Trust layer

Comparison Methodology

How CalcuTools should compare financial products without turning a comparison table into a hidden advertisement.

1. Define criteria

State the measurable fields first: rate, tenure, fees, liquidity, tax treatment, eligibility, lock-in, or another relevant attribute.

2. Date the data

Show an effective date or “user-entered assumption” so visitors can distinguish current sourced data from estimates.

3. Separate facts from conclusions

Show the numbers and trade-offs. Avoid “best” claims unless they are narrowly defined and evidence-backed.

4. Label monetization

Affiliate or referral relationships must be disclosed and must not alter the underlying calculation.

Scoring rule

CalcuTools should avoid a single universal “winner” score for financial products. Where a score is useful, show the component metrics and explain what the score means.

Compare like with like, then show what the comparison leaves out

A fair comparison uses the same amount, time horizon, tax basis and cash-flow timing across options. Inputs should be sourced, dated and labelled as user-entered, official terms or illustrative assumptions. A headline interest rate is rarely enough: fees, exit restrictions, risk, liquidity and tax may change which product suits a particular objective.

Limits and conflicts should remain visible

Rates and rules can change after a comparison is published. Readers should be able to distinguish paid placement or affiliate links from editorial analysis and calculations. A methodology note explains how numbers are compared; it does not certify a provider, validate every input or replace reading the official contract or product disclosure.

How to make a fair comparison

Compare options using the same horizon, principal, timing and outcome measure. A monthly payment, total repayment, after-tax return and liquidity are different measures, so a single number should not be treated as a complete ranking.

Record the source, effective date, fees and exclusions for any rate-sensitive input. Where data are missing, label the assumption rather than filling the gap with an invented value. This methodology explains the model; it does not imply that every option can be ranked for every person.