Ranking methodology (v1)
Our ranking is public, versioned and verifiable. No payment ever enters the calculation.
1. Unit
We rank a tool for a given use case, never "in general".
2. Reviews
- Each review is tied to a use case, a profile, usage duration and plan.
- At least one con is mandatory.
- Optional numeric outcome (followers, meetings, clients, hours) with period.
3. Verification levels
| Level | Proof | Weight w |
|---|---|---|
| 1 — Declared | Identified account (LinkedIn OpenID or work email) | 1 |
| 2 — Proof of use | Invoice/dashboard screenshot checked by agent then deleted | 2 |
| 3 — Verified outcome | Before/after stats verified (agent + human spot checks) | 3 |
4. Formula
review_rating = (2×effectiveness + ease + value) / 4 review_weight = w(level) × freshness freshness = 0.5^(age_months / 12) R = Σ(weight×rating) / Σ(weight) Score = (C×m + Σweight×R) / (C + Σweight) (Bayesian average; m = use-case mean, C = 10) If Σweight < 5: "provisional ranking"; agent test (if run) may count up to 20%. Non-compliant tools are excluded.
5. Anti-fraud
- 1 review per account, tool and use case.
- Detection: review spikes, text similarity, new accounts, IP/device clusters, vendor domains.
- No vendor incentives; vendor-solicited reviews labeled.
- Public vendor replies; no removal on request.
- Monthly transparency report.
6. ToS compliance
Badges: ✓ Compliant, ⚠ Grey area, ✕ Non-compliant (tools automating invites, messages, likes, comments or scraping on LinkedIn/Instagram — LinkedIn User Agreement §8.2, Instagram Terms). Non-compliant tools are excluded.
7. Sponsorship
Paid placements sit outside rankings, are labeled "Sponsored" and listed in the public register. Sales (including our agents) has no write access to scores or reviews.
8. Disclosure
AIGrowthRank is founded by Marion Carneiro, who also runs an Instagram growth agency (My Name Is Bond). Her own services are not ranked.
Changelog
v1 — 2026-10-15: initial release.