Introduction

This methodology explains how Summit Commercial Solutions estimates Commercial General Liability (CGL) insurance costs for Canadian businesses we serve. It documents our data windows, normalization rules, assumptions, controls, and refresh cadence. Use alongside our product primer: Commercial General Liability (CGL) Insurance.

Scope and Use

Purpose: provide transparent, directional CGL price benchmarks for common industries and business sizes in Canada (for the markets we serve).
Not a quote: estimates are illustrative and subject to underwriting, eligibility, and market conditions. For a binding quote, speak with a broker.
Audience: business owners, risk managers, and researchers comparing expected CGL costs across industries and regions we serve.

Data Sources

De‑identified Summit brokerage data: quotes and bound policies from our operating systems (premium, limits, class/operations, revenue/payroll, location, loss history).
De‑identified market quotes returned by partner insurers via our placement workflows.
Internal market intelligence: carrier appetite updates and rate change communications.
Client-verified corrections captured during remarketing and renewals.
Privacy: all analysis uses de‑identified records in line with our Company Privacy Policy.

Data Window

Primary window: trailing 24 months from the refresh date to capture current market conditions while smoothing short‑term volatility.
Extended window: up to 36 months for sparse segments (e.g., niche classes or unusual limit structures). Extended-window observations are reweighted to avoid overweighting stale experience.

Normalization Defaults

We normalize observed premiums to a consistent basis before aggregating and publishing benchmarks.

Dimension	Default basis	Notes
Form	Occurrence CGL	Claims-made observations are excluded unless explicitly stated.
Limits	CAD 2,000,000 per occurrence; general aggregate equal to per-occurrence unless otherwise noted	If an observed policy uses different limits, we scale using limit factor curves derived from market quotes.
Deductible/Retention	No liability deductible	If a liability deductible or SIR exists, we adjust premium to a no-deductible equivalent using observed credit/debit factors.
Territory	Canada (markets we serve)	Cross-border exposures are excluded from normalization.
Revenue/Exposure	Current annualized revenue (or appropriate exposure base)	When only prior-year values are available, we annualize and index to the refresh date.
Taxes/Fees	Excluded from displayed estimates	Provincial taxes, policy fees, and broker service fees are excluded; carrier financing charges are excluded.

Adjustments and Modeling

Limit scaling: we fit class-specific limit factor curves from de-identified quotes across common limit sets (e.g., CAD 1M, 2M, 5M). Where class data is thin, we blend to a broader segment curve to avoid instability.
Exposure scaling: premiums are modeled against the relevant exposure base (typically revenue for service/trade classes). We use log-linear fits with regularization and caps to avoid implausible extrapolations.
Geography: regional relativities are estimated using postal-code clusters and metropolitan/rural tags. We suppress and/or blend small-sample relativities to protect privacy and prevent noise.
Claims history: surcharges or credits for losses are applied at the cohort level using anonymized loss counts and severities; large single-event outliers are winsorized (see Outliers & Quality Controls).
Market conditions: a market factor is applied to align historical observations to the refresh month, based on observed carrier rate changes in our placement data.

Assumptions

Submission completeness: analyses use records with sufficient underwriting detail (class, exposure base, limits). Incomplete records are excluded.
Coverage scope: estimates reflect standard CGL insuring agreements and commonly requested endorsements for the class. Specialized or manuscript endorsements are excluded from the baseline.
Packaged policies: when CGL is packaged with property/auto, we allocate premium to liability using carrier-provided breakouts where available; otherwise we impute using class-average liability shares.
Broker compensation: estimates reflect insurer premium before provincial taxes/fees. Our compensation structure is described here: How We Get Paid.

Outliers & Quality Controls

Eligibility checks: remove records failing basic validity (e.g., zero exposure, placeholder limits).
Outlier trimming: winsorize premium-per-exposure metrics at the 5th/95th percentiles within class/limit cohorts to reduce undue influence from atypical placements.
Small-sample safeguards: suppress any cell that does not meet minimum observation thresholds; blend to parent segments when needed.
Reconciliation: compare modeled medians against current in-market quotes from multiple carriers for sentinel profiles prior to publication.
Audit trail: each refresh is versioned with reproducible configuration (data cut, filters, coefficients) and peer-reviewed by a broker lead.

Refresh Cadence and Versioning

Scheduled refresh: quarterly (targeted for January, April, July, October).
Ad‑hoc updates: we may publish interim updates following material market changes (e.g., broad carrier rate actions) or methodology improvements.
Version labels: each published table references this page and displays the methodology version (YYYY.Q) and refresh date.
Last reviewed: October 23, 2025.

Segment Coverage

Industries: aligns with Summit’s public industry pages (e.g., construction and realty, professional services, technology, manufacturing, retail and wholesale, health and wellness). See Business Insurance for definitions.
Business size bands: micro, small, mid-market bands defined by revenue ranges; exact cutoffs are documented alongside each published table.
Exclusions: highly specialized risks (e.g., wrap-up liability, U.S. operations, hazardous products) and jurisdictions not served by Summit are outside scope.

How to Interpret Our Cost Tables

Point estimates: represent the modeled median premium for the cohort after normalization; actual quotes vary by carrier, risk controls, contracts, and loss history.
Ranges: interquartile ranges (IQRs) contextualize expected variability. Wide IQRs indicate heterogeneous risk within the cohort.
Apples-to-apples: when comparing your quote, align on form, limits, endorsements, exposure base, and geography; for coverage details, see our CGL guide.

Change Log (Abbreviated)

2025.10: clarified tax/fee exclusion; added explicit small-sample suppression rules; updated quarterly refresh targets.
2025.07: introduced class-specific limit factor curves; improved geography blending.

Contact and Corrections

To request a deeper class cut, report a data issue, or obtain a tailored proposal based on your operations, speak with our team via the contact options on our Business Insurance page.