Methods & disclosure
Show the inputs.
Explain the limits.
Our guides are built around a specific advertising measurement question. A useful answer should let you reproduce the arithmetic and see what evidence is still missing.
Authorship and editorial review
MetricClue articles are written and edited under the pen name M. Clue. Each guide is checked against the formulas, stated assumptions and linked primary sources shown on the page. We do not claim independent subject-matter review, personal campaign results or credentials that are not documented.
How to read an example
Worked examples use illustrative inputs, not client results, site traffic or industry benchmarks. Formula relationships are derived from the definitions shown. When an example uses a margin, close rate or retention assumption, it is an input for that example, not a recommended value.
What our calculators include
ROAS divides compatible attributed revenue by positive ad spend. The order-level break-even model subtracts entered non-ad costs from order revenue. A positive remainder is the maximum break-even CPA; revenue divided by that remainder is the break-even ROAS.
Inputs must be finite and nonnegative. Blank fields are not treated as zero. Results are rounded for display; decisions near a boundary should use the underlying unrounded calculation. Zero spend makes ROAS undefined. Nonpositive order contribution leaves no positive acquisition allowance.
The target and traffic calculators are scenarios, not forecasts. They do not model taxes on profit, fixed overhead, future purchases, causal advertising lift or auction behavior unless explicitly stated.
Contribution targets and campaign comparisons
Target CPA = order revenue − entered non-ad costs − contribution to retain. Target ROAS = order revenue ÷ target CPA, where target CPA is positive. These are cost-based planning boundaries, not platform bid recommendations.
Campaign contribution is calculated only after you confirm that the entered order economics represent the campaign: campaign revenue × pre-ad contribution margin − ad spend. The spend allowance for a retained-contribution target assumes the same revenue and cost mix; it does not forecast what additional spending would produce. Saved comparisons remain in page memory only.
Local file analysis
The advertising CSV analyzer uses the browser FileReader API. The selected file is parsed in page memory and is not sent to MetricClue. It requires an explicit column schema, skips invalid rows rather than inventing values, and reports the number of valid and skipped rows. Its warnings test arithmetic consistency only; they do not validate platform attribution or prove incrementality.
Testing and change records
Interactive tools are tested with fixed examples that cover ordinary inputs, zero denominators, blank fields and invalid negative values. A substantive formula, source or limitation change updates the page date. Styling-only changes do not claim a new methodological review. Source-based tools identify the publication used and should be checked again when the source owner releases a new year or policy revision.
Downloadable examples
Downloadable templates contain fictional rows created to demonstrate the required schema. They are not campaign benchmarks, client data or evidence of expected performance. Third-party templates and anonymous datasets may inform field coverage, but MetricClue does not republish their layouts or describe synthetic data as observed results.
Sources and updates
Platform-specific claims link to official documentation. Account settings and product behavior can change. Check the cited source for your implementation. Page dates reflect publication or substantive content changes, not a claim that all linked platforms were independently retested that day.
Corrections
Report a correction with the page, formula or statement and enough non-sensitive context to reproduce the issue. If you are using a different revenue or cost basis, include that definition.
Commercial disclosure
The current guides contain no affiliate tracking links. If affiliate recommendations are added, the relationship will be disclosed near the relevant recommendation or link. Commercial compensation should not be represented as evidence of product performance.