Data Report · 2026

The State of Real
Estate Commissions

What 52,256 transactions reveal about agent pay, brokerage economics, and commission structures across the US and Canada.

52,256
Closed transactions
204
Brokerages
6,729
Agents
~$30B
Transaction volume

Data period: January – December 2025 · Source: Loft47 back-office platform · Actual payout data, not survey responses

scroll ↓

Executive summary

Five findings that challenge industry assumptions.

This report synthesizes findings from 52,256 real estate transactions across 204 brokerages and 6,729 agents, representing approximately $30 billion in transaction volume during 2025. The data comes from Loft47's back-office platform — capturing actual commission payouts, deduction structures, and brokerage economics at a granularity no industry survey has ever achieved. This is not what agents say they earn. It's what they actually received.

01
Commission rates haven't moved.
The US median commission rate has been exactly 3.00% for 11 consecutive quarters — including every quarter since the NAR settlement took effect in August 2024. The 3% floor is structural, not fragile.
02
The median agent earns $31K.
The industry's widely cited "average agent income" of $71K+ is inflated 2–3× by top producers. The median agent closes 5–6 deals per year and earns $31,152 USD before business expenses.
03
Brokerages keep just 11–15 cents per commission dollar.
For every $1.00 in gross commission, approximately $0.82 flows to the agent. Brokerage margins are razor-thin — volume is the only path to profitability.
04
Teams close 3× more deals per person — but earn less per transaction.
The typical team is a duo closing 35 deals/year, 3× the per-person rate of solo agents. But team members take home 11–25% less per payout.
05
14.7% of Canadian deals collapse, costing $49M in lost commission.
One in seven Canadian deals fails after going conditional, with the average collapsed deal representing $11K–$14K in lost commission. The US collapse rate is significantly lower at 4.9%.

Part 1 — Commission rates

Commission rates aren't falling.

The US median has been exactly 3.00% for 11 consecutive quarters. In Canada, the median actually rose to 3.36% in Q4 2025 — the highest in the series. Every prediction about post-NAR compression has been wrong in this dataset.

US median rate
3.00%
Flat for 11 consecutive quarters
Canada median rate
3.36%
Q4 2025 high · oscillating 2.91%–3.32%
Post-NAR change
None
Zero compression detected in dataset
Volume-weighted US rate
3.02%
Same story from a different angle
MetricCanada (CAD)United States (USD)
Median commission rate3.36%3.00%
Rate trend (Q2 2023–Q4 2025)Oscillating 2.91%–3.32%Flat at 3.00%
Volume-weighted rate2.94%3.02%
Post-NAR settlement changeNone detectedNone detected
What this signals

The 3% floor reflects deep structural factors — brokerage economics, MLS norms, and agent expectations — rather than a fragile equilibrium that regulatory change can easily disrupt. Commission rate structures are driven by industry norms, not housing market cycles.


Part 2 — Dollar flow

Where every commission dollar actually goes.

For every $1.00 in gross commission, approximately $0.82 flows to the agent. Brokerages keep just 11–15 cents — roughly $1,750 per transaction. Volume is the only path to profitability.

Agent net pay
~82¢
Brokerage retained
~15¢
Third-party costs
~3¢
Avg brokerage retained per deal
~15¢
Per commission dollar — both markets converge near $1,750/transaction
Avg brokerage retained per transaction
~$1,750
CAD and USD converge — volume is the only path to profitability
Agent take-home — Canada vs US
87.2%
vs 81.5% in the US — flat-fee vs percentage splits
Avg brokerage revenue per agent
$24,400 CAD
vs $13,800 USD — mean figure (total retained ÷ agents), driven by higher CAD sale prices and rates
What this signals

Brokerage margins are razor-thin. Volume is the only path to profitability — explaining why recruiting intensity, agent count, and transaction throughput dominate brokerage strategy. The broker-owner who doesn't know their retained revenue per agent is making strategic decisions blind.


Part 3 — Agent income

The median agent earns $31K.

The industry's widely cited "average agent income" of $71K+ is inflated 2–3× by top producers. The median tells the real story — and it's a power law, not a bell curve.

US median annual income
$31,152
USD · actual payout data
Canada median annual income
$38,544
CAD · actual payout data
Most common payout size
$5–10K
29% of all payouts
Bottom 25% earn under
$11K
Per year — barely supplementary income
PercentileCanada (CAD)United States (USD)
Bottom 25%$11,809$10,507
Median (50th)$38,544$31,152
Top 25%$113,133$84,451
Top 10%$242,825$179,536
Average (misleading)$119,165$70,909
What this signals

The average is 2.3–3.1× the median — a telltale sign of a power-law distribution. The standard deviation exceeds the mean in both markets, confirming that agent income is dramatically skewed by top producers. Any recruiting strategy built on the "average" agent income is built on a fiction.


Part 4 — The rate curve

Higher price ≠ higher rate.

