Rank Kiwi Research Methodology & Editorial Standards
How Rank Kiwi evaluates data, calculates derived ratios, attributes published research, and ensures empirical honesty across all creator performance analyses.
1. What Counts as Rank Kiwi Research?
Rank Kiwi Research is an editorial and quantitative initiative dedicated to studying creator content distribution, format dynamics, and performance outliers. We do not publish generic content-marketing listicles. Every piece must answer a specific empirical question with verifiable data.
To maintain strict honesty with our readers, we classify all research into three distinct tiers:
Rank Kiwi independently extracts, synthesizes, and models calculations from large published benchmark studies (such as datasets from Buffer, vidIQ, or Socialinsider). While the underlying census was collected by the cited publisher, the questions asked, derived equations, comparative ratios, visualizations, and creator implications are produced by Rank Kiwi.
Used exclusively when the data was genuinely collected through documented Rank Kiwi extension telemetry, aggregated user research, or internal database scans. We never apply this label to third-party studies.
Direct industry findings summarized for creator context. Numerical values are presented exactly as published by the original author, fully cited, and prominently linked.
2. Mandatory Study Transparency Elements
Every single research study published under the Rank Kiwi imprint must expose the following metadata elements directly in the body of the article:
3. Derived Value & Rounding Rules
Whenever Rank Kiwi derives a ratio, spread, or comparison from published source values, we do not present the result as if it were directly measured by the source. We show the exact equation:
Rounding & Precision Standards
- Ratios: Retain exactly two decimal places in methodology and analytical sections (e.g.,
10.82×,4.11×,2.08×). - Percentages: Display up to two decimal places when reflecting precise growth rates (e.g.,
0.13%,2.18%). - Probabilities & Rarity: State exact sample context (e.g., "Roughly 1 in 770 channels in this specific analyzed sample had at least 1M subscribers", rather than making universal census claims).
4. Causation vs Association & Algorithmic Folklore
We strictly enforce two non-negotiable editorial guardrails:
- “Posting 12 videos a month causes your channel to grow.”
- “The Instagram algorithm hates static images.”
- “Secret algorithm hacks to go viral in 30 days.”
- “Higher upload frequency was associated with higher median growth in this dataset.”
- “Reels achieved 36% more reach than carousels in Buffer's analyzed sample.”
- “Platform documentation highlights personalized viewer satisfaction signals.”
Recommendation systems (YouTube, Instagram, TikTok) are personalized ranking systems based on individual viewer behavior, satisfaction surveys, and completion signals. They do not operate on universal upload quotas. We ground all platform claims in official engineering documentation.
5. Median vs. Arithmetic Mean (Average)
Content performance follows power-law distributions. A single post with 5,000,000 views will violently distort the arithmetic average of a creator whose typical video receives 10,000 views.
Where third-party datasets provide arithmetic averages (such as Socialinsider's follower tier benchmarks), we label them clearly as averages and highlight the potential upward skew. Where datasets report medians (such as vidIQ's view growth studies), we preserve the median terminology. Rank Kiwi's own Outlier Score uses median baselines exclusively to ensure outlier detection remains robust against extreme hit distortion.
6. Corrections, Updates & Versioning
Scientific and statistical integrity requires transparent versioning:
- Publication Dates: We never alter the initial publication date of a research study when making minor edits or typo corrections.
- Updated Dates: Only updated when methodology changes, fresh source data is incorporated, or a material calculation is amended.
- Material Corrections: If an error is discovered in a calculation or attribution, we append an explicit correction notice explaining the change rather than silently rewriting history.