This is the document behind every review we publish. It lays out the workflow we follow, the fact-checks we run, how we handle corrections, and the rules the commercial side cannot break. It is written as a working policy, not as marketing about how trustworthy we are.
Four ideas shape everything below.
Accuracy. Every number in a review traces back to a primary source or a timestamped test record. If we cannot produce the source, the number does not go in.
Independence. No casino, affiliate network, or commercial partner has a say in review scoring, cons lists, or section ordering. Draft reviews are not shared with operators before publication.
Transparency. Byline on every review. Dates on every review. Methodology on every review (links to How We Test and How We Rate). Commercial relationship spelt out on Affiliate Disclosure.
Currency. The Australian online casino market moves. A review written in March is often stale by September. We re-test every 90 days on a rolling schedule, and we flag the re-test date visibly at the top of each review.
Before we invest a testing week, we run a screening pass. The casino must hold a current licence from a recognised regulator, listed on the regulator's own registry. For Jackpot Jill that means the Curaçao Gaming Control Board under the post-LOK framework, verifiable on the GCB public register. We also check the operator's jurisdiction, the parent company (including sister casinos), and recent complaint patterns on AskGamblers and Casino Guru. A casino with unresolved payment-dispute patterns or an expired licence does not proceed to a full review; it is flagged for a short-form "Why we are not reviewing this" note if readers ask.
The author registers a real account, uploads real ID documents, makes a real deposit, plays through a real bonus, submits a real withdrawal, and times everything. No demo accounts, no operator-supplied balances, no "VIP treatment" for reviewers. The full procedure is at How We Test. A single test cycle runs 5–10 days depending on KYC queue length and withdrawal rail.
The person who tested writes the review. The output is a draft of roughly 2,000–2,500 words with inline references to screenshots, transaction IDs, chat transcripts, and timestamps held in an editorial file. First-person voice stays in — if I tested a withdrawal and it cleared in 24 hours, the review says "my withdrawal cleared in 24 hours", not "withdrawals clear in 24 hours".
Before publication, every verifiable claim is checked against a primary source. Bonus amount and wagering against the casino's own Terms & Conditions page. Licence status on the regulator's register. Provider list against the casino's own lobby at the date of check. Withdrawal times against the timestamped test record. If the draft and the source disagree, the draft loses and gets corrected. Fact-check turnaround is normally 24–48 hours.
Reviews go live under the author's name, with the test date, the current score, and the scoring framework linked. No anonymous reviews. No "written by the editors" copouts. The byline points to a full author page with credentials (Oliver Taylor). Structured data (Article schema with author as Person) is emitted so search engines receive the attribution cleanly.
Every published review goes through a compressed re-test every 90 days. The compressed re-test covers: does the bonus still exist at the same terms, are the payment methods unchanged, does the licence remain active, has anything changed in the cons list. If something moves, the review is updated and the "Last updated" date on the page changes to match. If nothing has moved, the re-test date in the editorial file is updated but the public page does not change. Every twelve months, the review goes through a full end-to-end re-test identical to the original.
"Fact-check" is a vague word outside journalism. Inside our workflow it means the following sources, checked for the following claims:
If a claim cannot be traced to one of the above sources, the claim is removed from the draft. "Industry reports say" and "according to forums" do not survive fact-check. Where we cite secondary sources — a news report about a regulator action, an AskGamblers complaint thread — we link to them directly and let the reader evaluate.
Commercial relationships — the affiliate agreement with Jackpot Jill — are handled on the business side of Ovoda. The editorial side (Oliver as writer, the small freelance bench for copy-edit, the fact-checker) does not negotiate commercial terms. The commercial side (one person handling partnership conversations and invoice admin) does not review drafts, assign scores, or approve publication.
In practice this means the affiliate manager at Jackpot Jill learns what we wrote when it goes live. Commission renegotiation does not produce review updates. Partnership termination would not produce review deletion — we would continue to maintain the review and disclose the changed commercial status. These are not abstract principles; they are the rules that get triggered whenever a partner's interests diverge from a reader's.
One concrete example from the last review cycle: Jackpot Jill increased wagering from 40× to 50× in late March 2026 and quietly dropped PayID support. We updated the review within four days of detecting the changes, dropped the score from 4.3/5 to 4.1/5, and added the PayID absence to the cons list as a "standout omission". Our affiliate manager pushed back. The score stayed at 4.1. That is what the firewall looks like when it is actually load-bearing.
Readers find errors in our reviews. It happens. When you email [email protected] with a correction, the process is:
When a casino closes, a bonus is discontinued, or a licence is revoked, the review gets a banner at the top stating what happened and when. We do not delete historical reviews; they stay available as a record of what the product was at the time.
Ovoda's byline standard is simple. To publish a review on this site, an author must:
Current editor: Oliver Taylor, iGaming Reviewer with 10 years in the industry and 200+ operators personally tested. Future contributors, if we add any, will meet the same bar.
Where we stand on AI-generated text: we do not publish it. Reviews are drafted by the author who ran the test. AI tools are used narrowly for spell-check, grammar-check, and occasional structural suggestions on long drafts — the same way a working journalist uses Grammarly. No AI is in the loop for claim generation, scoring, or fact-checking. This is a deliberate policy, partly because Google's December 2025 Core Update made it explicit that AI-drafted gambling content without author expertise is "Lowest Quality", and mostly because the reviews we want to publish require direct experience that a language model does not have.
The commercial side of the same system: Affiliate Disclosure. The specific testing procedure: How We Test. The scoring framework: How We Rate. Author credentials: Oliver Taylor. Questions about this policy: [email protected].
