I don't think it's fair to call that the "profitability" of the posts. The highest-earning promoted posts are mostly by authors whose posts all do well regardless.
What would make this a fairer comparison?
Maybe something like this:
- Take the author's previous 5 posts
- Find the average of those posts (median or mean)
- Take the value of the promoted post and subtract this average
- Compare the answer to how much was spent promoting that post
For instance:
A user's previous 5 posts earned: 200, 100, 0, 150, 50
The average of those 5 posts is $100.
This user's promoted post made $120 after spending $30 promoting it.
According to your calculations, this would be a $90 profit.
$120 (post earnings)-$100(average post)-$30(money spent promoting)= $10 loss
There are still some problems with this, but I think it's a much better way to calculate profitability.
Problems
- Some of the user's past 5 posts may have also been promoted; this would skew the results. You want to compare non-promoted posts.
- A 5 post sample is tiny, but you don't want to go back too far because a lot can change in weeks or a month. The math would get more complicated, but you could use the last 20 posts, giving more weight to recent posts.
- Users are likely to promote what they feel are their best posts, which are more apt to perform better regardless.
- Past posts may have earned more or less because the price of steem was higher or lower, respectively. Maybe this could be overcome by comparing the number of Steem earned, not the value of that Steem.
- Many followers may have been gained since 5 posts ago, thus making posts more likely to perform better.