• checklist
  • dataset

Review page checklist for affiliate sites (what 13 sites do)

Section order, how many of 13 review sites have each block, thresholds, a do-not-copy list and a 28-point page acceptance. Checklist, CSV and example.

  • v1.0
  • updated 9 November 2026
  • CC-BY-4.0
  • Vladislav Krivorutško

This checklist lists what the review pages of 13 competitor sites put on the page, in what order, and which blocks almost none of them have. It has 49 rows with counts written as "n of d", 12 rules with thresholds, a list of what not to copy, and a 28-point acceptance you run on a finished page. The data comes from a snapshot taken in 2026-11 of two national niches, "casino affiliate (BG, SK)": 7 Bulgarian sites and 6 Slovak ones. The niche is only the source of the counts. The structure fits any site that reviews one product or provider per page.

Of the 49 blocks, 10 are standard, 20 are found on some sites, 14 were counted and found on none, and 5 are our own rules with no count behind them. The gaps are the useful part. Timed test results, for example, appear on 1 of 6 Bulgarian review sites and on none of the 8 Slovak reviews that were analysed. An offer-expired plate that switches on by itself was found in none of the 4 Slovak templates.

Files in this resource

4 files · or all in one zip
  • checklist.md

    Section order, 49 rows with counts, rules, do-not-copy list and the 28-point acceptance

    15 KB
  • checklist.csv

    The same 49 rows for a spreadsheet: counts for two markets, status, basis, rule, comment

    10 KB
  • example-filled.md

    The checklist and the acceptance filled in on an invented page

    4.4 KB
  • README.md

    Three steps, how to read the counts, limits

    1.9 KB

Who it is for

You run or build a site with a page per product, provider or offer (hosting, software, banks, courses, a regulated niche), and you want to know what to put on that page and in what order. You get a list to build from and a test to run on a page before it goes live.

How to use it

  1. Open checklist.md and read "How to read the numbers" first. BG counts are in sites, SK counts are in templates.
  2. Run the 28-point acceptance on one of your review pages and write yes or no for each point. example-filled.md shows a filled copy.
  3. Fix the no answers, starting with points 1 to 6. Filter checklist.csv by status = gap to choose where to be first.

What is inside

  • Section order of a review page in ten steps, from the H1 to the sibling reviews.
  • 49 rows in 11 sections: top of page, passport, offer, test results, body, scoring, verdict, author, proof and markup, legal and trust, review set. Each row has the BG count, the SK count, a status (standard, edge, gap, my take), and whether it was measured or is our hypothesis.
  • 12 rules with thresholds: one button, a score as a number before the text, a passport with a credential row and a check date, test results with timings and a method link, a fixed table row order, fixed axes, honest cons, an author with proof of experience, one date field that drives the offer expiry plate.
  • Do-not-copy list of 14 patterns seen in the sample.
  • Acceptance of 28 yes-or-no points.

Rather have this done for you?

Send the site address and what you want to know. We run the audit and send the report.

Send a request

Limits and risks

  • Presence is not effect. We counted what sites do. We did not measure rankings or conversion, so a gap is a gap in the sample, not proof that filling it works.
  • Two small niches in one regulated industry. Other niches may order blocks differently.
  • Different bases. BG is counted in sites that have reviews (6 of 7 sites), SK in templates (4, because three sites share one layout). Do not add the two.
  • A dash is not a zero. "-" means the source has no count.
  • Our own thresholds. The 90-day re-check, the 4-line minimum for test results and the 2-con minimum are starting values.

Tested on

2026-11-09: the checklist was applied by hand to one invented hosting-review page (example-filled.md), which scored 26 of 28 on the acceptance, and the CSV was parsed with Python 3 (49 rows, 8 columns, no row of a different length). The counts were re-read against a separate verification of the source data on the same date, which corrected four figures: the SK leader has 22 review pages (21 per the register plus one for an operator that lost its licence), timed test results exist on 1 of 6 BG sites, none of the analysed SK reviews has them, and no SK template has an automatic offer-expired plate. No part of the checklist was tested on a live production site. We did not test whether following it changes search results.

Files are provided as is: run them on your own data and check the result.

Want an audit like this on your site?

If you would rather have it done for you, send a request and describe the site.

Send a request
Download checklist.md