Find What Pays Find yours

Methodology

How opportunities get selected, how facts get verified, how ranking works, and what we refuse to publish.

How we verify platform facts

Every fee, payout term and requirement traces to a cited source with a date. Currently 14 of 14 platform records have at least one fact confirmed from the platform's own documentation.

Primary sources beat secondary ones. When a platform's own documentation contradicts a review site, we publish the platform's figure. This is not a theoretical policy — it changed what we published twice while building the first fourteen records:

Both contradictions came from sites earning affiliate commission on signups. That is the norm in this category, not an aberration.

What we refuse to publish

Earnings estimates. You will not find "earn $500/month" anywhere on this site. Outcomes in every one of these markets are extremely skewed and depend on factors outside your control — your postcode, your demographics, local demand. Published averages are overwhelmingly survivorship-biased and usually produced by someone with a commission at stake.

Figures we could not confirm. 14 of our 14 platform records carry an explicit "what we could not verify" list. Leaving a gap visible is more useful than filling it with something that might be wrong.

How ranking works

The finder scores in two stages, and the distinction matters.

Blockers are hard constraints. If you say you will not ship physical items, opportunities requiring shipping are excluded, not quietly ranked lower. We show you what was excluded and why, so you can change an answer if we misread you.

Signals are soft fit. Asset match carries the most weight, then effort alignment, privacy headroom, face requirements and startup cost. Each contributing signal produces the human-readable reason you see under each recommendation — those are generated from the same values that produced the score, not written separately.

Scoring is pure and deterministic: the same answers always produce the same ranking. Your answers stay in your browser.

Why affiliate payouts cannot affect ranking

This is enforced by architecture rather than promised as a policy.

Affiliate configuration lives in a separate module from the content records. The data structure passed to the scoring engine contains no revenue field, and the scoring module does not import the affiliate config. Payout is not something the ranking code declines to consider — it is data the ranking code has no path to.

Platform tables are sorted alphabetically. Where we recommend one platform over another, the reason is stated in the text and traceable to a cited fact.

How opportunities get added

An opportunity is only added if all six of these hold:

  1. A real market exists.
  2. People can realistically participate.
  3. We can add meaningful decision value beyond what already exists.
  4. Reliable sources exist.
  5. It has monetization potential or materially improves how people find us.
  6. Maintaining it will not require disproportionate ongoing labour.

"It gets search traffic" is not on that list. We do not generate keyword permutation pages, and we currently cover 12 opportunities rather than several hundred thin ones.

Freshness

Every record carries a verification date, shown to you next to the facts it covers. Records flag themselves for re-checking once that date gets old. Platform terms change without notice, so confirm current fees on the platform itself before committing money or time.

Corrections

If a figure here is wrong, it should be fixed rather than defended. The data behind every page is source-controlled, and corrections are tracked as issues on the project repository.