How a Comparison Table Earns Its Order

Every comparison table is an argument. Something decided that this product sits at the top and that one sits fourth, and the reader is usually left to guess what it was. Sometimes the answer is a genuine assessment. Sometimes it is whoever pays the highest commission per sale. The two look identical from the outside, which is precisely the problem.

So here is how ordering works on this site, followed by how to test any other comparison site you read.

What decides the order here

One number: the score. Each product is assessed against a four-part rubric with fixed weights, the rubric produces a score out of 10, and the table is sorted by that score in descending order. There is no manual override, no promoted slot, and no tie-breaking by commercial value. The weights are performance at 30 percent, features at 25 percent, value at 25 percent, and support and trust at 20 percent, and they are set out in full in our rating methodology.

A few consequences are worth spelling out, because they are what make the claim testable rather than decorative.

• Products we earn nothing from can and do outrank products we earn from.
• A product with no affiliate programme at all still gets ranked on merit.
• A price rise can push a product down the table without anything else changing, because value carries a quarter of the score.
• Where we have not finished assessing a product, it appears without a score and marked as a draft rather than being quietly placed somewhere convenient.

Where the money sits instead

We earn commission when a reader buys through a labelled link. That funds the site, and it is disclosed on every page that carries such a link and explained in full on our how we make money page. The separation is structural rather than a promise about our characters: commission rates are not an input to any part of the rubric, so there is no arithmetic path by which a better rate can move a product up.

What a brand can do is send us a correction. If a price is wrong or a feature has changed, we check it against the source and update the page. What a brand cannot do is preview a review, approve copy, buy a position, or have an accurate criticism removed.

Why weighted scoring rather than a simple ranking

Unweighted rankings hide judgement. If you simply declare a winner, every disagreement about the result is unfalsifiable. Weights make the judgement explicit: you can look at 25 percent for value and decide that value matters more to you than it does to us, then read the table differently. That is a feature. A methodology you can argue with is more useful than a verdict you cannot.

It also forces consistency. The same rubric applied to every product in a category means a score is a comparison rather than an impression, and it means a product's position changes only when something measurable changes.

What a score does not tell you

A score is relative to its category. An 8.4 in one category and an 8.4 in another are not equivalent, because the criteria that matter differ between categories. A score also cannot know your circumstances: the highest-scoring option is the best general choice, not automatically the best choice for a specific budget, team size or requirement. That is why every entry carries a short statement of who it suits and what to watch, and why the reviews are longer than the table.

How to audit any comparison site in five minutes

Apply these to us as well as to anyone else.

1. Look for a stated sorting rule. If the page never says what determines order, assume the order is commercial until shown otherwise.
2. Check whether the disclosure is specific. "We may receive compensation" in a footer is the minimum the law tolerates. A page explaining the model, near the links, is a sign someone is trying.
3. Hunt for unmonetised entries. If every product in the table has an affiliate button, the list may have been assembled from an affiliate network's catalogue rather than from the market.
4. Read the negatives. A table where every product has three strengths and no weaknesses is advertising with a scoring font.
5. Check the dates on prices. Undated prices are unverifiable, and in most categories they are stale.
6. Follow one factual claim to its source. Pick a specific number in a review and see whether anything supports it. This single test separates researched comparisons from rewritten marketing copy faster than anything else.
7. See what happens to a bad product. If the lowest-ranked entry still reads like a recommendation, ranking is not doing any work.

Holding us to it

If you find a page on this site where the order does not match the scores, or a score that the review text does not support, tell us and we will fix it or explain it. Corrections go to the editorial desk under our editorial standards, and the contact page has the address.