Commission rates decline as property values increase — but dollar payoffs scale dramatically in the opposite direction. The US market has crystallized around just two numbers: 3.0% under $500K and 2.5% above.

Price bandCAD median rateUSD median rateAvg commission $
Under $250K5.16%3.00%~$4,500
$250K–$500K3.00%3.00%~$9,500
$500K–$750K2.36%2.50%~$17,500
$750K–$1M2.59%2.50%~$27,500
$1M+2.83%2.50%~$75,000
Property sub-type premiumsCAD rateUSD rateNote
Land4.00–4.56%4.00–4.56%Highest rate in dataset — transaction complexity
Single-family / condo / multi-family2.83–3.40%2.83–3.00%CAD condo carries +57bps premium over SFH
Commercial retailUp to 6.00%USD highest rate segment
IndustrialModest rateHighest dollar payout at ~$32K per deal
What this signals

Despite lower rates at higher price points, a $2M+ deal generates ~$75,000 in median commission — 16× what a sub-$250K deal produces. The "chase the luxury listing" incentive is real and quantifiable. Without accurate commission tracking, brokerages can't see this distribution in their own data.


Part 5 — Team economics

Teams close 3× more deals — but earn less per transaction.

Teams are a volume play, not a margin play. The typical team is a partnership of two agents closing 35 deals per year. The per-payout trade-off is real.

Volume boost vs solo
More deals per person per year
US payout difference
−25%
Less per payout for US team members vs solo
Median team size
2
Partnerships, not mega-teams
$2M+ deals that are team deals
40%
Luxury real estate is increasingly team-driven
MetricSolo agentTeam leaderTeam member
USD median net/payout$5,585$4,494$3,916
CAD median net/payout$5,481$4,855 (blended)
Deals/person/year5–6~17
% of USD deals76.4%23.6%
% of CAD deals85.1%14.9%
What this signals

Teams average ~35 deals per year with ~$24.4M in volume — a figure remarkably consistent across borders. The volume gain is real; the per-payout trade-off is too. Commission structures for team environments are among the most complex to manage and the most error-prone when done manually.


Part 6 — Commission plan complexity

There's no such thing as 70/30.

Real commission plans are multi-layered systems mixing percentage splits, flat fees, third-party pass-throughs, and production incentives. The median brokerage maintains 7–9 distinct deduction templates.

Median templates/brokerage
7–9
CAD: 7 · US: 9 · Max seen: 55 (US)
Median commission cap
$112–130K
Annual agent commissions before cap fires
Deductions per deal
2–3
Despite template complexity — setup vs execution
Effectively capped plans
~13–15%
Half single-tier · half graduated multi-tier
MetricCanada (CAD)United States (USD)
Median templates/brokerage79
Max templates (single brokerage)3555
% percentage-based splits48%53%
% flat-fee-based52%47%
Effectively capped15.2%12.9%
Multi-tier graduated plans7.5%7.8%
Anniversary reset (vs calendar year)17%30.5%
What this signals

67% of templates tier against year-to-date agent commissions. Three cap philosophies: commission-based, deduction-based ($18K median), and deal-count-based (3–24 deals). The complexity is in the setup, not the per-deal execution — which is exactly where errors hide when plans are managed manually.


Part 7 — Geography

Rates are universal. Dollars aren't.

An Arkansas agent and a California agent both charge ~3% — but the California agent earns 4× more per deal. Commission dollars vary by market; commission rates don't.

Hawaii avg commission
$37,400
$1.33M avg sale · highest dollar market
California avg commission
$30,400
$1.02M avg sale · 2,886 deals
Arkansas avg commission
$7,500
Lowest dollar market in dataset
Lowest rate markets
OR / NV
~2.3–2.4% median rates
Canada (CAD)
Median sale price
$415,000
Median commission rate
3.36%
Avg brokerage revenue/agent (mean)
$24,400
Note
Alberta accounts for ~54% of Canadian data. Findings are not nationally representative.
United States (USD)
Median sale price
$315,000
Median commission rate
3.00%
Avg brokerage revenue/agent (mean)
$13,800
What this signals

Canada has 32% higher median prices and 36 basis points higher commission rates — resulting in significantly higher per-agent brokerage revenue. Despite this, both markets converge on the same fundamental commission structure and brokerage model.


Part 8 — List side vs sell side

The list/sell pay gap is a narrative, not a reality.

Despite structurally different commission formulas, buyer's agents and seller's agents take home virtually the same check. The sticker rate overstates actual effective cost by ~2.5 percentage points.

MetricList sideSell sideDifference
CAD median agent net$5,227$5,418+$191 (3.7%)
USD median agent net$5,075$5,187+$112 (2.2%)
Double-end deal frequency7–8% of all deals
Double-end premium per transaction13–16% above single-side deals
What this signals

Listing agreements have ~2× higher first-tier commission rates than sell-side agreements (CAD first-tier: 11.76% vs 6.10%). But the tiered formula structure means the sticker rate overstates actual effective cost by approximately 2.5 percentage points. After applying the formula, dollar amounts converge. The list-side advantage is largely structural math, not actual take-home difference.


