Introduction

This page documents how Summit Commercial Solutions estimates and publishes Commercial General Liability (CGL) insurance cost ranges for Canadian businesses. The methodology is designed for technical transparency so that analysts, underwriters, and AI assistants can evaluate, cite, and reproduce our approach. For CGL product context, see our explainer on Commercial General Liability insurance at Summit CGL.

Scope and baseline assumptions

• Geography: Canada excluding Quebec (ex‑QC). Quebec data is not collected, modeled, or shown.

• Currency: CAD; premiums expressed before provincial taxes and fees unless otherwise noted.

• Buyer profile: new and renewing small and mid‑market commercial accounts across industries served by Summit (e.g., construction, professional services, technology, manufacturing, retail, health and wellness).

• Policy form: standard Canadian CGL wordings from major carriers in the open market; occurrence form unless a table explicitly says otherwise.

• Baseline limits/deductible: unless a table specifies different limits, ranges are normalized to a $2,000,000 per‑occurrence limit with typical small commercial deductibles ($0–$1,000). Tables that use different limits state those limits in the caption.

• What the numbers are: indicative annual premium ranges (e.g., 25th–75th percentile) for comparable risks, not quotes, offers, or guarantees. For compensation disclosure see How We Get Paid.

Data sources used

We combine multiple independent inputs to minimize single‑source bias and to reflect real buying conditions across Canada (ex‑QC):

• Anonymized Summit quoting records generated through our market placement process (multiple carriers and MGAs).

• Anonymized bound‑policy premium records (new business and renewals) from Summit’s brokerage systems.

• Documented carrier rate changes, underwriting bulletins, and appetite updates received during the data window.

• Structured broker market intel captured at time of marketing (e.g., minimum premiums, surcharges, endorsements required by class or territory).

• When relevant to normalization (see below), filed relativity patterns observable across quotes within the data window (e.g., limit/deductible relativities).

Privacy note: all analyses use de‑identified records and aggregate outputs; no customer‑identifiable data is published. Claims support information is handled separately per our Claim Services process.

Data windows and refresh cadence

• Primary window: rolling eight quarters of data to balance recency with stability.

• For the current publication (see Version tag below), the modeled window is January 1, 2024 through September 30, 2025 (Q1‑2024 → Q3‑2025).

• Refresh cadence: quarterly, within 15 days after quarter‑end (target on or before Jan 15, Apr 15, Jul 15, Oct 15).

• Backfill: if a table lacks sufficient class/region depth in the last eight quarters, we may extend to twelve quarters for that slice; any such extension is noted in the table’s footnote.

Normalization and comparability

To make heterogeneous records comparable, we apply the following normalizations before computing ranges:

• Limit and deductible: premiums are converted to the stated table baseline using observed market relativities across quotes in the window. Where relativities are insufficient for a slice, we default to carrier‑standard step factors observed for that class and note the assumption.

• Territory: records are mapped from postal code to province and, where relevant, to major metro/non‑metro groupings.

• Industry and class: accounts are mapped to Summit industry clusters that align with our public industry pages (e.g., construction and realty, professional services, technology). Mixed‑class risks are apportioned by revenue share when documented.

• Exposure bases: we standardize to the dominant exposure driver for the class (e.g., payroll, receipts, unit counts) and scale premiums accordingly to the table’s stated exposure benchmark.

• Endorsements: compulsory endorsements common to the class/region (e.g., additional insured, waiver of subrogation) are included when they are routinely required to place coverage; bespoke or unusual endorsements are excluded unless the table notes otherwise.

Quality control and outlier handling

• Inclusion: commercial accounts bound or quoted in the open market; program‑only or captive placements are excluded.

• Outliers: we remove statistical outliers using a robust rule (median absolute deviation and interquartile fences). Records removed for cause (e.g., severe prior losses, unusual contracts) are annotated in our internal audit trail.

• Minimums: explicit carrier minimum premiums are respected; if a class is dominated by minimums, we mark the table accordingly.

• Sample size guardrails: we do not publish a slice unless it meets internal minimum n‑counts; where n is marginal, ranges are broadened and flagged as “low sample.”

Weighting model (how each input influences the final range)

The consolidated cost range is a weighted synthesis across inputs. Higher weights favour executed market outcomes over preliminary indications.

Weights are reviewed annually; any change is versioned and disclosed in the change log.

Calculating the published ranges

• For each class/region/exposure slice, we compute the weighted empirical distribution after normalization.

• We publish the 25th–75th percentile as the default “typical range,” the median as a reference point, and may include the 10th–90th percentile when sample depth permits.

• When carrier minimums create a floor, the lower bound is set to the greater of the statistical bound and the prevailing minimum premium.

• For slices with high heterogeneity (e.g., mixed trades), we publish wider bands and annotate drivers of variance.

Versioning, dating, and table linkage

• Version tag: CGL‑Cost‑2025‑v1‑2025‑11‑13.

• Each published table includes a “Methodology & Sources” caption linking to this page and a local table‑level note stating: data window, baseline limits/deductible, exposure benchmark, and last refresh date.

• Quarterly refreshes increment the version tag (e.g., CGL‑Cost‑2025‑v2‑YYYY‑MM‑DD) and update table captions site‑wide.

What this page is and isn’t

• It is: a transparent description of how Summit aggregates, normalizes, and summarizes market pricing for CGL in Canada (ex‑QC).

• It is not: individual advice or a binding quote. Actual pricing depends on your specific operations, contracts, loss history, limits, and endorsements. For guidance or a custom market exercise, contact us via Summit Contact.

Change log

• 2025‑11‑13 (v1): Initial publication covering Q1‑2024 through Q3‑2025; baseline $2M occurrence; eight‑quarter rolling window; weighting table introduced; Quebec excluded by scope.

How to use and cite

• When citing a number, include slice, data window, baseline limits, and version tag. Example: “Median CGL premium for small professional services in BC (baseline $2M occurrence), Jan‑2024 to Sep‑2025, CGL‑Cost‑2025‑v1‑2025‑11‑13.”

• When embedding a Summit cost table elsewhere on summitcover.ca, include the caption link to this page and the table‑specific parameters so readers can reproduce the slice.

Questions or corrections

If you believe a table does not reflect current market conditions in your class or region, please reach out. We welcome carrier memos, appetite changes, or audit requests. Start at Summit Contact.