Methodology

Step one

1. Sources

Our comparison dataset comes from the offer walls that publish their offer lists publicly. On 2026-08-14, two of the four partners on this site did:

• EarnLab — 6,207 offer rows captured from its public pages on 2026-08-14, no login required.
• Gain.gg — 2,403 offer rows captured from its public pages on 2026-08-14, no login required.
• Freecash — measured in part: 4 featured offers published in USD (Sunshine Island $391 · Match Masters $280 · Board Kings $32 · Dice Dreams $5). Featured offers only — the full wall requires a login, so these highlights are shown as captured and kept out of the spread statistics.
• Swagbucks — listed, not measured: its payout wall is not publicly accessible (login required), so we publish no payout figure for it.

We use publicly served responses only. We do not create accounts to extract gated data, and where a site’s robots.txt disallows a path or it refuses an unauthenticated request, we record that refusal as a finding rather than working around it. That is why two of the four partners carry a status line instead of a payout figure.

2. Currency and point conversion

Both ranked walls price offers in coins. We convert using each site’s own published rate — 1,000 coins = $1.00 on both walls at capture time — never an assumed one.

Where a site serves us a currency we cannot convert from a primary source, we exclude the figure from cross-site comparison rather than apply an FX rate of our own choosing to make a table look complete (user-facing copy is USD-only, GR-20260810-5).

Step two

3. The matching rule

Two offers are treated as the same offer only when the advertiser and the device platform (android, ios or desktop) both match. For milestone-level comparison we additionally require the milestone tier to match, so a “finish level 500” tier is never compared against a “finish level 30” tier. This is the strictest comparison the public data supports.

Difference (which we call spread) is calculated as (higher − lower) / lower, expressed as a percentage. We call a difference material at 20% or more.

The dataset

4. What the 2026-08-14 capture produced

Every figure above renders from a single versioned capture dataset dated 2026-08-14. When a new capture lands, the dataset is swapped and every page on the site re-renders from it — no figure is hand-edited anywhere else.

The honest part

5. Known confounds and limits

These are the reasons a reader should treat our numbers as evidence rather than verdict. We would rather state them than have them found:

1. Different underlying networks. The two ranked walls do not source offers from the same advertiser networks. A difference is therefore between wall-and-network combinations, not proof of one site’s generosity.
2. Two of four partners fully measured. Our comparison rests on the two walls with milestone-comparable public data. Freecash publishes only featured-offer highlights without a login (shown as captured, excluded from spread statistics); Swagbucks’ wall requires a login. Widening the claims waits on more data, not the other way around.
3. Direction is not uniform. One wall was higher in 163 of 174 whole-offer pairs — most, not all.
4. Small denominators inflate percentages. A milestone moving from $0.0040 to $0.0600 is a 1,400% difference and also five cents. We publish the underlying dollar figures next to every percentage for this reason.
5. Single geography. The 2026-08-14 capture ran from a US connection. Offer availability and rates vary by country.
6. Listed, not credited. We compare what sites advertise. We hold no accounts, so we have not verified that a listed rate is what actually credits on completion.
7. A snapshot, not a feed. Every figure carries a capture timestamp because these rates change without notice. An undated comparison table is a stale one.

Keeping it current

6. Dataset freshness

The site renders from one versioned capture dataset — currently the 2026-08-14 capture. A refresh is a dataset swap, not a rewrite: the new capture is ingested, every figure and partner status on the site re-renders from it, and the date stamp above moves with it. A partner whose wall becomes publicly measurable flips from “listed” to ranked automatically; one that closes its public data flips back.

The standing offer

7. Corrections

If a figure here is wrong, we want to fix it and say that we did. Site operators and readers can reach us via the contact page. Corrections change the page and its last-updated stamp; we do not quietly edit numbers.