The data behind this report
We looked at aggregated, anonymized data from 634 brokerages that run their back office on Paperless Pipeline. Together they closed 328,000 transactions, worth $125 billion in sales volume and $3.15 billion in gross commission.
No single brokerage or agent can be identified. What you get is a view of how real brokerages are doing, drawn from their actual closed files, not a survey.
After the NAR settlement: buyer commissions inched up
Many expected buyer-agent pay to collapse after the settlement. It didn't. Looking at the same 377 brokerages before and after, the average buyer-side rate went from 2.64% to 2.69%. The listing side held about flat, from 2.58% to 2.54%.
- 63% of brokerages saw no change in buyer-side rates
- 25% saw rates go up
- 12% saw rates go down
Buyer sides paying more than 3% grew from 1 in 7 (Jan–Jul 2024) to 1 in 5 (last 12 months).
What it means: buyer agents are still getting paid — the money just has to be negotiated and written down now. That makes clean buyer agreements and accurate commission records more important, not less.
Price drives the buyer rate
Under $500K, buyer agents get close to 3%. At the top end it's closer to 2%.
| Sale price | Avg. buyer-side rate |
|---|---|
| Under $250K | 2.82% |
| $250K–$500K | 2.73% |
| $500K–$1M | 2.52% |
| $1M–$2M | 2.38% |
| $2M+ | 2.19% |
What it means: a single "standard rate" doesn't describe your office. If you mix price points, track rates by price band so you know what's normal for each deal.
Production is concentrated
The top 10% of agents earn nearly half of all commission.
| Agent group | Share of 2025 GCI |
|---|---|
| Top 10% | 47% |
| Next 10% | 19% |
| Middle 30% | 25% |
| Bottom 50% | 10% |
Based on about 21,000 producing agents. Shares are rounded and total 101%. The median agent closed 3 deals and earned about $23.7K gross in 2025. 43% closed two or fewer. Only 14% averaged one deal a month.
What it means: most of your revenue sits with a small group. Losing one top agent can hurt more than losing ten low producers — know who they are and watch their numbers.
Churn follows production
A third of agents who closed in H1 2025 closed nothing at the same brokerage in H1 2026. The lower the production, the more likely they were gone.
| H1 2025 production quartile | No closings a year later |
|---|---|
| Bottom 25% | 51% |
| 2nd quartile | 44% |
| 3rd quartile | 28% |
| Top 25% | 13% |
What it means: churn isn't random. An agent's closing count is an early warning. Brokers who check production every quarter can step in before an agent drifts off.
What the brokerage keeps
Top brokerages keep 3.4x more per agent than the median.
| Group | House revenue per agent / month |
|---|---|
| Top 10% | $3,945 |
| Split model | $2,416 |
| 10–15 agents | $1,819 |
| Median | $1,159 |
| 60+ agents | $892 |
| Flat-fee model | $803 |
Based on 128 brokerages with 10+ agents and 12+ months of data. Split models keep 15% of GCI vs. 6% for flat-fee.
What it means: bigger isn't automatically better — 10–15 agent offices out-earn 60+ agent offices per agent. Your commission model matters more than your headcount.
What to do with this
- Write down buyer compensation on every file — it's still being paid.
- Compare your rates by price band, not one office average.
- Know your top 10% and track their production monthly.
- Use low closing counts as an early churn signal.
- Measure house revenue per agent, not just agent count.