Predicted vs. real performance
Every publish feeds real metrics back into the model to recalibrate account- and niche-level weights. (Mock data in this MVP slice.)
How regular people use AI to boost productivity
A · Save-first
published
Predicted DAVS
86.8
58.0k
Views
2.3k
Likes
4.8k
Saves
360
Comments
980
Shares
720
Follows
Save rate well above the niche average — the checklist + end-of-video save CTA worked. Next: keep the serialized structure and add a stronger follow conversion.
Why most people fail at AI productivity
B · Comment-driven
published
Predicted DAVS
83.2
41.2k
Views
2.0k
Likes
1.5k
Saves
1.1k
Comments
540
Shares
410
Follows
Comment rate spiked from the contrarian hook; saves lagged. Best when the goal is reach via comments, not library-building.
Build an AI work system in 7 days — Day 1
C · Follow-first
published
Predicted DAVS
85.5
33.5k
Views
1.7k
Likes
2.1k
Saves
480
Comments
360
Shares
1.3k
Follows
Highest follow conversion of the three — serialization + 'Day 1/7' framing drove subscriptions. Strong base for an IP series.