Methodology
How Charity Compass computes and presents figures from Canadian charity filings.
What we do
Charity Compass presents information from charities' F2024 T3010 returns. Tier 1 figures are computed from open CRA data. The underlying returns are self-reported and we do not independently verify them.
Cents to cause
For operating charities, program spending (T3010 line 5000) is divided by total spending (line 4950). Filings with impossible or unreconciled allocation figures are not eligible to lead ranked results.
Operating overhead
Management and administration (line 5010) plus fundraising (line 5020), divided by total spending (line 4950). A lower percentage does not establish greater effectiveness.
Reserves
Cash, bank balances and short-term investments (T3010 line 4100) plus long-term investments (line 4140) are divided by annual program spending (line 5000) to estimate how many years liquid holdings could cover. Liabilities, restricted funds and endowments are not netted out, because the open data does not separate them; a charity that spends what it raises each year will show a low figure by design. This is a financial context measure, not a prediction of need or stability.
Foundations
Foundations are shown through the share of spending disbursed to qualified donees rather than an operating-charity program ratio.
Third-party analysis
Charity Intelligence ratings, when available, are labelled Tier 2 and attributed to its analysts. They are not blended into Charity Compass's Tier 1 calculations.
Ranking and search
Search combines the filters and priorities selected by the user. When a donor describes a cause in their own words, the ranking weighs cause match (40%), analyst impact where available (25%, neutral where unrated), thin reserves (20%), and spending within the 5–35% overhead range (15%). Every ingredient is shown on each result. Two switches change the weights and say so on the page: "Needs money now" doubles the weight of thin reserves (40%) and requires the reserves figure to exist; "Stricter match" raises cause match to 70% and lists more results as partial fits. Cause searches with no size chosen apply a $100K revenue floor, shown as a removable note. An algorithmic ranking is a comparison aid, not an assessment of wrongdoing, suitability, or real-world impact.
Last reviewed September 15, 2026.