Five dimensions. Fixed weights. Calculated before the article is written, and never influenced by an affiliate relationship. This page shows you the gears.
The score is locked before a single sentence of the review is drafted. It is never reverse-engineered to fit a conclusion.
Affiliate relationships never touch the number. A tool that pays a commission is scored on the identical axes as one that pays nothing.
Every weighted average is independently recomputed in Python or Excel before it appears on the site.
A star rating with no published weights can be nudged to protect a commission, and readers have no way to catch it. The WillStreet Score exists to remove that option. The five dimensions and their weights are public, they never change per review, and the math is simple enough to check by hand.
Every tool is tested on real Canadian tasks: CAD pricing, TFSA and RRSP math, CRA context. The same instrument, every time. That is the whole trick.
The five dimensions never fork by category. A budgeting tool and a tax tool are scored on identical axes, because cross-tool comparability is the entire point.
Fixed since day one. If these weights ever change, the change and the reason are published.
Can the tool do the job, correctly, without inventing numbers? This is the heaviest weight because a cheap, friendly tool that gets your TFSA math wrong is worse than useless.
Pricing is judged in CAD, after FX and provincial tax, against what the tool delivers and what its alternatives charge. A USD sticker price is never presented as the Canadian cost.
How long from signup to a useful result. A powerful tool that takes a weekend to learn scores lower than one that gets you an answer in ten minutes.
Does the tool speak Canadian money natively? CAD support, TFSA, RRSP and FHSA account types, and whether its tax answers hold up against CRA rules rather than IRS rules with a maple leaf on top.
What happens when something breaks. Documentation depth, response time, and whether the community around the tool can actually answer a question.
Score an imaginary tool on each dimension from 1 to 10. The weighted average and verdict compute live, using the exact formula every published review uses. No hidden adjustments. Nowhere to hide one.
Score = (AI × .30) + (Value × .25) + (Ease × .20) + (Canadian × .15) + (Support × .10) rounded to one decimal
The tool earned it across the weighted axes. The review still names its real flaws, because a score is not a free pass.
Good for a specific reader in a specific situation. The conditions are always stated. If they do not describe you, keep your money.
The review still runs, because a documented failure is useful to the next Canadian who searches for the tool.
Every number below came out of this exact methodology. Nothing here is a sample. Click through to see the full test behind each score.
Dimension scores are locked before drafting begins. The article explains the number. The number never bends to fit the article.
Some links on WillStreet may be affiliate links, and this is always disclosed. The score is calculated the same way whether a relationship exists or not. Claude and ChatGPT have no public consumer affiliate program, so WillStreet earns no commission from either.
Every score comes from hands-on testing on Canadian tasks, with the raw findings shown in the review. If a result is $192.55 off, the review says $192.55, not roughly $200.
Weighted averages are recomputed in Python or Excel before publish. A score never ships on a single calculation.
The five dimensions never change by category. That keeps every score on the site comparable to every other score.
Pricing changes over 10%, a new model release, or new CRA guidance triggers a re-test. Each review carries its next scheduled review date.