How review platforms make money — and why it shapes what you read

Reviews are free to read. Someone else is paying, and what they get for their money explains most of what is wrong with this category.

Model one: businesses pay for tools

The dominant model. Businesses subscribe for invitation sending, analytics, widgets, integrations. Reading stays free.

The pressure it creates: the paying customer is the business being reviewed. Not necessarily corrupting — the tools are real and the money has to come from somewhere — but every product decision passes a question about whether it will annoy paying customers, and criticism annoys them.

What to check: whether any paid feature touches what readers see. Tools for collecting and responding are fine. Anything that changes which reviews appear, in what order, or whether a rating displays, is the rating being sold.

We use this model. The subscription buys invitation quota, widgets, an API and analytics. It buys nothing about what appears on a profile, and there is no tier where that changes.

Model two: advertising

The platform sells placement — sponsored positions in category listings, or ads on competitors' profiles.

The pressure it creates: a direct financial interest in which businesses look good, because a well-rated business is a better advertising prospect. It also produces the specific unpleasantness of a business's own profile carrying advertisements for its competitors.

What to check: whether sponsored placements are labelled, and whether the ordering of a category page is influenced by spend.

Model three: affiliate and lead generation

The platform earns a commission when a reader clicks through and buys. Common in insurance, broadband, energy and software comparison.

The pressure it creates: the sharpest of the four. The platform earns more when readers choose higher-commission products, and rankings are the mechanism. A "best of" list on an affiliate site is a commission table unless proven otherwise.

What to check: whether commission relationships are disclosed per listing, and whether products that pay nothing can rank first.

Model four: consumers pay

Rare. A subscription to read.

The pressure it creates: almost none of the above, which is why the model is theoretically attractive. It fails commercially — very few people pay for review access — and it creates its own problem, that a paywalled rating cannot be checked by the people it describes.

The questions worth asking

Whatever the model, four things reveal whether it has bent the product:

Can a business pay to remove, hide or reorder a review? If yes, nothing on the site means anything. This is the only question with a single acceptable answer.

Is removal logged and counted publicly? Every platform removes reviews and should. A platform that will not say how often is asking for trust it has not earned.

Does the ordering favour anyone who paid? Including subtly — "relevance" that correlates with subscription tier.

Can a business leave? If a profile cannot be closed and the business has no way out, the platform is holding its reputation hostage while selling it tools to manage it. That is the complaint underneath most of the anger directed at this industry, and it is a fair one.

Where we sit

Businesses pay for tools. £39 a month buys invitation quota, widgets, an API, analytics and a verified badge that means the company was checked against Companies House and proved control of its domain.

It does not buy removal, ordering, suppression of criticism, or advance sight of a review. There is no tier where it does, and if we ever added one this paragraph would be the evidence that we changed.

Every removal is logged, appealable, and counted in published totals. A business can close its profile and leave.

We are not asking you to take that on trust. We are asking you to check it — and to ask the same four questions of everyone else.