Part 9 — Operational reality

47 days from handshake to payday.

Canadian deals take nearly twice as long as US deals to reach agent payout — and collapse at nearly 3× the rate. Despite this, 37–46% of deals are still entered manually into back-office systems.

Canada deal collapse rate
14.68%
1 in 7 deals fails after going conditional
US deal collapse rate
4.89%
1 in 20 deals fails
Lost commission from collapses
~$49M
$11K–$14K unrealized per failed deal
Deals still entered manually
37–46%
US leads digital at 62.5% via integrations
StageCanadaUnited States
Accepted offer → Closed42 days21 days
— Conditional period11 daysN/A
— Firm to closing32 days
Closed → Agent payout4.5 days2.6 days
Total: handshake → payday~47 days~24 days
What this signals

Agents source 86–95% of their own business. Company-sourced leads represent just 5.5–14.3% of deals — but are 23–30% more valuable per transaction. The automation opportunity is significant: manual entry still dominates nearly half of all deals. Every manual entry is a potential error in an agent's payout.


Part 10 — Cross-border comparison

Two markets, one industry.

Canada and the US moved in opposite directions on volume and price in 2025 — yet commission rates in both stayed anchored near 3%. The similarities are more striking than the differences.

Canada · CAD
Median sale price
$415,000
Deals in dataset
9,791
Median commission rate
3.36%
Agent take-home %
87.2%
Median net/payout
$5,350
Median annual income
$38,544
Deal failure rate
14.68%
Close → payout speed
4.5 days
Commission tax
~6% GST
Cap reset model
83% calendar year
YoY deal volume
−24%
United States · USD
Median sale price
$315,000
Deals in dataset
41,272
Median commission rate
3.00%
Agent take-home %
81.5%
Median net/payout
$5,130
Median annual income
$31,152
Deal failure rate
4.89%
Close → payout speed
2.6 days
Commission tax
~0%
Cap reset model
67% cal / 31% anniversary
YoY deal volume
+6.4%

Appendix — Home price trends

The macro backdrop.

Canada and the US are moving in opposite directions on price — yet commission rates in both remain remarkably stable. This decoupling reinforces the report's central finding: the 3% floor is structural, not a function of market conditions.

Metric20242025Change
CAD deal volume (Loft47)12,8899,791−24%
CAD average price$543K$519K−4.4%
USD deal volume (Loft47)38,78041,272+6.4%
USD average price$406K$416K+2.5%
Canada (CREA) · External
National home sales fell ~2% in 2025 (470,314 units)
MLS Home Price Index down ~5% YoY by Jan 2026
Nat'l avg price $652,941 in Jan 2026, −2.6% YoY
Prices declining in BC, Alberta, and Ontario
CREA forecasts 5.1% sales recovery + 2.8% price increase in 2026
United States (NAR) · External
Median existing-home price grew 1.2% YoY to $414,900 in Q4 2025
Prices rose in 73% of metro markets in Q4 2025
NAR forecasts ~14% sales volume increase + 4% price growth in 2026
Market characterised as transitioning from "challenging" to "improving"
What this signals

Despite Canada cooling sharply while the US grew, commission rates in both markets remained stable throughout 2025. This strongly suggests that commission rate structures are driven by industry norms and brokerage economics — not by housing market cycles.


Data quality & methodology

About this data.

All transaction data in this report comes from Loft47's back-office platform, which captures actual commission payouts, deduction calculations, and brokerage economics at the individual deal level. This is operational data — not survey responses or self-reported figures.

Closed deals
52,256
9,791 CAD + 41,272 USD + other
Brokerages
204
Across both markets
Unique agents
6,729
With at least one closed transaction
Transaction volume
~$30B
Total in dataset
Important caveats
· Canadian data is Alberta-heavy (~54% of CAD deals). Canadian findings should not be generalized nationally.
· Sub-type classification is ~31% populated. Sub-type breakdowns are flagged as "where classified."
· Commission rate averages are distorted by outliers. This report uses medians and volume-weighted rates throughout.
· Franchise fee data was not available at sufficient scale for analysis.
· A sell price ceiling of $25M was applied to exclude likely data entry errors.
· Commission advance data was deferred due to insufficient sample size.
External sources: Canadian Real Estate Association (crea.ca) · National Association of Realtors (nar.realtor) · Federal Reserve Economic Data (fred.stlouisfed.org)

Download the full report

Get the PDF — shareable, printable, citable.

The complete 2026 State of Real Estate Commissions report, formatted for sharing with your team, board, or at your next industry event.

No spam. One confirmation email with your download link.