Skip to main content
AI Tool Radar

How We Rate AI Tools

Our transparent, data-driven rating methodology. No guesswork, no paid placements, no inflated scores.

Data over opinions

We don't guess. Our ratings combine real user data from G2 (the largest B2B software review platform), Trustpilot, and community feedback with our own systematic feature analysis.

Every tool is evaluated on the same five criteria with fixed weights. This ensures a ChatGPT guide is scored on the exact same scale as a Koala Writer guide. No category gets special treatment.

Where G2 sub-ratings are available (37 of our 41 tools have G2 profiles), those numbers feed directly into our scoring. For the remaining tools, we rely on Trustpilot scores and manual community research.

Tools covered41
With G2 sub-ratings37
Total G2 reviews analyzed50,000+
Rating categories5

Five dimensions, weighted by impact

Each tool is scored 1.0 to 5.0 across five categories. The weighted sum produces the overall rating.

Features

30% weight

Completeness of feature set compared to category leaders. Based on official documentation, changelog analysis, and feature-by-feature comparison with the top 3 tools in each category.

Source: Manual research + documentation review

Ease of Use

20% weight

Sourced from G2 'Ease of Use' and 'Ease of Setup' ratings where available. For tools with fewer than 20 G2 reviews, we supplement with community feedback from Reddit and Trustpilot.

Source: G2 Ease of Use + Ease of Setup ratings

Value for Money

20% weight

Calculated by comparing features-per-dollar against the category average. Free tiers, pricing transparency, and billing flexibility (monthly vs annual, credit rollover) are all factored in.

Source: Pricing analysis + category benchmarking

Performance

15% weight

Based on the G2 'Meets Requirements' rating and community reports about speed, reliability, output quality, and uptime.

Source: G2 Meets Requirements + community reports

Support

15% weight

Average of G2 'Quality of Support' rating and Trustpilot score where available. We evaluate response times, support channels (chat, email, phone), and documentation quality.

Source: G2 Quality of Support + Trustpilot

How the overall score is calculated

Overall = (Features x 0.30) + (Ease of Use x 0.20) + (Value x 0.20) + (Performance x 0.15) + (Support x 0.15)

The result is rounded to one decimal place. Each sub-score ranges from 1.0 to 5.0 (matching the standard 5-star scale). G2 ratings that use a 10-point scale are converted to our 5-point scale by dividing by 2.

Example calculation

A tool with Features 4.5, Ease of Use 4.7, Value 4.0, Performance 4.3, Support 4.6:

(4.5 x 0.30) + (4.7 x 0.20) + (4.0 x 0.20) + (4.3 x 0.15) + (4.6 x 0.15)
= 1.35 + 0.94 + 0.80 + 0.645 + 0.69
= 4.425 = 4.4

G2 sub-ratings we track

G2 provides granular sub-ratings beyond the headline score. We map these directly to our rating categories.

G2 Sub-RatingMaps ToG2 ScaleOur Scale
Ease of UseEase of Use (20%)0 - 10/ 2 = 0 - 5
Ease of SetupEase of Use (20%)0 - 10/ 2 = 0 - 5
Quality of SupportSupport (15%)0 - 10/ 2 = 0 - 5
Meets RequirementsPerformance (15%)0 - 10/ 2 = 0 - 5

Most-reviewed tools on G2 (our coverage)

Grammarly13,187 G2 reviews
Notion10,149 G2 reviews
Canva AI6,896 G2 reviews
Semrush3,287 G2 reviews
Synthesia2,375 G2 reviews
Loom AI2,318 G2 reviews
Writesonic2,089 G2 reviews

Where our data comes from

PlatformWhat We PullCoverage
G2Ease of Use, Ease of Setup, Quality of Support, Meets Requirements37 of 41 tools have G2 profiles
TrustpilotOverall satisfaction, support quality, billing transparencySupplementary source for tools with limited G2 data
Official docsFeature lists, pricing pages, changelogs, API documentationAll 41 tools
Reddit / communitiesReal-world experiences, workarounds, complaintsSupplementary for all tools

What we don't do

No paid placements

We don't accept payment for higher ratings. Some of our links are affiliate links (clearly disclosed), but affiliate relationships never influence our scores.

No inflated scores

If a tool has weak support or confusing pricing, the score reflects that. We don't round up to be kind.

No unresearched ratings

We don't rate tools we haven't researched. If we can't find sufficient data to justify a score, we say so.

No hidden methodology

The formula, weights, and data sources are all documented on this page. You can verify our math.

What we can't verify

No methodology is perfect. Here is what we are transparent about:

  • --G2 review authenticity. G2 has its own verification process, but we cannot independently confirm every review is genuine. We mitigate this by looking at review volume (tools with 100+ reviews are more reliable than those with 5).
  • --Real-time accuracy. AI tools change rapidly. A tool may ship a major update the day after we publish. We update ratings monthly, but there will always be a lag.
  • --Subjective features scoring. The Features category (30% weight) involves judgment calls about what constitutes a complete feature set. We compare against category leaders to anchor this as objectively as possible.
  • --Tools without G2 profiles. 4 of our 41 tools (Pika, Koala Writer, and others) lack G2 profiles entirely. Their Ease of Use, Performance, and Support scores are estimated from Trustpilot, Reddit, and our own research. We note this in the individual guides.

How we keep ratings current

Monthly recalculation

Ratings are recalculated monthly when new G2 and Trustpilot data becomes available. The "Last Updated" date on each tool page reflects the most recent data pull.

Major update triggers

Significant events like pricing changes, new model releases, or major feature additions trigger an immediate re-evaluation outside the monthly cycle.

Version history

Each guide includes a changelog of significant rating changes, so you can see how scores have evolved over time and why.

How community feedback shapes our guides

Reddit discussions, forum posts, and community feedback are used as supplementary qualitative data across all tool evaluations. They do not directly affect the numerical score, but they inform our analysis in important ways:

What community data influences

  • +Strengths and Limitations section (real user pain points)
  • +Use case recommendations (who benefits, who does not)
  • +Hidden costs and gotchas not visible on pricing pages
  • +Support quality assessment (response times, resolution)

What community data does NOT influence

  • -The numerical rating formula (scores are calculated, not subjective)
  • -Feature completeness assessment (based on official docs)
  • -Pricing data (always from official pricing pages)

Third-party data attribution

G2 Ratings: All G2 ratings, review counts, and sub-ratings referenced on this website are sourced from publicly available G2 product pages. We are not affiliated with, endorsed by, or partnered with G2.com. G2 is a registered trademark of G2.com, Inc. For the most current ratings, visit g2.com.

Trustpilot: Trustpilot ratings referenced on this website are sourced from publicly available Trustpilot business pages. We are not affiliated with Trustpilot. Trustpilot is a registered trademark of Trustpilot A/S.

Reddit & Community Sources: Community insights referenced in our guides are sourced from publicly available Reddit posts and discussion threads. Individual user experiences may not be representative of the overall product quality.

Pricing Data: All pricing information is sourced directly from official product pricing pages and verified at the date noted in each guide. Prices may change without notice. Always check the official website for current pricing.

Affiliate Disclosure: This website contains affiliate links. We may earn a commission at no extra cost to you. This does not influence our ratings, which are calculated using the formula documented above. Read our full disclosure.

Questions about our methodology?

We believe in transparency. If you have questions about how we scored a specific tool, or if you think we got something wrong, let us know.