How to Spot Fake Software Reviews and Sponsored Rankings Before You Shortlist

The hotel listed first on a booking site may be the best in town, or the one paying the highest commission. Seasoned travellers look for the small “sponsored” label and read the complaints first. Software buyers need the same habits, because a poor platform costs years, not a night. This guide shows how to spot fake reviews, test whether a ranking was earned or bought, and build a shortlist you can defend.

Why distorted reviews are hard to see

Most distortion comes from how reviews are collected and how rankings are funded, not from outright fabrication. A software company may invite only its happiest customers, reward reviews in ways that nudge sentiment, or let staff review without disclosure.

Comparison sites often earn money from the companies they list, through sponsored placement, payment per lead or affiliate commission. These models can be legitimate, but they change what “top” means. The result is a shortlist shaped by marketing budgets.

Regulators have acted against the worst practices. The US Federal Trade Commission's rule banning fake reviews has applied since 21 October 2024, and in December 2025 the FTC warned 10 companies. The UK banned fake reviews under the Digital Markets, Competition and Consumers Act 2024, with fines of up to 10% of global turnover since April 2025.

India's IS 19000:2022 sets a voluntary standard for online reviews. None of this removes sponsored placement, which is lawful when disclosed, so careful reading remains the buyer's task.

Six patterns that signal unreliable reviews

One suspicious review proves little; a pattern across dozens tells you a great deal. Assess the profile as a whole before reading any single review closely.

For a single review, apply a specificity test. Note whether it names a concrete task, gives a time frame, mentions a limitation and compares with a previous tool. A review meeting two of these is probably written from experience; one meeting none tells you nothing usable.

The earned-or-bought test for rankings

A ranking is an editorial product, not a review. A “leader” label is only as meaningful as the method behind it. Put every list through five checks before it shapes your shortlist:

• Label: whether any position is marked sponsored, featured or partner, and how visibly.
• Method: whether the site publishes how positions are set, and whether payment affects order.
• Money: how the site earns revenue, and whether order tracks spend per lead or click.
• Independence: whether a company in the category controls the site. The FTC rule bans presenting such a site as independent.
• Reproducibility: whether the order changes sensibly when you filter by size and sector.

If a list fails two or more checks, use it only to discover names. Badges such as “top rated” deserve the same scrutiny. A badge built on review volume inherits every weakness of those reviews.

Worked example: re-ranking a top five (hypothetical)

A 600-person logistics company needs a contract management tool. The team applies the pattern checks, the earned-or-bought test and one evidence request: a current SOC 2 Type II report or ISO/IEC 27001 certificate covering the product. A SOC 2 Type II report is an independent CPA firm's view of how controls operated over a period.

Product A is not rejected for sponsorship, which is lawful and says nothing about quality. It is held because its visibility is not yet matched by evidence. Product D has 300 reviews, but only one from a comparable firm and none critical. The committee receives B, C and E, each with a one-line reason.

From reviews to a defensible shortlist

• Start wide: use rankings to gather eight to ten names, not to choose three.
• Filter to your context: keep reviews from your size, sector and roles, written in the last two years.
• Read critical reviews first: they show how the product fails and how the company responds.
• Choose your own references: request a customer list rather than accepting nominated advocates.
• Request independent evidence: reviews are opinions, not audits.
• Record the reason: one line per product explaining why it is in or out.

More of this reading is now done by machines. In Gartner research presented in May 2026, 45% of B2B buyers used generative AI in a recent purchase. An assistant that summarises reviews inherits their distortions unless you ask about sponsorship and sources.