concerns. The Q4/January data covers only 7-8 weeks post-close — a very short window. Integration disruptions typically compound over time as operational changes cascade through the organization. The platform migration in July is not just a one-day event; it requires months of preparation, training, and adaptation that will create ongoing friction from March onward. The 5% cushion (17,000 agents) is generous, but if a major franchise system (e.g., a large Century 21 franchisee) defects to a competitor, that could represent thousands of agents in a single event. The unprecedented scale means base-rate reasoning is inherently uncertain.

Balancing the positive early data against remaining risks. The most important new information is the Anywhere retention data — all-time high in January means the franchise-side agents, who are the majority of the combined count (~300K of 340K), are stable. This is the single most informative data point because franchise-side stability was the greatest prior uncertainty. If 300K franchise agents maintain 95%+ retention, Compass only needs to avoid catastrophic losses on its own 30-40K agent base to stay above 323,000. The platform rollout risk is real but contained to the owned-operations side, which is the minority of total agents.

Retention data is excellent across the board. Compass 96.8%, Anywhere at all-time highs, record agent adds. The 5% buffer is generous. The Rocket/Redfin partnership adds agent value. The platform rollout in July is a risk but the franchise-side majority is insulated. Moving from 67% to 78% reflects the strong early data.

Positive early data is encouraging but the July platform rollout is the critical event that has not yet occurred. Early retention being high is expected — the real test comes when operational changes reach agents directly. The all-time high Anywhere retention is the strongest signal since franchise agents are the majority. Still assigning meaningful probability to NO due to unprecedented scale and unknown dynamics of platform migration across 7 brands.

The weight of evidence has shifted clearly toward YES. Q4/January data is positive, record agent adds provide cushion, and the Rocket/Redfin partnership is a unique value proposition. The prior 67% was appropriately cautious with no data; now with favorable early signals, 75% reflects both the positive data and the remaining risk from the July platform rollout. The resolution date (Nov 30) gives time for any Q3 disruption to stabilize.