How Navneet Reviews AI Tools

By Navneet Arya ·

Key takeaways

  • Every review follows the same 5-step process: docs, verified reviews, pricing, competitor benchmarking, and freshness checks.
  • At least 100 verified reviews per tool are read from Trustpilot, G2, and Capterra before a verdict is written.
  • Affiliate status is checked after the rating is set, not before — it never influences the score.
  1. Official documentation review — Every feature claim is checked against the tool's official documentation and changelog. Marketing copy is not used as a source. If the docs say a feature is paid-only, limited, or in beta, the review says so.
  2. Verified user review aggregation — At least 100 verified reviews per tool are read from Trustpilot, G2, and Capterra. Reddit threads are checked too, for real-world sentiment that doesn't show up on structured review platforms.
  3. Pricing verification — Pricing is confirmed on the tool's live pricing page, not from cached data or third-party aggregators. Free-plan limits are written out explicitly. INR equivalents are included for Indian creators and freelancers.
  4. Competitor benchmarking — No tool is reviewed alone. Each one is compared against its 2–3 closest alternatives on feature parity, pricing, and use-case fit, using the same criteria each time.
  5. Review freshness — updated on tool changes — Reviews are updated when tools ship major features or change pricing. Anything untouched for over 6 months gets flagged for re-verification.

The 6-step research process, quantified

  1. Official documentation & interface research (~20 min) — Free account access and official docs are used to research standard workflows. Every feature claim is checked against the docs before it goes in a review.
  2. Free plan claim verification (~15 min) — Each free-tier claim is checked against what the tool actually delivers: whether a card is required, and where "free" features are rate-limited.
  3. Analysis of 100+ verified user reviews (~45 min) — At least 100 verified reviews from Trustpilot, G2, and Capterra are pulled and tagged for recurring complaints and praise.
  4. Reddit sentiment check (~20 min) — The top posts in relevant subreddits from the past year are read for long-term frustrations that don't surface on review platforms for months.
  5. Live pricing page verification (~10 min) — Every tier's price, billing frequency, and cancellation terms are documented straight from the pricing page.
  6. Feature comparison against 3 nearest competitors (~30 min) — Key features are mapped against the closest alternatives to build the comparison tables used in every review.

Frequently asked questions

Does Navneet Arya independently research every AI tool?

Yes. Every tool reviewed on AI Nexus is researched based on features, pricing, official documentation, and verified user reviews before a review is published.

How long does research take before a review goes live?

At least 2–4 weeks for a single tool review. Comparison articles take 4–6 weeks since they require side-by-side testing of 2–3 tools.

Does an affiliate commission change the rating?

No. Tools are rated before checking whether an affiliate programme exists. Critical reviews have been published for tools with affiliate programmes, and positive reviews for tools without one.

Reviewed by Navneet Arya — Independent AI Tools Researcher, AI Automation & Performance Testing Leader at BOLD. Researching AI tools since 2022.
Follow: X/Twitter · LinkedIn · GitHub · About the reviewer →