Author Playbooks

Updated monthly. Click any tile to expand the full guidance.

Directional Allison Palmer · Mind-Body

100% of articles use no tracked headline formula. Overall median is 76%ile—39 pts above team—but formula impact can't be measured without consistent format use.

→ Apply a repeatable headline formula (number lead, 'What to know', or possessive named entity) to at least the next 5 articles. That creates a testable signal. Best platform right now: Apple News (76%ile).

Details ↓
Directional Lauren Schuster · Experience

100% of articles use no tracked headline formula. Overall median is 44%ile—7 pts above team—but formula impact can't be measured without consistent format use.

→ Apply a repeatable headline formula (number lead, 'What to know', or possessive named entity) to at least the next 5 articles. That creates a testable signal. Best platform right now: Apple News (44%ile).

Details ↓
Directional Ryan Brennan · General/Discovery

96% of articles use no tracked headline formula. Overall median is 40%ile—3 pts above team—but formula impact can't be measured without consistent format use.

→ Apply a repeatable headline formula (number lead, 'What to know', or possessive named entity) to at least the next 5 articles. That creates a testable signal. Best platform right now: Apple News (52%ile).

Details ↓
Directional Hanna Wickes · General/Discovery

78% of articles use no tracked headline formula. Overall median is 27%ile—10 pts below team—but formula impact can't be measured without consistent format use.

→ Apply a repeatable headline formula (number lead, 'What to know', or possessive named entity) to at least the next 5 articles. That creates a testable signal. Best platform right now: SmartNews (34%ile).

Details ↓

Variant Production—Tracker Overview

Based on 422 clusters in the Tracker. Cluster = 1 original (Parent-P) + its CSA-generated variants (Child-C). This analysis uses the Tracker directly — no platform performance join required.

397 of 422 clusters (94%) have ≥2 articles (original + at least one variant). Mean cluster size: 2.85. Median: 3. Max: 6.

Cluster size distribution

Variants per cluster Clusters Share
1256%
215036%
313933%
48119%
5+276%

Variant production by author (top 10 by cluster count)

Author Clusters Mean variants/cluster
Ryan Brennan1543.1
Hanna Wickes1443.1
Lauren Schuster472.2
Lauren J-G402.0
Allison Palmer191.9
Samantha Agate112.9
Hanna WIckes42.0
Lauren JG21.0

Cluster size = 1 (original) + count of Child-C rows sharing the same Parent_ID. Only Parent-P rows are counted here—each represents one cluster. Child-C rows without a valid Parent_ID are excluded from cluster counts.