[DoreenMichele]

I don't really understand your point. This is what he means:

I often write a placeholder title, write the essay, and then at the very end, spend a good chunk of time iterating on titles until there’s a good one.

It isn't uncommon for me to handle blog posts that way as well, or even paid freelance articles if no title was provided. You put in a title that suggests what you are talking about, then you write the piece, then you try to figure out what the most important detail is or the most compelling hook or the shortest way to make the essential point.

Writing good titles is quite hard and good titles often grow out of the piece after it is written. You get to the end and you have written multiple paragraphs to give a good lead up and then your final paragraph draws a conclusion in a way you didn't have in mind when you started. And therein lies clues to a good title.

I fairly often pull ideas for a title from the last paragraph or two of a blog post. And I usually don't know what that last paragraph will be until I have written the entire piece.

[DigitalJack]

I must have missed that. Thanks for pointing it out, I was bothered by the seemingly contradictory advice :)

all I noticed was:

  >"Titles are 80% of the work, but you
  write it as the very last thing. 
  It has to be an compelling opinion 
  or important learning"

followed by

  >"Most of my writing comes from
  talking/reading deciding I strongly
  agree or disagree. These opinions
  become titles. Titles become essays."

and then the example

  >"The best example of this in
  my work is 'Growth Hacker is the
  new VP Marketing' which started
  out as a tweet with 20+ shares, and
  then was developed into an essay
  afterwards."